How does a company earn more when revenue no longer grows? This chapter works through the levers at the same revenue, uses a field experiment to show what better processes measurably achieve, checks what studies say about cutting and reorganising, orders twelve levers with three worked examples and a calculator, and describes how to support and restructure firms without overlooking dependencies and people. Every key statement names its source (as of 29 October 2026); how well it is documented is in the tab “What holds up”. The chapter gives no recommendation for a particular firm.
When revenue no longer grows, progress comes from inside the firm: through more effectiveness (doing the right things) and more efficiency (doing things right). If revenue stays similar, profit rises only if internal processes get better and the company thinks differently. That is not a round of savings in the classic sense but a reordering of existing systems; it reaches into habits and responsibilities and therefore often meets resistance. This chapter works through the levers, checks what studies show about cutting and reorganising, and describes how to support and restructure firms without overlooking dependencies and people. It gives no recommendation for a particular firm.
Six fields, three questions
When revenue no longer grows, progress comes from inside the firm: through more effectiveness and more efficiency.
Three questions run through the chapter:
How to read this. The charts carry the statements, the text supports them. Every key statement names its evidence; calculations, invented examples and models are marked as such (labels at the top of every card).
With rating and the tabs in which they are checked
Five working hypotheses carry the chapter. Scenario: revenue stays at a similar level, yet profit is to rise. Each hypothesis is checked in the tabs named and receives a rating under the evidence rule of the encyclopaedia.
| No. | Working hypothesis | Checked in the tabs | Rating of the chapter |
|---|---|---|---|
| 1 | At the same revenue, profit can rise through prices, process costs, errors and the capital tied up. | Profit without growth, Field experiment, Three examples | Partly documented |
| 2 | Reorganising changes processes and systems, cutting only lowers costs. The effects of the two differ. | Cutting or reorganising, Twelve levers | Partly documented |
| 3 | Change hits people and habits, hence the resistance. It can be shaped but not defined away. | Staff, People | Partly documented |
| 4 | What looks like duplicated work from outside can serve a hidden task. Before any change the dependencies must be clear. | What growth leaves behind, Dependencies | Popular model |
| 5 | Whoever supports a firm first understands its processes, then sets a target picture and changes in steps that can be checked. | Supporting a firm, Restructuring | Partly documented |
How to read this. The rating is the judgement of this chapter: well documented, partly documented, simplifying, popular model or refuted. A click on the level opens the reasoning in the ladder in the tab “What holds up”.
The sixteen tabs after this overview
What the tabs show. A click on the name opens the tab.
| Tab | What it shows | Charts |
|---|---|---|
| Effectiveness and efficiency | What effectiveness and efficiency mean exactly, and why the order matters. | Four-field table |
| Profit without growth | Four levers at the same revenue, the profit bridge of an invented firm and what one per cent does. | Driver tree, profit bridge, bars |
| Field experiment | A randomised experiment with advice in Indian textile plants, with the p-value explained. | Bars with p-values |
| Cutting or reorganising | What studies show about downsizing, efficiency and restructuring. | Study matrix |
| Processes | Where time goes in a process, bottleneck logic and methods compared. | Time bars, sequence of steps |
| What growth leaves behind | Why organisational hurdles grow as a firm grows. | Bars, Greiner’s staircase |
| Twelve levers | Twelve levers for more margin, where they work and how well they are documented. | Lever map |
| Three examples | Three invented firms in saturated markets, worked through. | Three profit bridges, columns |
| Margin calculator | The profit bridge with your own figures. | Calculator |
| Staff | Staff: first order the work, then decide about posts; legal position in Austria. | Sequence of steps |
| Supporting a firm | Four approaches compared and a working model in five phases. | Comparison, phase model |
| Restructuring | Crisis stages, restructuring plan, early-warning indicators in Austria and time spans. | Stages, time bars |
| Dependencies | Hidden tasks, Meadows’ leverage points and questions to ask before any change. | Process scheme, ranking |
| People | What shapes reactions to change, and two common claims, checked. | Scheme of influences |
| What holds up | 23 statements with rating, state of evidence of the hypotheses, gaps. | Ladder |
| Glossary | The technical terms of the chapter in plain language, with a search. | List |
For whom, for what, with which limits
How to read this. Evidence levels belong to a statement (ladder in the tab “What holds up”), reading levels to a source (sources block), labels to a card. The three scales are not mixed.
Effectiveness asks whether a measure reaches the goal. Efficiency asks with how much effort it reaches it. Both count, but the order decides: efficiency at the wrong goal remains worthless. This tab clarifies the terms against the sources and fixes how the chapter uses them.
Four-field table of effect and effort
Whoever only lowers effort can move from the top left field to the top right, but also from the bottom left to the bottom right: then the work has become cheaper and still does not reach its goal. So check the effect first, then lower the effort. When revenue no longer grows, both are needed: effectiveness where prices, customers and offer limit the result, efficiency where processes, errors and capital tied up eat into it.
Gabler, Drucker, Porter
The definitions in the Gabler business dictionary are narrower than the well-known short formula “doing the right things, doing things right”. There, effectiveness means: a measure is suitable to reach a given goal; the term says nothing about the way in which the goal is reached (the part on health care management adds explicitly: regardless of the effort required). Efficiency means: the measure is suitable to reach the goal in a certain way, for example while staying economical. In both definitions the goal is given; they do not ask whether it is the right one. That is a question of strategy (tab Supporting a firm).
The word pair of the short formula already appears in Peter Drucker: in his article “Managing for Business Effectiveness” (Harvard Business Review, May 1963) he writes, according to the excerpt at Quote Investigator, of the confusion between effectiveness and efficiency “that stands between doing the right things and doing things right”, and, in substance, that nothing is as useless as doing very efficiently what should not be done at all. Whether the short formula as a definition originated there is not examined by Quote Investigator; the page only traces the origin of the second sentence. In 1996 Porter separates “operational effectiveness” (performing similar activities better than rivals) from strategic positioning (choosing different activities or performing similar ones differently). Better execution alone evens out among competitors, because good practices spread quickly, and its benefits, according to Porter, often go to customers and to the makers of the equipment (his example is the printing industry in the United States).
A caution about the words: Porter’s “operational effectiveness” covers efficiency in the narrow sense and further improvements such as fewer defects or faster development. It is therefore closer to what this chapter calls efficiency than to effectiveness in the sense of the German dictionary.
| Source | How it uses the terms | Note on the evidence |
|---|---|---|
| Gabler, Effektivität (effectiveness) | Suitability of a measure to reach a given goal; says nothing about the way it is reached (in the part on health care management: regardless of the effort). | Dictionary entry; the page gives no date. |
| Gabler, Effizienz (efficiency) | Suitability to reach the goal in a certain way, for example while staying economical. | Dictionary entry; the page gives no date. |
| Drucker 1963 | Confusion between effectiveness and efficiency “between doing the right things and doing things right”; in substance, nothing is as useless as doing very efficiently what should not be done at all. | Excerpt at Quote Investigator (secondary source); we did not read the article itself. The page does not examine the short formula as a definition. |
| Porter 1996 | “Operational effectiveness”: performing similar activities better. Strategic positioning: different activities, or similar activities performed differently. | Article in the Harvard Business Review, read in large parts on the publisher’s page. |
In this chapter: effectiveness means that measures reach the goals that were set, efficiency that they do so with little effort. First check the effect, then lower the effort.
Rating. Statement on this topic in the ladder, with rating: Effectiveness means doing the right things, efficiency doing things right.
Evidence. Gabler, Effektivität (effectiveness); Gabler, Effizienz (efficiency); Quote Investigator on Drucker 1963; Porter 1996. The full references are in the sources block below.
If revenue stays the same, the simple calculation leaves four levers: price quality and mix, variable costs, fixed costs and the capital tied up. Arithmetically that is certain. What is open is how large the effect is and whether it lasts. This tab works the logic through for an invented firm; what a measurement says about it is in the tab “Field experiment”.
Driver tree of the profit before tax
Profit before tax = revenue − variable costs − fixed costs − interest
The calculation is an accounting identity. It says nothing about the size and duration of an effect; that needs evidence, which follows in the next tabs. Hardly any lever works in one place only: fewer errors save material and working time, less stock lowers interest and the space needed.
Rating. Statement on this topic in the ladder, with rating: At the same revenue, prices, variable and fixed costs and the capital tied up move the profit.
Income statement of an invented year
Example Ltd is invented; its figures only show the logic of the calculation. It has a revenue of 10 million euros a year and earns 5% of it before tax.
| Item (Example Ltd, invented) | Thousand euros | Share of revenue |
|---|---|---|
| Revenue | 10,000 | 100.0% |
| variable costs (materials, bought-in services, freight) | −6,500 | 65.0% |
| Contribution margin | 3,500 | 35.0% |
| fixed costs (staff, premises, IT, administration) | −2,900 | 29.0% |
| Operating profit | 600 | 6.0% |
| Interest | −100 | 1.0% |
| Profit before tax | 500 | 5.0% |
Own calculation, invented example. The contribution margin is what is left after the variable costs to carry the fixed costs.
Example Ltd, invented figures in thousand euros
The profit bridge shows what the levers of the driver tree contribute in the example: profit goes from 500 to 685 thousand euros although revenue stays the same. The one-off implementation costs of 40 weigh on the first year; from the following year they fall away, and profit would then be 725, 45% more than today. The example shows two things: each lever changes a different line of the calculation, and change costs money before it works.
| Step | What lies behind it (invented) | Calculation in thousand euros |
|---|---|---|
| Price discipline | Discounts without anything in return are dropped; on average 0.4% more of revenue remains. | 10,000 × 0.4% = +40 |
| Purchasing and offcuts | Pooled purchasing and better cutting lower material costs. | 6,500 × 0.77% ≈ +50 |
| Errors and rework | Errors cost 300 a year today (scrap, rework, credit notes); a fifth of that falls away. | 300 × 20% = +60 |
| Lead time | Shorter waiting times save overtime, express freight and time spent searching. | +50 a year (assumption) |
| Capital tied up | Stock falls by 500; interest and storage costs on the capital tied up are 5% a year. | 500 × 5% = +25 |
| One-off costs | Training, advice and changeover, in the first year only. | −40 |
Rating. Statement on this topic in the ladder, with rating: Working hypothesis 1: at the same revenue, profit can rise through prices, process costs, errors and the capital tied up.
Effect of one per cent each on the profit of Example Ltd
One per cent more on the price works most strongly in the example, because it arrives directly in the profit without any additional cost. The order depends on the cost structure. The leverage of a price change is the reciprocal of the margin: at a margin of 5%, one per cent of price brings 20% more profit (own calculation, holds at the same volume and costs). This also works the other way: whoever cuts prices loses in the same way. More volume works less strongly than more price, because variable costs rise with volume; only the contribution margin counts.
In 2003 McKinsey works it through on the average income statement of a company in the S&P 1500 share index (Standard & Poor’s), a model calculation and not a measurement: a one per cent higher price at the same volume would raise operating profit by about 8%, nearly half as much again as one per cent lower variable costs and more than three times as much as one per cent more volume. For price, variable costs and volume that is the same order as in the example; the text we read gives no figure for fixed costs. Nor does it state the margin on which the 8% rest; arithmetically it implies a margin of about 12.5% (own calculation). McKinsey is a consultancy source and not an independent study.
Rating. Statement on this topic in the ladder, with rating: One per cent more on the price works more strongly than one per cent less cost, and at a low margin every lever works strongly.
Evidence. McKinsey 2003. The full references are in the sources block below.
The figures of this tab are invented and show the logic of a calculation. Whether a lever brings as much in a real firm only its own measurement shows; the chapter gives no recommendation for a particular firm.
The strongest empirical support for better processes comes from a randomised experiment: in Indian textile plants, a lottery decided which plants received free advice on management practices. This tab shows the results, explains what the p-values mean there and names the limits of the experiment.
Estimated effect of the advice against the control plants, from Table II of the study
Bloom, Eifert, Mahajan, McKenzie and Roberts (Quarterly Journal of Economics, 2013) selected 20 plants in 17 textile firms near Mumbai. 14 of them received five months of free advice: one month of diagnosis and four months of implementing 38 management practices in five areas (production, quality control, inventory, human resources, and sales and orders). 6 control plants received only the diagnosis and then measurement visits with little advice.
Compared with the control plants, quality defects fell by 43.1% and inventory by 21.7%, output rose by 9.4% and productivity by 16.6% (given in the introduction of the study as 17% in the first year). Productivity here means: output measured against capital and labour hours together, in the study total factor productivity. The authors estimate the profit at about 325,000 US dollars per plant and year; it was not measured, because they could not obtain accounting data on the firms’ profits and losses. They inferred it from the measured improvements in quality, inventory and efficiency.
Evidence. Bloom et al. 2013; Bloom et al. 2013, summary. The full references are in the sources block below.
A permutation test, explained in words
The p-value answers a single question: how often would a difference at least this large come out if the advice had had no effect and only the random assignment counted? In the field experiment a lottery decided which firms were advised. The usual calculation estimates from the spread of the data how strongly a result would vary by chance alone (the standard error); with only 17 firms that estimate is uncertain. The authors therefore also calculated an exact test: they assumed that the advice had no effect and worked through every possible assignment of the advice to the firms (a permutation test). The p-value of this test is the share of the assignments in which a difference at least as large comes out. The p-values in this tab come from this test.
| Measure | Effect | p-value in the permutation test | in words | significant at the threshold of 0.05? | usual calculation (standard errors in Table II): significant |
|---|---|---|---|---|---|
| Quality defects | −43.1% | 0.001 | about 1 in 1,000 assignments | yes | at the threshold of 0.05 |
| Inventory | −21.7% | 0.060 | 6 in 100 assignments | no, just above | at the threshold of 0.05 |
| Output | +9.4% | 0.026 | just under 3 in 100 assignments | yes | at the threshold of 0.05 |
| Productivity | +16.6% | 0.061 | just over 6 in 100 assignments | no, just above | only at the threshold of 0.10 |
Table II gives p = 0.001 for quality defects; the text of the study gives 0.01 at one point. Both values are below 0.05. The effects in per cent are converted from the values in the table, as the authors do in the text.
What the p-value does not say. It says nothing about how large or how important an effect is; that is what the effect in per cent beside it is for. It is not the probability that the result is chance, and a p-value just above 0.05 does not mean there is no effect, but that the experiment cannot separate it from chance with enough certainty.
How to read this. Example inventory: p = 0.060 means that in 6 out of 100 assignments a difference at least as large would come out if the advice had had no effect. In the permutation test this counts as narrowly not significant by the usual threshold of 0.05 and as significant by the looser threshold of 0.10; with the usual standard errors of the table, inventory is significant at 0.05. With values this close, the verdict depends on the method of calculation; the direction of the effect is the same in both.
What the study shows and what it does not
The limits are clearly stated: one industry in one region, only 17 of 66 firms approached took part, and the sample is small. In the permutation test, inventory and productivity lie just above the usual threshold of 0.05 with p = 0.060 and 0.061; with the usual standard errors of the table, inventory is significant at 0.05 and productivity only at 0.10. The consultancy collected the data itself; the research team supervised the collection.
The study shows that better processes can work measurably in a plant. It does not show how large the effect is in other industries, and it does not separate how much each of the 38 practices contributes. In the experiment the control plants also received the diagnosis; the differences measured are therefore mainly due to the supported implementation (tab Supporting a firm).
Rating. Statements on this topic in the ladder, with rating: Supported introduction of management practices reduced defects and raised output; Working hypothesis 1: at the same revenue, profit can rise through prices, process costs, errors and the capital tied up.
Cutting lowers costs, reorganising changes how work runs. Whoever only cuts jobs gains no significant advantage over the industry in the studies we read. Firms do better when they restructure their assets along with the cuts, when they save through efficiency rather than cuts and invest at the same time, or when they improve their processes.
Findings from six studies, our own division
The four levels are a division by this chapter. Cutting saves across the board. Improving streamlines processes within the existing system. Reorganising redraws responsibilities, interfaces and systems. Restructuring changes business, assets or financing. The gap in the column “Reorganising” says something about the research, not about practice: for the reordering of responsibilities and systems none of the studies we read is direct. The tab “Processes” relies there on methods and the field experiment, the tab “Dependencies” on systems thinking; both are models and not measured findings.
Rating. Statement on this topic in the ladder, with rating: Working hypothesis 2: reorganising changes processes and systems, cutting only lowers costs; the effects of the two differ.
Five studies on downsizing, efficiency and recovery
| Study | Sample | Finding | Limits |
|---|---|---|---|
| Guthrie and Datta 2008, Organization Science | US manufacturing firms; size not stated in the abstract we read | Downsizing through layoffs went together with lower later profitability, especially in industries with a high share of research, high growth and low capital intensity. | Abstract read; an association, not a demonstrated cause. |
| Slavković et al. 2021, Acta Economica | Survey of managers from 75 companies | Staff cuts were statistically significantly associated with weaker performance. Knowledge management was associated with better performance and weakened the negative association during the cuts. | Small sample, self-report, a one-off survey (association); the abstract does not name the country. |
| Datta, Guthrie, Basuil and Pandey 2010, Journal of Management | Review of studies on the causes and effects of downsizing | The review orders the causes and effects of downsizing and names recurring patterns and contradictions in the research; the abstract we read says nothing about the effect on performance. | Only the abstract was read. |
| Cascio, Young and Morris 1997, Academy of Management Journal | 5,479 employment changes at companies in the S&P 500 share index (Standard & Poor’s), 1980 to 1994 | Pure downsizing brought no significantly higher returns than the average of the firm’s own industry. Downsizing combined with a restructuring of assets brought a higher return on assets and higher share returns than the industry. | Abstract read through a secondary source; large US companies. |
| Gulati, Nohria and Wohlgezogen 2010, Harvard Business Review | 4,700 listed companies, three recessions (1980 to 1982, 1990 to 1991, 2000 to 2002) | Of the companies that survived the recession, about 80% needed more than three years to regain their earlier growth rates in revenue and profit. Only 9% of the surviving firms did better afterwards on key financial measures than before the recession and beat the rivals in their industry by at least 10% in growth of revenue and profit. According to the authors, the best prospects of joining them came from a combination: cutting costs mainly through operational efficiency rather than through more job cuts than comparable firms, and investing significantly more than rivals in research and marketing, while also investing in plants and machinery. | Only the summary was read; an association in times of crisis, listed companies. |
Conclusion. The studies are associations, not randomised experiments (except the field experiment by Bloom et al.), and several look at large listed companies. For small and medium-sized firms it is open whether they carry over. Together they show one direction: cuts alone bring no significant advantage; firms do better when they restructure their assets along with the cuts (Cascio et al.) or save through efficiency instead of deeper cuts and invest at the same time (Gulati et al.).
Rating. Statements on this topic in the ladder, with rating: Downsizing alone brings no significant advantage over the industry; The best prospects of coming out of a recession stronger lay with firms that saved through efficiency and invested at the same time.
Evidence. Guthrie and Datta 2008; Slavković et al. 2021; Datta et al. 2010; Cascio, Young and Morris 1997; Gulati, Nohria and Wohlgezogen 2010. The full references are in the sources block below.
Whoever wants to improve processes first needs a picture of how work actually flows: where it waits, where it is done twice and where a narrow point limits throughput. Methods help with this; they do not replace understanding.
Order handling at an invented firm, times in working days
In the invented example the order waits 10 of 12.5 days; it is worked on for only 2.5 days, that is 20% of the lead time. That is the core idea of value stream mapping in the lean approach: time in which nobody works on the order is of no use to the customer. Whoever only speeds up the work steps can at most gain part of the 2.5 days in the example; the waiting remains. Which shares hold in a real firm only its own measurement shows. This chapter gives no standard value.
Theory of Constraints after Goldratt, in substance after the TOC Institute
↺ after step 5 back to step 1
Bottleneck logic adds a rule about order to the value stream: improving pays off where the throughput of the whole system is limited. In the example the order waits longest before manufacturing (3.5 days). That is where a narrow point would be suspected, which a measurement must confirm or refute.
Which question each answers, and how well it is documented
| Method | Answers the question | Prerequisite (assessment of the chapter) | Evidence and limits |
|---|---|---|---|
| Value stream mapping (lean) | Where does work wait, where do duplication and hand-overs arise? | Someone draws the actual process from receipt to delivery. | Womack and Jones 1996: five steps (value, value stream, flow, pull, perfection). The page of the Lean Enterprise Institute that we read gives no figures or studies on effects. |
| Bottleneck analysis (Theory of Constraints) | Which point limits throughput, and what follows from it? | Data or observation of where work piles up. | Goldratt, described in the novel “The Goal”. The page of the TOC Institute gives only an unsupported rule of thumb (the first three steps typically expose at least 30% hidden capacity), no study. |
| Process mining | How does the process actually run, and where does it deviate from the plan? | Event logs from the IT systems. | Manifesto of the IEEE Task Force on Process Mining (2011): discovering process models from event logs, checking them against the actual process and extending them. A position paper, not a study of effects. |
| Redesigning processes (reengineering) | Is the process fundamentally built the wrong way? | · | Hammer 1990 in the Harvard Business Review: only the title and short text were read, so no statement on effects here. Kotter 1995 observed such programmes (tab “People”). |
| Introducing proven management practices | Which practices are missing, for instance in quality control, stock and order handling? | Diagnosis and supported implementation. | Bloom et al. 2013 (tab “Field experiment”): randomised experiment with 38 practices in one industry. |
None of the methods is evidence that improving works; they order where to look. The evidence that better processes can work rests, in this chapter, on the field experiment.
Rating. Statement on this topic in the ladder, with rating: Value stream and bottleneck logic show where improving pays off.
Evidence. Lean Enterprise Institute; TOC Institute; Process Mining Manifesto 2011; Hammer 1990; Bloom et al. 2013. The full references are in the sources block below.
As a firm grows, its organisational hurdles grow too: more posts that coordinate, more rules, approvals, reports, variants and systems, each of which solved a problem of its time. While revenue grows, their costs are spread over more revenue. If it stays the same, this dilution falls away, and every superfluous round of coordination comes straight out of the margin. When the market is saturated, the room for manoeuvre therefore lies inside the firm, in what growth has left behind.
Possible pairs, mathematics and not a measurement
possible pairs = n × (n − 1) ÷ 2; with 10 people 10 × 9 ÷ 2 = 45, with 100 people 100 × 99 ÷ 2 = 4,950
With every new person, department or level, the number of points that could coordinate grows faster than size itself: ten times as many people, a good hundred times as many possible pairs. Not every one of these links is needed; organisations build departments, levels and rules precisely to bundle the lines. These bundles in turn cost coordination. That is a consideration of this chapter, not a measurement.
Rating. Statement on this topic in the ladder, with rating: As a firm grows, its organisational hurdles grow.
Model after Greiner 1972, our own drawing without scale
According to the secondary source, Greiner described in the Harvard Business Review in 1972 that organisations grow in phases that each lead into a crisis. The publisher’s short description of the 1998 reprint says: management practices that work well in one phase can trigger a crisis in another. The names of the phases and crises come from the secondary source; the article itself was not read. We did not find a test of the model against data in this web search.
Rating. Statement on this topic in the ladder, with rating: Organisations grow in phases that each lead into a crisis.
Variety in a car plant, bureaucracy in a survey
Two findings support the idea from another angle. In a General Motors car plant, more strongly fluctuating option content went together with more total labour hours and more overhead hours per car, more downtime, rework and inventory, but in the short run not with more direct labour hours; stations with stronger fluctuation did, however, hold more staff as a buffer. The average option content was not related to it (Fisher and Ittner 1999, abstract read, one plant). Variety therefore also costs at places that are not directly attributed to a single variant (reading of this chapter).
In a survey to which the Harvard Business Review had invited its readers, Hamel and Zanini (2017) describe bureaucracy, after more than 7,000 responses, as markedly stronger in larger companies and as apparently increasing (according to the summary). The results with their figures are behind the paywall; we read only the summary and the introduction. Because participation was self-selected, the responses are not a random sample (classification by this chapter).
What follows from this is ordered in the tab “Twelve levers”: variants, approvals, double entry and waiting times are levers of their own there.
Rating. Statement on this topic in the ladder, with rating: Strongly fluctuating option variety goes together with costs at places to which it is not directly attributed.
Evidence. Fisher and Ittner 1999; Hamel and Zanini 2017; Greiner 1972; Wikipedia on Greiner. The full references are in the sources block below.
When the market is saturated, the room for manoeuvre lies inside the firm. This tab orders twelve levers that growth often leaves behind: discounts, small orders, variants, errors, purchasing, waiting times, bottlenecks, approvals, stock, receivables, management practices and staff. For each there is an invented example, the measure by which to track it, the point at which it can do harm, and what speaks for it in the sources we read.
Twelve levers, where they work in the calculation and how well they are documented
| Lever (number as in the table below) | Line of the income statement in which the lever works | State of evidence | |||
|---|---|---|---|---|---|
| Price and mix | variable costs | fixed costs | capital and interest | ||
| 1 Review discounts and special terms | works directly | consultancy source | |||
| 2 Price small orders to cover their costs | works directly | works indirectly | own calculation | ||
| 3 Clean up variants and the product range | works indirectly | works indirectly | works directly | works directly | one plant, indirect |
| 4 Reduce errors, scrap and rework | works directly | works indirectly | field experiment | ||
| 5 Pool purchasing, review specifications | works directly | works indirectly | no source read here | ||
| 6 Cut waiting times and hand-overs | works directly | works indirectly | thinking model | ||
| 7 Exploit the bottleneck, then invest | works indirectly | works directly | thinking model | ||
| 8 Cut approvals and double entry | works directly | survey, indirect | |||
| 9 Lower stock levels | works indirectly | works directly | field experiment, study | ||
| 10 Collect receivables faster | works directly | study | |||
| 11 Introduce management practices | works indirectly | works indirectly | works indirectly | works indirectly | field experiment, survey of firms |
| 12 Redeploy or reduce staff | works directly | studies, tab “Staff” | |||
Hardly any lever works in one place only: fewer errors save material and working time, fewer variants save set-up time and stock. The assignment is our own, not a measurement; it helps to look at the right lines of the calculation in a real firm.
Rating. Statement on this topic in the ladder, with rating: The twelve levers of the catalogue can raise the margin at the same revenue.
Twelve levers in detail, examples invented
| Lever | Example (invented) | What to measure | What to watch | Evidence |
|---|---|---|---|---|
| 1 Review discounts and special terms | A wholesaler finds that cash discount, annual bonus and free freight together cost 6% of the list price without anyone seeing all three at once. | Price actually received after all discounts, per customer and product | Keep discounts with a real return (volume, prepayment) | Marn and Rosiello 1992: volume discounts, cash discounts and loyalty bonuses make the price shrink between invoice and payment (only the publisher’s description read); McKinsey 2003, consultancy source with model figures (both with Michael V. Marn as an author) |
| 2 Price small orders to cover their costs | Every order costs a metalworking firm about €140 for offer, planning, set-up and invoice, whatever its size. Below an order value of €600 hardly anything is left. | Contribution margin per order by order size | Small orders that grow into large ones | Own calculation, no study read here |
| 3 Clean up variants and the product range | 12 of 40 colour variants bring 2% of revenue together but tie up stock and set-up time. In future they are available to order only. | Contribution margin, stock and set-up time per variant | Variants that carry the purchase of the main product | Fisher and Ittner 1999, one car plant: more strongly fluctuating option variety went together with more overhead hours, rework, downtime and inventory, the average option content did not; as a remedy the abstract names bundling options, not cutting variants (only indirect for this lever) |
| 4 Reduce errors, scrap and rework | A plant records every piece of rework with its cause for a month. A checklist at cutting removes two of the three most frequent causes. | Error rate, cost of rework and credit notes | Do not hide errors, or only the reported figure falls | Bloom et al. 2013: quality defects −43.1%, field experiment in one industry |
| 5 Pool purchasing, review specifications | Five departments buy office supplies, licences and vehicles separately. Pooled and put out to tender again, prices fall and duplicate licences go. | Spending per category, number of suppliers | Dependence on one supplier, quality | No study read here (advisory practice) |
| 6 Cut waiting times and hand-overs | The order in the tab “Processes” waits 10 of 12.5 days. Approvals are bundled, the order runs through without being put aside. | Lead time, share of waiting time, number of hand-overs | Waiting time that serves a hidden task (tab “Dependencies”) | Value stream (lean), thinking model; the page we read gives no figures or studies on effects |
| 7 Exploit the bottleneck, then invest | The paint shop is full. Before a second one is bought, breaks are staggered, set-ups are prepared and faulty parts are sorted out beforehand. | Throughput and downtime at the bottleneck | The bottleneck moves; measure again afterwards | Theory of Constraints, thinking model; the page we read gives only an unsupported rule of thumb, no study |
| 8 Cut approvals and double entry | Every purchase order over €500 needs two signatures, orders from the web shop are transferred to accounting by hand. | Time per case, number of approval levels, transfer errors | Approvals that check something on the side (tab “Dependencies”) | Hamel and Zanini 2017: survey on bureaucracy in general, stronger in larger companies; only the summary and introduction read, no figures on savings (only indirect for this lever) |
| 9 Lower stock levels | Three months of stock are held for every article. For fast-moving articles two weeks are enough. | Days of stock cover, shortages | Ability to deliver; exempt bottleneck parts | Bloom et al. 2013: inventory −21.7% (p = 0.060 in the permutation test, significant at 0.05 with the usual standard errors); Deloof 2003: fewer days of stock went together with higher profitability |
| 10 Collect receivables faster | Invoices go out at the end of the month, reminders after 60 days. In future the invoice goes out with the delivery, the reminder after 14 days. | Days until payment is received | Customer relationship, agreed payment terms | Deloof 2003: 1,009 large Belgian non-financial firms, 1992 to 1996, shorter receivable periods went together with higher profitability |
| 11 Introduce management practices | A plant puts up three figures every morning (output, errors, on-time delivery) and discusses deviations for ten minutes. | Whether the figures are kept up and discussed | A few figures the team can influence itself | Bloom and Van Reenen 2007: survey with a newly developed instrument of 732 medium-sized manufacturing firms in four countries, better practices went together with productivity, profitability, growth and survival; Bloom et al. 2013 (field experiment). Which single practice works, the sources we read do not separate. |
| 12 Redeploy or reduce staff | After the reorganisation, two posts of transfer and checking work fall away. One is not refilled at the next retirement, one person moves to customer service, where work is piling up. | Staff costs per unit of output, knowledge that depends on single people | First order the work, then the posts; employment law | Cascio et al. 1997, Guthrie and Datta 2008, Slavković et al. 2021 (tabs “Cutting or reorganising” and “Staff”) |
For errors, stock, receivables and management practices there are studies, for price discipline consultancy sources, for value stream and bottleneck thinking models, for small orders and purchasing no source read here. The examples are invented; how much a lever brings in a real firm only its own measurement shows. Three firms are worked through in the tab “Three examples”, your own figures are calculated by the margin calculator.
Rating. Statements on this topic in the ladder, with rating: The twelve levers of the catalogue can raise the margin at the same revenue; Shorter receivable and inventory periods go together with higher profitability; Better management practices go together with productivity and profitability.
Evidence. Marn and Rosiello 1992; McKinsey 2003; Fisher and Ittner 1999; Bloom et al. 2013; Hamel and Zanini 2017; Deloof 2003; Bloom and Van Reenen 2007; Cascio, Young and Morris 1997; Guthrie and Datta 2008; Slavković et al. 2021. The full references are in the sources block below.
Three invented firms in saturated markets: a metalworking firm, a wholesaler and an agency for marketing and design. For all of them revenue stays the same or changes by less than 1% through price steps. Each profit bridge shows what the individual levers contribute; the dashed column shows the profit from the second year, when the one-off costs fall away. All figures are invented and in thousand euros.
Invented firm, revenue 6,000 thousand euros
The metalworking firm cuts its cost of errors by a quarter, charges a handling fee on small orders, takes downtime and overtime off the paint shop and lowers its material stock. Profit rises from 180 to 288, from the second year to 336. Revenue grows only through the fee, by 42 to 6,042; the return is calculated on this revenue.
Invented firm, revenue 15,000 thousand euros
The wholesaler drops discounts without anything in return, keeps 600 rarely sold articles only to order, collects its receivables 15 days earlier and, with an interface from the web shop to accounting, saves a planned hire. Profit rises from 225 to 375, from the second year to 405. At a starting return of 1.5% every lever works strongly: half a per cent more of revenue alone brings a third more profit.
Invented firm, revenue 1,600 thousand euros
The agency charges a surcharge for work outside the order, fixes two rounds of corrections in the offer, builds offers from templates and cancels unused licences. The hours set free replace bought-in hours at €80. Profit doubles in the first year from 64 to 128 and rises from the second to 142; revenue stays the same, because most customers do without the extra requests.
Return on sales today, in the first and from the second year
| Example | Step | Assumption (invented) | Calculation in thousand euros a year |
|---|---|---|---|
| Metalworking, revenue 6,000 | Errors, rework | Errors cost 240 a year (4% of revenue), a quarter of that falls away | 240 × 25% = +60 |
| · | Small orders | 300 small orders a year carry a handling fee of €140 in future | 300 × 0.14 = +42 |
| · | Bottleneck: paint shop | The plant stands still less often, overtime and outside painting fall | +30 |
| · | Lower stock | Material stock falls by 300, interest and storage costs are 8% a year | 300 × 8% = +24 |
| · | One-off costs | Checklists, training, advice | −48 |
| Wholesaler, revenue 15,000 | Terms and discounts | Discounts without anything in return are dropped, that is 0.5% of revenue | 15,000 × 0.5% = +75 |
| · | Product range | 600 rarely sold articles are available to order only, stock falls by 375, interest and storage costs 8% | 375 × 8% = +30 |
| · | Receivables | Customers pay after 40 instead of 55 days, capital tied up falls by about 616, interest 5% | 15,000 × 15 ÷ 365 × 5% ≈ +31 |
| · | Double entry | An interface from the web shop to accounting makes a planned hire (0.8 posts at 55) unnecessary | 0.8 × 55 = +44 |
| · | One-off costs | Programming the interface | −30 |
| Agency, revenue 1,600 | Extra requests | Work outside the order is available only for a surcharge; most customers do without it, the 375 hours set free replace bought-in hours at €80 | 375 × 0.08 = +30 |
| · | Rounds of corrections | Clear briefs and two rounds of corrections in the offer save 300 bought-in hours | 300 × 0.08 = +24 |
| · | Offers from modules | Offers from templates instead of freely written ones save 200 hours | 200 × 0.08 = +16 |
| · | Licences | Unused software licences are cancelled | +8 |
| · | One-off costs | Modules, templates, training | −14 |
How to read the examples. They are invented and not typical cases. They show the logic of the calculation from the tab “Profit without growth”: the lower the starting return, the more strongly every improvement works on the profit; at the agency it doubles in the first year. Whether a lever brings as much in a real firm only its own measurement shows; the studies we read give no sizes for single levers that could be carried over to other firms. The most important assumption lies in the price steps: the calculation assumes that customers stay when discounts are dropped or small orders carry a fee. That must be checked beforehand, for instance first with a small group of customers.
Rating. Statements on this topic in the ladder, with rating: Working hypothesis 1: at the same revenue, profit can rise through prices, process costs, errors and the capital tied up; One per cent more on the price works more strongly than one per cent less cost, and at a low margin every lever works strongly.
Your own figures are calculated by the margin calculator: it takes the three firms and Example Ltd as templates.
The calculator builds the profit bridge from your figures: revenue and profit today, plus seven levers for one year. It calculates what you enter and assumes that everything else stays the same, above all volume and customers. It does not say what will happen and replaces neither planning nor advice.
Amounts in thousand euros, one year
Choose a template or enter your own figures; amounts in thousand euros. Every change recalculates at once; the profit bridge and the table below show each step with its calculation.
How to read this. In the calculation the levers work independently of each other and add up; in a real firm they can overlap (fewer errors also lower the stock) or slow each other down. The most important assumption is that customers stay when prices or terms change. That must be checked beforehand, for instance first with a small group of customers.
The calculator is a model with openly stated assumptions; its result is neither a promise nor a forecast. For a real firm, its own measurement before and after counts.
Staff cuts are one possible lever, but according to the studies we read not one that carries on its own. What follows is an order, not a rule: first clarify which work is still needed after the reorganisation, then plan the posts and first use the routes where nobody has to leave. Where cuts are still needed, the how counts, and in Austria there are rules of their own for it.
Our own order, derived from four studies
In Cascio, Young and Morris (1997) pure downsizing brought no significantly higher return than the industry, while downsizing combined with a restructuring of assets did. In Guthrie and Datta (2008) downsizing went together with lower later profitability. In the survey by Slavković et al. (2021) staff cuts went together with weaker performance, and knowledge management softened this association during the cuts. And according to Gulati, Nohria and Wohlgezogen (2010), companies had the best prospects of coming out of a recession stronger when they cut costs mainly through efficiency rather than through more cuts than comparable firms, and invested at the same time. All four show associations, not experiments (tab “Cutting or reorganising”).
Rating. Statement on this topic in the ladder, with rating: Downsizing alone brings no significant advantage over the industry.
How it is done, and knowledge
How a change is carried out is, according to a review of 79 studies, among the influences on how those affected react (Oreg, Vakola and Armenakis 2011, tab “People”). And knowledge that depends on single people leaves with them; knowledge management softened the negative association between staff cuts and performance in Slavković et al.
How to read this. The abstract of the review that we read does not say how strongly each factor works; acting early, openly and with knowledge secured is a conclusion of this chapter from the findings named, not a measured recipe.
Evidence. Oreg, Vakola and Armenakis 2011; Slavković et al. 2021. The full references are in the sources block below.
Works council and early-warning system, legal position
Works council. Where a works council exists, it must be notified before every dismissal. It can give its opinion within one week, and a dismissal given before this period has expired is legally ineffective, unless the works council has already given its opinion (§ 105 of the Labour Constitution Act, Arbeitsverfassungsgesetz; state of legislation according to the page 26 April 2026).
Early-warning system. Whoever intends to end the employment of several employees within 30 days, for instance by dismissal, must above a certain number first notify the regional office of the Public Employment Service (Arbeitsmarktservice; early-warning system under § 45a of the Labour Market Promotion Act, Arbeitsmarktförderungsgesetz). The page of the Supreme Court of Justice names as a threshold at least five employees who have reached the age of 50; whether further thresholds apply is set out in § 45a itself, which we did not read. In the case decided, an employer had dismissed seven employees over 50 without notification and then dismissed the same people again in several steps. The lower courts declared the second dismissal of the claimant ineffective, and the Supreme Court rejected the appeal: the steps rested on a single intention to end the employment relationships (9 ObA 33/21z of 29 April 2021).
Legal position according to the pages read, not checked by a legal expert. Have any staff measure checked under employment law beforehand; this chapter does not replace legal advice.
Evidence. § 105 Labour Constitution Act; Supreme Court of Justice, 9 ObA 33/21z. The full references are in the sources block below.
Supporting a firm means working in an order: first understand, then set a target picture, then change in steps that can be checked and finally anchor the change. Four very different approaches all begin with analysis, and in the field experiment the lead was measured against control plants that also received the diagnosis, so it is due to the supported implementation. Support makes visible how the firm actually works, orders decisions and checks assumptions before money is spent; whoever provides it does not change things themselves but helps the firm to change (assessment of this chapter).
Reorganisation, restructuring plan, Theory of Constraints, field experiment
The four approaches come from very different worlds: a dictionary entry on the reorganisation of organisations, a German professional standard of public auditors for restructuring plans (for IDW S 6 these are the components in the order in which they are listed, not working steps), a method for bottlenecks and a field experiment. What they share: the situation is surveyed before changes are decided and carried out; in Gabler the approach itself is planned before that. In the field experiment the control plants also received the one-month diagnosis; the differences measured are therefore mainly due to the four months of supported implementation. A diagnosis alone therefore does not replace implementation (conclusion of this chapter).
Evidence. Gabler, Reorganisation (reorganisation); Tabak on IDW S 6; TOC Institute; Bloom et al. 2013. The full references are in the sources block below.
Our own working model, derived from the four approaches
↺ next round: the measurement shows whether the understanding was right
The working model puts the shared steps together. Two phases hardly appear in the sources: the target picture (as a guiding picture only in the restructuring standard) and anchoring (not a step of its own in any of the approaches; the fifth step of the Theory of Constraints and the measurement phase of the field experiment support only measuring again). They stand here all the same, because without a goal one cannot check whether a change works, and because without anchoring the effect can be lost again. That is a reasoning of this chapter, not a finding.
| Phase | Result that can be checked | Supported by |
|---|---|---|
| 1 Understand | Picture of the actual processes with figures: waiting times, errors, bottleneck, dependencies | Gabler (recognise the problem, survey the current state), IDW S 6 (starting situation, crisis stage and causes), Theory of Constraints (identify), Bloom et al. (diagnosis) |
| 2 Set a target picture | A measurable goal and limits, for instance: halve the lead time, keep delivery on time | IDW S 6 (guiding picture of the restructured company); not a step of its own in the other approaches |
| 3 Options and order | List of levers with effect, cost, risk and dependencies (tab “Twelve levers”) | Gabler (develop, assess and decide on alternatives), IDW S 6 (measures, integrated planning) |
| 4 Implement in steps | Change implemented, measured before and after | Gabler (implement), Theory of Constraints (exploit, subordinate, elevate), Bloom et al. (four months of implementation) |
| 5 Anchor and measure | New standards, fixed responsibilities, key figures in daily work | Only measuring again: Theory of Constraints (prevent inertia, back to step 1) and Bloom et al. (measurement phase); anchoring is not a step of its own in any of the approaches |
Rating. Statement on this topic in the ladder, with rating: Working hypothesis 5: whoever supports a firm first understands its processes, then sets a target picture and changes in steps that can be checked.
Two ways of changing an organisation
Developing or restructuring? Both change the organisation, but under different conditions. This chapter calls the rebuilding in a crisis “restructuring”, as the Gabler business dictionary describes it under Sanierung (restructuring of an ailing firm) and turnaround; the web search found no entry of its own for “Restrukturierung” there.
| Feature | Developing (reorganisation in the sense of the Gabler business dictionary) | Restructuring in a crisis (Sanierung) |
|---|---|---|
| What it is | An existing organisational structure is changed | Organisational and financial-technical measures meant to restore the performance of insolvent companies, above all to avert inability to pay |
| Trigger | Changes in the environment or the offer, a change of key people | A crisis with causes inside (for instance financing) or outside (for instance the economy) the company |
| Goal | More profit at the same revenue (scenario of this chapter) | Turn out of losses that threaten survival into a lasting profit situation that secures survival (turnaround) |
| Means | Processes, responsibilities, systems (tabs “Processes” and “Twelve levers”) | Additional financial steps: new capital, longer terms, converting debt into equity, reducing capital, changing the legal form |
| Time pressure | Time to try things out on a small scale (assessment) | Grows as the crisis advances (tab “Restructuring”) |
| Framework in Austria | No law of its own for this case (assessment of this chapter, not checked by an expert) | Business Reorganisation Act and insolvency law (tab “Restructuring”) |
Caution, same word: the Austrian Business Reorganisation Act uses reorganisation for a measure to improve the asset, financial and earnings position of a company whose existence is at risk, which enables it to continue in the long run (§ 1 (2)), that is restructuring in a crisis. In this chapter reorganisation in the dictionary’s sense stands for developing without a crisis.
Evidence. Gabler, Reorganisation (reorganisation); Gabler, Sanierung (turnaround of an ailing firm); Gabler, Turnaround; § 1 Business Reorganisation Act. The full references are in the sources block below.
Restructuring in the sense of this chapter means: the firm is already in a crisis. The German professional standard IDW S 6 of the Institut der Wirtschaftsprüfer (Institute of Public Auditors in Germany) distinguishes, as a rule, six crisis stages that build on each other and describes what a restructuring plan must contain; in Austria the Business Reorganisation Act sets two early-warning indicators. This tab orders the stages, the components, the legal position and the time spans to be expected.
Under IDW S 6, as Grant Thornton (2025) describes them
According to the description by Grant Thornton (2025), an audit and advisory firm, corporate crises usually run through six typical stages that build on each other. As a crisis advances, the pressure to act grows while the room for action shrinks; the earlier a firm reacts to warning signs, the better its chances of a turnaround. The standard itself was not read. For the subject of this chapter this means (reading of this chapter): whoever reorganises already in the stakeholder or strategy crisis still has time to try things out on a small scale. In the liquidity crisis the issue is the ability to pay, and little room is left for careful learning.
Rating. Statement on this topic in the ladder, with rating: Crises usually run through six stages that build on each other, and as a crisis advances the room for action shrinks.
Evidence. Grant Thornton 2025. The full references are in the sources block below.
Seven components under IDW S 6, 2023 version
According to the description by a tax advisory firm (Tabak, as of April 2026), the German professional standard IDW S 6 of the Institut der Wirtschaftsprüfer (Institute of Public Auditors in Germany) in its 2023 version requires seven components:
The structure matches the working model in the tab “Supporting a firm”: first the situation and the causes, then the target picture, then the measures. The article does not give a working order; it lists contents. We did not read the standard itself, and we did not check whether and how it is applied in Austria.
Evidence. Tabak on IDW S 6. The full references are in the sources block below.
Business Reorganisation Act, legal position
Under § 22 of the Business Reorganisation Act (Unternehmensreorganisationsgesetz; liability provisions) a need for reorganisation is presumed when the equity ratio is below 8% and at the same time the notional debt repayment period exceeds 15 years (JUSLINE, state of legislation taken into account 9 September 2017). The exact calculation is set out in §§ 23 and 24; they were not read here. If insolvency proceedings are opened over the assets of a legal person subject to audit that runs a business, the members of its authorised management body are liable to it for the debts the estate does not cover, up to €100,000 per person, if within the last two years before the insolvency application they received an auditor’s report with this finding and did not apply for a reorganisation procedure without delay or did not pursue it properly. They are equally liable if in this period they did not draw up the annual accounts or did so late, or did not commission the auditor without delay. Where the body can only act jointly, only the members who rejected starting the reorganisation procedure are liable. The same applies by analogy to certain registered partnerships (§ 22 (2)).
The Austrian Economic Chamber adds (as of 1 January 2025): a company whose existence is at risk but which is not insolvent can apply for reorganisation at the locally competent regional court (in Vienna the Commercial Court). Neither the opening nor the lifting or termination of the procedure is made public. The reorganisation plan has a deadline of 60 days, extendable by 30 days; the reorganisation auditor delivers the report within 30 days of receiving the plan. A period of two years is to be envisaged for the reorganisation.
Legal position according to the pages read, not checked by a legal expert; no legal advice. Read the current text in the federal legal information system and have it checked by an expert before any decision.
Evidence. § 22 Business Reorganisation Act; Austrian Economic Chamber on business reorganisation. The full references are in the sources block below.
Three periods from three sources
The three periods measure different things but show the same pattern: months for a supported implementation, years until recovery. The time until the report is our own calculation from the deadlines of the Economic Chamber (60 and 30 days for the plan, 30 days for the report, together at most 120 days). Kotter (1995) writes, after observing more than 100 companies, that change runs in phases which together usually take considerable time (tab “People”).
Evidence. Bloom et al. 2013; Austrian Economic Chamber on business reorganisation; Gulati, Nohria and Wohlgezogen 2010; Kotter 1995. The full references are in the sources block below.
What looks like duplicated work or mere waiting time from outside can serve a task that is written down nowhere. Whoever reorganises must know these dependencies before cutting anything. This tab shows this with an invented example, orders the places at which one can intervene in a system, and collects questions to ask before any change.
Invented example, continuing the order handling from the tab “Processes”
In the value stream the approval looks like pure waiting time (2 days, tab “Processes”). On the side, management checks whether the customer still has unpaid invoices; it gets the hint by word of mouth from accounting. Whoever knows only the described process does not see the dashed links. If the approval is dropped, the check of unpaid invoices is missing, and accounting is the first to notice. The example is invented.
An old rule of thought and two findings that come close to it
The rule of thought behind it is old: in “The Thing” (1929) Chesterton wrote, in substance, that a fence should only be taken down once one knows what it is for. That is a rule of thought, not a measured finding. We did not find in this web search a measurement of how often such hidden tasks make a change fail. Two findings come close: in a car plant, more strongly fluctuating option variety went together with more overhead hours, downtime, rework and inventory (Fisher and Ittner 1999, tab “What growth leaves behind”); that such costs are usually not attributed to variety is our interpretation. And during staff cuts, knowledge management softened the negative association with performance (Slavković et al. 2021, tab “Cutting or reorganising”).
Rating. Statement on this topic in the ladder, with rating: Working hypothesis 4: what looks like duplicated work from outside can serve a hidden task; before any change the dependencies must be clear.
Evidence. Chesterton 1929; Fisher and Ittner 1999; Slavković et al. 2021. The full references are in the sources block below.
Where cutting takes hold, and where reorganising
Meadows ranks twelve places at which one can intervene in a system by their leverage. Numbers and parameters, which is what cuts change, stand at the very bottom; information flows, rules and goals further up. According to Meadows, numbers and parameters only become strong when they set off a place higher up the list. She calls the list explicitly provisional and writes that the higher the place, the more a system resists changing it. That fits the experience that reorganising meets resistance, but it is a thinking model and not a measurement.
Rating. Statement on this topic in the ladder, with rating: Numbers and parameters are usually the weakest lever in a system.
Evidence. Meadows, Leverage Points. The full references are in the sources block below.
Seven questions, derived from Chesterton, Meadows and the field experiment
Seven questions to ask before any change to a process. They are derived by this chapter from the sources named, not a tested procedure.
| Question to ask before any change | Why | Background |
|---|---|---|
| What is it for, who introduced it and why? | What looks pointless may serve a task that nobody wrote down | Chesterton 1929, rule of thought |
| Who notices first when it is missing? | The consequences often show up somewhere other than the saving | Example above: accounting |
| What information flows here on the side? | According to Meadows, lines of information are a stronger lever than numbers | Meadows, place 6 |
| Which rule or incentive depends on it, even unwritten? | Rules steer behaviour | Meadows, place 5 |
| What happens during holidays, illness and peak load? | Buffers absorb fluctuations and disruptions; where they are missing, it often shows only in the exceptional case | Meadows, place 11 |
| What do we measure before and after, and who looks at it? | Without measurement one cannot check whether the change works | Bloom et al. 2013 evaluated four performance figures and recorded the adoption of the 38 practices (tab “Field experiment”) |
| Can the change be tried out on a small scale and reversed? | A mistake on a small scale is cheaper than on a large one | Derived by this chapter |
Reorganising hits habits, responsibilities and sometimes jobs. How people react to it is not just a question of their character: it also depends on how a change is carried out. This tab summarises a large review on this, Kotter’s observations on how long change takes, and two common claims that the sources do not bear out.
After Oreg, Vakola and Armenakis 2011, abstract
Oreg, Vakola and Armenakis (2011) ordered 79 quantitative studies from six decades. According to their abstract, reactions depend on characteristics of those affected and on the situation in the organisation, but also on how a change is carried out, whether it is perceived as useful or harmful and what it changes in substance. For supporting a firm this means (reading of this chapter): resistance cannot be defined away, but how the change is carried out is the part that management shapes most directly.
Rating. Statements on this topic in the ladder, with rating: Working hypothesis 3: change hits people and habits, hence the resistance; it can be shaped but not defined away; Reactions to change also depend on how it is carried out.
Evidence. Oreg, Vakola and Armenakis 2011. The full references are in the sources block below.
Kotter 1995 and anchoring
Kotter (1995) observed more than 100 companies during major changes. According to the short description on the publisher’s page, a few were very successful and a few failed outright; most lay in between, with a clear tilt towards the lower end. His two lessons: first, change runs in many phases that together usually take considerable time; skipping steps only creates the illusion of speed. Second, critical mistakes in any phase can slow the momentum and undo what has already been achieved. That is an account from experience, not a study with measurement; the individual errors that Kotter describes in the article were not read here.
Anchoring means: new standards, fixed responsibilities and key figures that are discussed in daily work (phase 5 of the working model in the tab “Supporting a firm”; a conclusion of this chapter). The Theory of Constraints sets a counterpoint in its fifth step: once a bottleneck is resolved, the search starts again, so that inertia does not itself become the bottleneck (tab “Processes”).
Evidence. Kotter 1995; TOC Institute. The full references are in the sources block below.
The 70 per cent figure and resistance
| Common claim | What the sources we read say | Rating of the chapter |
|---|---|---|
| “70% of all change initiatives fail.” | Hughes (2011) checked five publications that give this figure and found no valid and reliable empirical evidence for it. | Popular model |
| “Resistance lies with the people.” | According to Oreg et al. (2011), reactions also depend on how the change is carried out, on its perceived benefit or harm and on its content. | Simplifying |
How to read this. A missing proof does not mean the opposite is true: how many initiatives fail is open, not refuted. That is why the 70 per cent figure stands as “popular model” and not as “refuted”.
Evidence. Hughes 2011; Oreg, Vakola and Armenakis 2011. The full references are in the sources block below.
What holds up? Here are 23 statements from this chapter and from conversations about profit without growth, including the five working hypotheses, each with the evidence, the sources and a rating. The rating is the judgement of this chapter and not the truth; calculations, invented examples, models, studies and legal positions are marked separately. None of it is a recommendation for a particular firm.
Summary and state of evidence of the five working hypotheses
At the same revenue, profit rises through prices, processes, errors, variants and the capital tied up. Best documented are better management practices and fewer errors; several studies speak against cutting alone. Supporting a firm means: first understand, then change in steps that can be checked, and carry the change out so that those affected can come along.
How to read this. The ratings are the judgement of this chapter under the evidence rule of the encyclopaedia (well documented, partly documented, simplifying, popular model, refuted). They say how well a statement is covered by what we read, not what a particular firm should do.
23 statements, five levels
Each row contains a statement from this chapter or from conversations about profit without growth (23 rows, including the five working hypotheses), the key evidence for it and the rating. A click opens the row: what there is, what the evidence does not show, how we rate it, and the sources. The level applies to the statement as it is worded in the row. It is the judgement of this chapter and not the truth. The buttons filter the rows by area and by level.
| Level | Meaning | Rows |
|---|---|---|
| Well documented | A calculation, a legal norm or several measurements, the same direction, reference bases known. The statement is worded as narrowly as the evidence reaches. | 2 |
| Partly documented | Part of the statement is covered by studies or calculation and part is not, or the studies show only associations. | 13 |
| Simplifying | The statement simplifies: it holds for one part, one case or one reading and is generalised beyond that. | 2 |
| Popular model | A widespread statement or model for which we found no study that tests exactly it (web search, no specialist database). Provisional, and no verdict on whether there is anything to it. | 6 |
| Refuted | Studies or legal norms that we read speak against the statement. | 0 |
No statement in this chapter receives the level “refuted”: where evidence speaks against a wording or is missing, the row says “simplifying” or “popular model” and explains why.
What there is. An accounting identity: each of the four levers changes one line of the income statement. The worked example of Example Ltd shows the logic with invented figures.
What the evidence does not show. The identity says nothing about how large an effect is in a real firm or whether it lasts. That needs measurements (tab “Field experiment”).
Rating. Well documented: arithmetically the statement is certain. The rating applies to the calculation, not to the size of an effect.
In the chapter. Four levers at the same revenue
Evidence. Own calculation; an identity needs no source.
What there is. Own calculation at the same volume and costs: the leverage of a price change is the reciprocal of the margin. McKinsey (2003) works it through on the average income statement of a company in the S&P 1500 share index (Standard & Poor’s), a model calculation: a 1% higher price would bring about 8% more operating profit, nearly half as much again as 1% lower variable costs and more than three times as much as 1% more volume. For price, variable costs and volume this is the same order; the text gives no figure for fixed costs.
What the evidence does not show. The calculation does not say whether customers will bear a higher price; it assumes the same volume. The order depends on the cost structure, and price cuts work just as strongly in the other direction. McKinsey is a consultancy source with model figures, not an independent study.
Rating. Well documented as a calculation at the same volume and costs. How much price a market will bear is a different question, which this rating does not answer.
In the chapter. What one per cent does
Evidence. McKinsey 2003. The full references are in the sources block below.
What there is. The calculation is certain (row “At the same revenue … move the profit”). The order of magnitude is supported by a randomised experiment in 20 plants of one industry (Bloom et al. 2013) and, for the price lever, a consultancy source with model figures (McKinsey 2003).
What the evidence does not show. How large the effect is in other industries and in small firms; whether customers stay when discounts are dropped; Bloom et al. only estimated the profit, they did not measure it.
Rating. Partly documented: the direction is covered by the calculation and the field experiment, the size measured for one industry only.
In the chapter. Profit bridge: from 500 to 685
Evidence. Bloom et al. 2013; McKinsey 2003. The full references are in the sources block below.
What there is. A field experiment with random assignment: 14 plants received five months of free advice on 38 practices, 6 control plants only the one-month diagnosis and then measurement visits with little advice. The figures were checked against Table II of the publisher’s version. With the usual standard errors of the table, quality defects, output and inventory are significant at 0.05, productivity at 0.10.
What the evidence does not show. One industry in one region, only 17 of 66 firms approached took part; the consultancy collected the data itself, and the research team supervised the collection. Whether it carries over to other industries is open.
Rating. Partly documented: a strong method, but a small sample and one industry; in the permutation test two of the four findings lie just above the usual threshold.
In the chapter. The field experiment in Indian textile firms
Evidence. Bloom et al. 2013; Bloom et al. 2013, summary. The full references are in the sources block below.
What there is. A survey with a newly developed survey instrument; better practices went together with productivity, profitability, growth and survival.
What the evidence does not show. It is an association, not a demonstrated cause. Which single practice works is not separated in the abstract we read.
Rating. Partly documented: a large survey, but only the abstract was read and there is no proof of cause.
In the chapter. Lever, example, measure, caution
Evidence. Bloom and Van Reenen 2007. The full references are in the sources block below.
What there is. An association in company data: firms with fewer days until payment and fewer days of stock were more profitable.
What the evidence does not show. Whether shorter periods raise profitability or profitable firms simply manage their capital more tightly is not separated by the association; large firms in one country and one period.
Rating. Partly documented: an association, only the abstract was read.
In the chapter. Lever, example, measure, caution
Evidence. Deloof 2003. The full references are in the sources block below.
What there is. For errors, stock, receivables and management practices there are studies (Bloom et al. 2013, Deloof 2003, Bloom and Van Reenen 2007), for price discipline consultancy sources (Marn and Rosiello 1992, McKinsey 2003; both with Michael V. Marn as an author), for value stream and bottleneck thinking models, for variants indirectly one plant (Fisher and Ittner 1999), for approvals only indirectly a survey on bureaucracy in general (Hamel and Zanini 2017) and for staff several association studies.
What the evidence does not show. For small orders and purchasing we read no source. The studies we read give no sizes for single levers that could be carried over to other firms; the figures of the examples are invented.
Rating. Partly documented: the levers are supported to different degrees; the catalogue as a whole is an ordering by this chapter.
In the chapter. The lever map
Evidence. Bloom et al. 2013; Deloof 2003; Bloom and Van Reenen 2007; Marn and Rosiello 1992; McKinsey 2003; Fisher and Ittner 1999; Hamel and Zanini 2017. The full references are in the sources block below.
What there is. Quote Investigator quotes the passage from Drucker’s article “Managing for Business Effectiveness” (Harvard Business Review, May 1963) saying that the confusion between effectiveness and efficiency stands “between doing the right things and doing things right”. The page does not examine the short formula itself or its origin. The Gabler business dictionary defines effectiveness as the suitability of a measure to reach a given goal, and efficiency as the suitability to reach it in a certain way, for example while staying economical.
What the evidence does not show. In both dictionary definitions the goal is given; they do not ask whether it is the right one. “Doing the right things”, by contrast, moves effectiveness into the choice of the goal, that is into strategy. We did not read the 1963 article ourselves.
Rating. Simplifying: the word pair is documented as Drucker’s wording of 1963, the formula as a definition is not. It catches the core but is wider than the dictionary entries.
In the chapter. What the terms mean exactly
Evidence. Quote Investigator on Drucker 1963; Gabler, Effektivität (effectiveness); Gabler, Effizienz (efficiency). The full references are in the sources block below.
What there is. Several studies on downsizing show no advantage of cuts alone. Advantages appear where a restructuring of assets comes with the cuts (Cascio et al.), where costs are lowered through efficiency rather than cuts and investment continues (Gulati et al.), or where processes are improved (Bloom et al.).
What the evidence does not show. For the reorganising of responsibilities and systems itself, none of the studies we read is direct. The studies are associations (except Bloom et al.), partly with large listed companies.
Rating. Partly documented: the part “cutting alone is not enough” is well supported, the part “reorganising works” only indirectly.
In the chapter. Four depths of intervention
Evidence. Cascio, Young and Morris 1997; Guthrie and Datta 2008; Slavković et al. 2021; Gulati, Nohria and Wohlgezogen 2010; Bloom et al. 2013. The full references are in the sources block below.
What there is. In Cascio, Young and Morris pure downsizing brought no significantly higher returns than the average of the firm’s own industry; downsizing combined with a restructuring of assets brought a higher return on assets and higher share returns. In Guthrie and Datta downsizing went together with lower later profitability; they call the earlier research on the effect of cuts equivocal. In Slavković et al. staff cuts were statistically significantly associated with weaker performance. The review by Datta et al. names patterns and contradictions in the research; the abstract we read says nothing about the effect on performance.
What the evidence does not show. All are associations, not experiments. Cascio et al. studied large listed companies; for the others the abstract we read does not give the size, and for small firms it is open whether the findings carry over. Whether cuts are needed in a single case they do not answer.
Rating. Partly documented: several studies point the same way, but they are associations and mostly only abstracts were read.
In the chapter. What the studies show
Evidence. Cascio, Young and Morris 1997; Guthrie and Datta 2008; Slavković et al. 2021; Datta et al. 2010. The full references are in the sources block below.
What there is. About 80% of the surviving firms needed more than three years to regain their earlier growth rates in revenue and profit. According to the authors, the best prospects of joining the winners came from a combination: cutting costs mainly through operational efficiency rather than through more job cuts than comparable firms, and investing significantly more than rivals in research and marketing, while also investing in plants and machinery.
What the evidence does not show. An association in times of crisis among listed firms; only the publisher’s summary was read.
Rating. Partly documented: a large sample, but an association and only the summary.
In the chapter. What the studies show
Evidence. Gulati, Nohria and Wohlgezogen 2010. The full references are in the sources block below.
What there is. Two widespread methods: value stream mapping makes visible where work waits; the Theory of Constraints says where to start (at the bottleneck).
What the evidence does not show. The pages of the Lean Enterprise Institute and the TOC Institute that we read give no tested data on effects; the TOC Institute gives only an unsupported rule of thumb on spare capacity. That better processes work rests here on the field experiment, not on these methods.
Rating. Popular model: widespread and plausible, but not tested against studies here. Provisional, and no verdict on whether there is anything to it.
In the chapter. Five focusing steps
Evidence. Lean Enterprise Institute; TOC Institute. The full references are in the sources block below.
What there is. The number of possible lines of coordination grows faster than size. In a survey to which the Harvard Business Review had invited its readers, Hamel and Zanini (2017) describe bureaucracy in larger companies as markedly stronger; Fisher and Ittner (1999) found in one car plant that more strongly fluctuating option variety went together with more overhead hours, downtime, rework and inventory; that these costs are not attributed to variety is our interpretation. Greiner’s phase model describes something related: what works in one phase can trigger a crisis in another.
What the evidence does not show. None of the sources we read measures directly how the hurdles grow with size. Of the survey we read only the summary and the introduction, the figures are behind the paywall, and it is not a random sample; Greiner’s model is not tested against data here.
Rating. Partly documented: several indirect supports in the same direction, no direct measurement.
In the chapter. Lines of coordination grow faster than the firm
Evidence. Hamel and Zanini 2017; Fisher and Ittner 1999; Greiner 1972. The full references are in the sources block below.
What there is. More strongly fluctuating option content went together with more total labour hours and more overhead hours per car, more downtime, rework and inventory, but in the short run not with more direct labour hours. The average option content was not related to it; according to a simulation by the authors, random fluctuation does more harm than variety itself.
What the evidence does not show. One plant, one industry; only the abstract was read. As a remedy it names bundling options, not cutting variants. That the costs of variety are not attributed to single variants is our reading of the finding.
Rating. Partly documented: a well-fitting finding, but one plant.
In the chapter. Two findings from another angle
Evidence. Fisher and Ittner 1999. The full references are in the sources block below.
What there is. A well-known model: according to the secondary source each phase leads into a crisis; according to the publisher’s short description, management practices that work well in one phase can trigger a crisis in another. The names of the phases and crises come from the secondary source.
What the evidence does not show. We did not read the article itself; we did not find a test of the model against data in this web search. In 1998 Greiner added a sixth phase; the 1972 version is shown.
Rating. Popular model (provisional): widespread, not tested against studies here.
In the chapter. Phases and crises after Greiner
Evidence. Greiner 1972; Wikipedia on Greiner. The full references are in the sources block below.
What there is. The order appears in three approaches as a sequence of steps (reorganisation after the Gabler business dictionary, Theory of Constraints, field experiment) and in the German professional standard IDW S 6 in the structure of the components, as Tabak lists them. In the field experiment the control plants also received the diagnosis; the differences are mainly due to the supported implementation.
What the evidence does not show. None of the sources we read tests whether exactly this order works better than another. The target picture appears only in the restructuring standard, anchoring in none of the approaches as a step of its own.
Rating. Partly documented: a common order and a field experiment, but no comparison of orders.
In the chapter. Working model in five phases
Evidence. Gabler, Reorganisation (reorganisation); Tabak on IDW S 6; TOC Institute; Bloom et al. 2013. The full references are in the sources block below.
What there is. An old rule of thought (first know what the fence is for) and a thinking model (information flows and rules usually work more strongly than numbers). Two findings come close: more strongly fluctuating variety went together with costs elsewhere in a car plant (overhead, downtime, rework, inventory), and during staff cuts knowledge management softened the negative association with performance.
What the evidence does not show. We did not find in this web search a measurement of how often hidden tasks make a change fail.
Rating. Popular model (provisional): plausible and widespread, not tested against studies here.
In the chapter. The approval that is more than waiting time
Evidence. Chesterton 1929; Meadows, Leverage Points; Fisher and Ittner 1999; Slavković et al. 2021. The full references are in the sources block below.
What there is. Meadows ranks twelve places at which one can intervene in a system by their leverage; numbers and parameters stand at the bottom, information flows, rules and goals further up.
What the evidence does not show. She calls the list explicitly provisional; it is a thinking model from experience, not a measurement. According to Meadows, numbers and parameters only become strong when they set off a place higher up the list.
Rating. Popular model: an influential thinking model, not a measurement.
In the chapter. Twelve leverage points after Meadows
Evidence. Meadows, Leverage Points. The full references are in the sources block below.
What there is. According to Grant Thornton (2025) the pressure to act grows as a crisis advances while the room for action shrinks; the earlier a company reacts to warning signs, the better its chances of a turnaround.
What the evidence does not show. We did not read the standard itself; the stages are an ordering model from a German professional standard of public auditors, not tested against studies here. Tabak names the stages differently (for instance crisis of success and maturity for insolvency) and supports only the number six.
Rating. Popular model (provisional): a widespread ordering model, not a measurement.
In the chapter. Six crisis stages
Evidence. Grant Thornton 2025; Tabak on IDW S 6. The full references are in the sources block below.
What there is. According to the review by Oreg, Vakola and Armenakis, reactions also depend on how a change is carried out; that is the part management shapes most directly (our reading). Kotter describes from experience that change runs in phases and takes time.
What the evidence does not show. How strongly and by what means resistance can be shaped, the abstract we read does not say. Kotter’s account is not a study with measurement.
Rating. Partly documented: the influence of how change is carried out by a large review, the scope for shaping it not measured.
In the chapter. What shapes reactions
Evidence. Oreg, Vakola and Armenakis 2011; Kotter 1995. The full references are in the sources block below.
What there is. The review orders the influences: characteristics of those affected, the situation in the organisation, how the change is carried out, whether it is seen as useful or harmful, and what it changes in substance.
What the evidence does not show. How strongly each factor works, the abstract we read does not say.
Rating. Partly documented: a large review, but only the abstract was read and no strengths of the factors.
In the chapter. What shapes reactions
Evidence. Oreg, Vakola and Armenakis 2011. The full references are in the sources block below.
What there is. Hughes checked five publications that give the figure and found no valid and reliable empirical evidence for it.
What the evidence does not show. A missing proof does not mean the opposite is true: how many initiatives fail remains open, not refuted.
Rating. Popular model: a widespread figure without valid evidence.
In the chapter. Two common claims, checked
Evidence. Hughes 2011. The full references are in the sources block below.
What there is. Characteristics of those affected are among the influences, but not the only ones: according to the review the situation in the organisation, the way the change is carried out, its perceived benefit or harm and its content also count.
What the evidence does not show. How strongly each factor works, the abstract we read does not say.
Rating. Simplifying: one part of the influences is made into the whole explanation.
In the chapter. Two common claims, checked
Evidence. Oreg, Vakola and Armenakis 2011. The full references are in the sources block below.
Three scales that are not mixed
Each card carries labels at the top that say what kind of content it holds. They are a third scale next to the evidence levels (of a statement) and the reading levels (of a source) and are not mixed with them.
| Label | Meaning |
|---|---|
| Own calculation | A calculation by this chapter with the working shown; its figures may be invented. |
| Invented example | Firm, figures and assumptions are invented. They show the logic of a calculation, not a typical case. |
| Model | A model or a rule of thought: it orders and explains but measures nothing. |
| Study | Finding of a study or review; which kind (field experiment, association, survey) is stated in the text. |
| Legal position | Legal position in Austria according to the statute text read or the page of an authority or a court, not checked by a legal expert; no legal advice. |
| Overview | Overview or summary of the chapter. |
| Calculator | A calculator with openly stated assumptions: it calculates what you enter and does not say what will happen. |
Gaps in the research, stated openly
What we looked for and did not find, or did not read. “Not found” means: not found in a web search without a specialist database; it does not say that no such thing exists.
This tab gives no recommendation for a particular company and does not replace business, legal or tax advice.
The technical terms of this chapter in plain language, arranged by topic, with figures from the same calculations and sources as the other tabs. Each entry lists the sources on which the explanation rests; what we explain or calculate ourselves is marked as our own explanation. An entry explains a word and gives no recommendation.
What a company keeps in a year before taxes on profit are deducted: revenue minus variable costs, minus fixed costs, minus interest.
The invented Example Ltd arrives at 500 thousand euros on revenue of 10,000 thousand euros.
Own explanation. Simplified income statement of this chapter; annual accounts have more lines.
In the chapter. Example Ltd in figures (Profit without growth)
What remains of revenue after the variable costs to carry the fixed costs and the profit. More volume only brings this along, not the whole revenue: that is why one per cent more volume works less strongly than one per cent more price.
Example Ltd: 10,000 − 6,500 = 3,500 thousand euros, that is 35% of revenue.
In the chapter. Example Ltd in figures (Profit without growth) · What one per cent does (Profit without growth)
Variable costs follow the volume: materials, bought-in services, freight. Fixed costs arise regardless of volume: staff, premises, IT, administration. The line is not sharp; overtime, for instance, is a staff cost that varies with volume.
Own explanation. Division used for Example Ltd; in a real firm it needs checking.
In the chapter. Four levers at the same revenue (Profit without growth)
Profit before tax in per cent of revenue. The lower it is, the more strongly every improvement works on the profit: the leverage of a price change is its reciprocal.
At a margin of 5%, one per cent more on the price brings 20% more profit, at the same volume and costs.
Own explanation. Own calculation.
In the chapter. What one per cent does (Profit without growth) · The three returns compared, and all assumptions (Three examples)
A chart that leads step by step from today’s profit to another profit: each column shows what a lever adds or costs. In this chapter improvements are green, one-off costs red and the profit from the second year dashed.
In the chapter. Profit bridge: from 500 to 685 (Profit without growth) · A profit bridge for your figures (Margin calculator)
Costs that a change causes once, such as training, advice, changeover or programming. They weigh on the first year and fall away afterwards; change costs money before it works.
In the chapter. Profit bridge: from 500 to 685 (Profit without growth)
Money that sits in stock and in unpaid receivables and therefore causes interest and storage costs. If it falls, these costs fall without any change in revenue.
Evidence. Deloof 2003.
In the chapter. Lever, example, measure, caution (Twelve levers) · Wholesaler: from 1.5 to 2.5 per cent (Three examples)
Making sure that discounts get something in return. Volume discounts, cash discounts and loyalty bonuses make the price shrink between invoice and payment, often without anyone seeing all of them at once.
Evidence. Marn and Rosiello 1992.
In the chapter. Lever, example, measure, caution (Twelve levers)
The time from receipt of an order to its completion, waiting included. In the invented example it is 12.5 working days, of which only 2.5 are spent working.
In the chapter. Where the time goes (Processes)
Value stream mapping draws the actual process from receipt to delivery and separates time in which the order is worked on from time in which it lies idle. Lean names, after Womack and Jones, five steps: value, value stream, flow, pull, perfection.
Evidence. Lean Enterprise Institute.
In the chapter. Where the time goes (Processes) · Methods compared (Processes)
The bottleneck is the point that limits the throughput of the whole system. The Theory of Constraints after Goldratt says: first identify and exploit it, subordinate everything else, only then elevate it, and afterwards search again.
Evidence. TOC Institute.
In the chapter. Five focusing steps (Processes)
A method that reconstructs the actual process from the event logs of the IT systems, compares it with the plan and looks for improvements.
Evidence. Process Mining Manifesto 2011.
In the chapter. Methods compared (Processes)
The fundamental redesign of a process instead of merely automating it; made known by Hammer (1990). Of the article we read only the title and the short text.
Evidence. Hammer 1990.
In the chapter. Methods compared (Processes)
Fixed ways of managing, such as putting up and discussing key figures daily, controlling quality, keeping stock. In the field experiment 38 such practices in five areas were introduced.
Evidence. Bloom et al. 2013; Bloom and Van Reenen 2007.
In the chapter. The field experiment in Indian textile firms (Field experiment) · Lever, example, measure, caution (Twelve levers)
The number of possible pairs in a group that could coordinate with each other. It grows faster than the group: 10 people form 45 pairs, 100 people 4,950.
Own explanation. Mathematics, not a measurement.
In the chapter. Lines of coordination grow faster than the firm (What growth leaves behind)
The number of versions of a product. In one car plant, more variable variety went together with more overheads, rework, downtime and inventory, but not with more direct assembly hours: variety also costs where it is not attributed.
Evidence. Fisher and Ittner 1999.
In the chapter. Two findings from another angle (What growth leaves behind)
An experiment in real firms in which chance decides who receives the measure. Because the groups then differ on average only by the measure, a difference can be attributed to it.
Evidence. Bloom et al. 2013.
In the chapter. The field experiment in Indian textile firms (Field experiment)
The group that does not receive the measure and serves as the comparison. In the field experiment it was 6 plants that received only the diagnosis and then measurement visits with little advice.
Evidence. Bloom et al. 2013.
In the chapter. The field experiment in Indian textile firms (Field experiment)
A number between 0 and 1: the share of cases in which a difference at least as large would come out if the measure had had no effect and only chance counted. The smaller it is, the worse the result fits the assumption “no effect”.
It is not the probability that the result is chance, and it says nothing about how large or important an effect is.
Own explanation. Explanation by this chapter.
In the chapter. What the p-value means here (Field experiment)
A test that works through all possible assignments of the measure, as if it had no effect, and counts how often a difference at least as large comes out. In the field experiment there were 12,376 assignments, all ways of choosing 11 of the 17 firms.
Evidence. Bloom et al. 2013. Own explanation. Check of the number.
In the chapter. What the p-value means here (Field experiment)
An estimate of how strongly a result would vary by chance alone, obtained from the spread of the data. Dividing the effect by its standard error shows whether it reaches the usual threshold; tables often mark this with stars. With few observations, such as 17 firms, the estimate is uncertain; that is why Bloom et al. also calculated a permutation test.
Evidence. Bloom et al. 2013. Own explanation. Explanation by this chapter.
In the chapter. What the p-value means here (Field experiment)
It is usual to call a finding with a p-value below 0.05 statistically significant; some papers also name a looser threshold of 0.10. That is a convention, not a law of nature.
Own explanation. Explanation by this chapter.
In the chapter. What the p-value means here (Field experiment)
Output measured against capital and labour hours together. If it rises, the same machines and hours produce more.
Evidence. Bloom et al. 2013.
In the chapter. The field experiment in Indian textile firms (Field experiment)
Two quantities are associated when they occur together; that does not yet mean that one causes the other. Most studies on downsizing show associations; only a randomised experiment separates cause from association more reliably.
Own explanation. Explanation by this chapter.
In the chapter. What the studies show (Cutting or reorganising)
A paper that orders and summarises many studies on one question, such as the one by Oreg et al. on 79 studies or the one by Datta et al. on downsizing. It is as good as the studies it orders.
Evidence. Oreg, Vakola and Armenakis 2011; Datta et al. 2010.
In the chapter. What shapes reactions (People)
Those who decide for themselves whether to answer are not necessarily like everyone else. The survey by Hamel and Zanini, to which the Harvard Business Review had invited its readers, with more than 7,000 responses, therefore shows what the respondents report, not how it is in all firms (classification by this chapter).
Evidence. Hamel and Zanini 2017.
In the chapter. Two findings from another angle (What growth leaves behind)
The suitability of a measure to reach a given goal; the term says nothing about the way (Gabler). The word pair of the short formula “doing the right things” already appears in Drucker (1963); the formula moves the question into the choice of the goal.
Evidence. Gabler, Effektivität (effectiveness); Quote Investigator on Drucker 1963.
In the chapter. What the terms mean exactly (Effectiveness and efficiency)
The suitability to reach a goal in a certain way, for example while staying economical (Gabler); in this chapter: with little effort. Efficiency at the wrong goal remains worthless.
Evidence. Gabler, Effizienz (efficiency).
In the chapter. First become effective, then economical (Effectiveness and efficiency) · What the terms mean exactly (Effectiveness and efficiency)
Four depths of intervention, a division by this chapter: cutting saves across the board, improving streamlines processes within the existing system, reorganising redraws responsibilities, interfaces and systems, restructuring changes business, assets or financing.
Own explanation. Division by this chapter, not a finding of the studies.
In the chapter. Four depths of intervention (Cutting or reorganising)
The change of an existing organisational structure, for instance after changes in the environment or the offer or a change of key people (Gabler). In this chapter: developing without a crisis.
Caution, same word: the Austrian Business Reorganisation Act uses reorganisation for a measure to improve the asset, financial and earnings position of a company whose existence is at risk (§ 1 (2)), that is restructuring in a crisis.
Evidence. Gabler, Reorganisation (reorganisation); § 1 Business Reorganisation Act.
In the chapter. Developing or restructuring (Supporting a firm)
Sanierung: organisational and financial-technical measures meant to restore the performance of insolvent companies, above all to avert inability to pay. Turnaround: the turn out of losses that threaten survival into a lasting profit situation that secures survival (Gabler). In this chapter: restructuring in a crisis.
Evidence. Gabler, Sanierung (turnaround of an ailing firm); Gabler, Turnaround.
In the chapter. Developing or restructuring (Supporting a firm)
As a rule six stages building on each other under the German professional standard IDW S 6 of the Institut der Wirtschaftsprüfer, as Grant Thornton describes them: stakeholder, strategy, product and sales, earnings and liquidity crisis, then insolvency. With every stage the pressure grows and the room for action shrinks.
Evidence. Grant Thornton 2025.
In the chapter. Six crisis stages (Restructuring)
The written plan by which a company is to come out of the crisis; under the German professional standard IDW S 6 in its 2023 version, as described by a tax advisory firm, with seven components, from the starting situation through the guiding picture to the statement on the ability to be turned around.
Evidence. Tabak on IDW S 6.
In the chapter. What a restructuring plan contains (Restructuring)
In simple terms the share of equity in total capital. Under the Business Reorganisation Act a need for reorganisation is presumed when it is below 8% and at the same time the notional debt repayment period exceeds 15 years; § 23 sets out the exact calculation.
Evidence. § 22 Business Reorganisation Act; Austrian Economic Chamber on business reorganisation.
In the chapter. Early-warning indicators in Austria (Restructuring)
How many years a firm would need to repay its debts from its current cash surplus. Above 15 years it is, together with an equity ratio below 8%, an early-warning indicator; § 24 sets out the exact calculation.
Evidence. § 22 Business Reorganisation Act; Austrian Economic Chamber on business reorganisation.
In the chapter. Early-warning indicators in Austria (Restructuring)
A procedure at the regional court under the Business Reorganisation Act which a company whose existence is at risk, but which is not insolvent, can apply for itself; it is not made public. The plan is to be submitted within 60 days (extendable by 30 days), the reorganisation auditor delivers the report within 30 days; a period of two years is to be envisaged for the reorganisation.
Evidence. Austrian Economic Chamber on business reorganisation.
In the chapter. Early-warning indicators in Austria (Restructuring) · Months to implement, years to recover (Restructuring)
Twelve places to intervene in a system, ranked by their leverage: from numbers and parameters (usually weak) through information flows and rules to goals and mindsets (strong). Meadows calls the list provisional.
Evidence. Meadows, Leverage Points.
In the chapter. Twelve leverage points after Meadows (Dependencies)
The rule of thought that a fence should only be taken down once one knows what it is for (Chesterton 1929). For reorganising: first know the hidden tasks of a process, then cut.
Evidence. Chesterton 1929.
In the chapter. The fence one should understand first (Dependencies)
Negative reactions to a change. According to a review of 79 studies, reactions depend not only on those affected but also on how the change is carried out, whether it is seen as useful and what it changes.
Evidence. Oreg, Vakola and Armenakis 2011.
In the chapter. What shapes reactions (People)
Making a change stick in daily work: new standards, fixed responsibilities and key figures that are discussed regularly, so that the effect is not lost again.
Own explanation. Phase 5 of this chapter’s working model.
In the chapter. Working model in five phases (Supporting a firm) · Change needs time and order (People)
Where a works council exists, it must be notified before every dismissal; it can give its opinion within one week. A dismissal before this period has expired is legally ineffective unless it has already given its opinion (§ 105 Labour Constitution Act).
Evidence. § 105 Labour Constitution Act.
In the chapter. Legal framework in Austria (Staff)
Whoever intends to end the employment of several employees within 30 days, for instance by dismissal, must above a certain number first notify the regional office of the Public Employment Service (§ 45a Labour Market Promotion Act). The page of the Supreme Court names as a threshold at least five employees aged 50 or over; whether others apply is set out in § 45a itself, which we did not read.
Evidence. Supreme Court of Justice, 9 ObA 33/21z.
In the chapter. Legal framework in Austria (Staff)
This tab explains words and does not replace business, legal or tax advice.