Steering the business · Series “Reporting that carries decisions”, part 1
Why nobody reads your management report — and what it costs
A management report nobody reads is not a cosmetic flaw. It costs a six-figure sum every year — and decisions that are never taken.
Philipp Kempter · 3 August 2026 · 8 min readMonday, 9 a.m., executive meeting. The monthly management report has been sitting in everyone's inbox since Friday: 42 pages, three annexes, a cover page with traffic lights. The CFO has read it. Nobody else has — the others leaf through it during the first agenda item, and the first comment is not about margin but about why revenue on page 7 does not match the sales report. If you recognise this scene, your management reporting is not badly made. It simply has no mandate.
Five symptoms that give it away
Reporting overload. There are more reports than questions. Every reorganisation, every new system and every wish of a former executive has left a report behind, and none has ever been switched off. Controlling does not know who reads what, and does not ask, because the answer would be uncomfortable.
Numbers without a message. The report shows revenue, costs, result and liquidity by entity, by unit, by month — complete and correct. What follows from it is written nowhere. The reader is expected to find the anomaly, and because the reader has no time, the anomaly is not found.
Every page looks different. On page 3 the budget is a dashed line, on page 9 a grey bar, on page 14 it is missing. Prior year sits on the left once and on the right the next time. Whoever reads the report relearns the notation on every page instead of understanding the number.
The commentary repeats the number. "Revenue rose 4.2 % against prior year." That is already in the table above. The comment explains neither why, nor whether this is good, nor what should be done now. It is a table turned into prose.
The meeting opens with a fight over figures instead of a decision. Cindi Howson describes this as the typical frustration of mature organisations: executive meetings start with a debate about how the numbers are compiled and whose are correct rather than with a discussion of the insights those numbers provide (Howson 2014, p. 111). When sales, production and finance each arrive with their own revenue figure, the meeting is half over before the first decision is on the table.
What reporting costs: a worked example
Reporting is the most labour-intensive controlling process, and most of the labour adds no value. Christian Langmann summarises the process analyses this way: around 70 % of total effort in the reporting process goes into data collection, data preparation, report production and validation — the steps before anyone interprets a number (Langmann 2019, p. 12). For analysis, commentary and deriving actions, little time is left, particularly in mid-sized companies.
Let us work it through for a fictitious but realistic group: three legal entities, 200 employees, a controlling team of two, plus unit heads who fill in Excel sheets for "their" reports themselves. Twenty regular reports, together around 260 hours of production effort per month, valued at a full cost of CHF 110 per hour. On the other side of the ledger are the decisions that demonstrably relied on these reports over the past twelve months — not meetings where the report was on the table, but decisions that would have been taken differently, or not at all, without it.
| Report family | Reports | Hours/month | Cost/year TCHF | Decisions that relied on it |
|---|---|---|---|---|
| Monthly executive report, consolidated | 1 | 60 | 79 | 4 |
| Unit reports: sales, production, logistics, purchasing | 6 | 70 | 92 | 3 |
| Quarterly board report | 1 | 24 | 32 | 2 |
| Liquidity and forecast | 2 | 30 | 40 | 5 |
| People and projects | 4 | 36 | 48 | 1 |
| Ad-hoc analyses and "legacy" reports | 6 | 40 | 53 | 0 |
| Total | 20 | 260 | 343 | 15 |
CHF 343,000 a year for fifteen decisions — just under CHF 23,000 per decision. By Langmann's rule of thumb, around CHF 240,000 of that goes into collecting, preparing and validating. Six reports, a quarter of the effort, carried not a single decision in twelve months. And the calculation is incomplete: the recipients' time is missing. Fifteen managers who each spend four hours a month reading, querying and reconciling add another 720 hours a year — valued at management rates, another six-figure sum. Werner Gladen names the reason why precisely this time is the most expensive: the scarcity of management time and attention is the underlying reason for every condensation of information (Gladen 2014, p. 12). A report that consumes attention without preparing a decision costs more than its production.
The figures in your group are different. The structure of the calculation is the same, and it can be built in an afternoon.
Why it happens: supply, demand and need do not coincide
Gladen describes the information supply of management as "scarcity amid abundance": what the data offers, what management asks for and what management actually needs for its tasks are three different sets that overlap only in part (Gladen 2014). The starting point for objective information need is the problems to be solved, derived from the company's objectives (Gladen 2014, p. 7) — not the recipients' wish list and not whatever the system happens to be able to output.
Ossola-Haring, Schlageter and Schöning turn the same three circles into a test with seven fields (Ossola-Haring et al. 2019, pp. 94 f.). Two of them explain most of the reporting overload. First: reports that consist of supply alone — technically available, wanted by nobody, needed for no decision. The authors call them zombie reports; they derive their right to exist purely from their history (p. 93). Second: reports that are supplied and demanded but contribute nothing to steering. This is the more dangerous field, because the report is read and therefore looks legitimate. Whatever can be reported gets produced, and usage then retroactively confirms the investment — the cat chases its own tail (p. 96).
The second reason is more fundamental. Reports do not steer. Ossola-Haring et al. put it bluntly: an automatically generated management report, a technically and visually polished management cockpit or a glossy presentation with many facts and charts does not steer a company (Ossola-Haring et al. 2019, p. 96). What steers is a process: someone looks at a variance in a fixed rhythm, assesses it, decides and records the decision. The report supplies that process. Without the process, the report remains a document — maintained, distributed, filed.
Which leads to the sentence this article rests on: a metric without a decision, an owner and a cadence is not steering. It is documentation.
What to do: put every metric to the test
The way out of reporting overload does not start with a new tool or a prettier layout. It starts with a question we ask first in every review: what are these numbers supposed to say, and what are they for? Anyone who cannot answer that in two sentences has a candidate for the cut list.
Concretely, every metric that wants to stay in the management report needs three things:
- A decision. Which decision changes if the number comes out differently? Pricing round, capacity, inventory, headcount, investment, credit line. If none comes to mind, the number is background knowledge — legitimate, but not on page 1 and not monthly.
- An owner. Who assesses the variance and carries the decision? Not the controller who delivers the number, but the person who acts. A metric without an owner has no recipient either.
- A cadence. How often do you look at it? Daily, weekly, monthly, quarterly. The cadence follows from the decision, not from the closing calendar. A number you can only react to once a quarter does not belong in the monthly report.
Whatever passes these three questions stays — and is presented differently from now on. The value itself is rarely of interest; what matters is the variance from target, from budget, from prior year. So we show the target as the baseline and the actual only as a deviation from it, with a threshold above which a variance deserves attention, say ±5 %. Whatever lies within the threshold needs no comment. Whatever lies outside needs one — and one that names cause and consequence rather than repeating the number.
Whatever fails the three questions goes on the cut list. Not deleted at once, but suspended for one cycle: the report is no longer distributed, and whoever misses it speaks up. Usually nobody does. This test is the first step of the framework underlying this series — from decision to metric — and it is deliberately not a technology project. Only once it is clear which metrics stay is it worth talking about data models, automation and presentation. The way we approach this in review and scoping engagements follows a fixed order: objective, problem, early warning, analysis, report, impact.
At an industrial company I once reviewed around two dozen existing reports that had accumulated over the years — logistics, warehouse, production, finance. For each one I asked the same question: what are these numbers supposed to say, and what are they for? For a part of the reports nobody could say any more which of the data was actually looked at; they went. For the rest it turned out that recipients did not need the value at all, only the variance from target. The result was markedly fewer pages, a single notation and reports that show variances instead of values. Nobody missed the reports that were cut.
What convinced me that this test is an investment appraisal rather than tidying up is a line from Ossola-Haring et al.: the benefit of a metric — or of a decision derived from it — must be significantly higher than the effort of generating it (Ossola-Haring et al. 2019, p. 88). A business case for a reporting system has to show, as for any capital investment, whether the gain in steering capability justifies the cost (p. 96). Hardly any group would run a machine costing CHF 340,000 a year without knowing what it produces. In reporting, that is the norm.
The next step
Tomorrow: take the report that is next to be distributed and write three words in the margin of every page — the decision, the owner, the cadence. Where one is missing, flag the page. The day after: build the calculation from this article with your own figures, hours per report and month, decisions of the past twelve months. Together they are your first cut list, and it is usually longer than you expect.
If you want a second opinion before you cut, that is what a project review is for: a few days, fixed price, written second opinion. We go through your reports with the same purpose question, set the effort against the decisions and deliver a prioritised cut list with what stays — and why.
Sources
- Gladen, Werner (2014): Performance Measurement. Controlling mit Kennzahlen. 6th edition, Springer Gabler, Wiesbaden.
- Howson, Cindi (2014): Successful Business Intelligence. Unlock the Value of BI & Big Data. 2nd edition, McGraw-Hill Education, New York.
- Langmann, Christian (2019): Digitalisierung im Controlling. Springer Gabler (essentials), Wiesbaden.
- Ossola-Haring, Claudia; Schlageter, Andreas; Schöning, Stephan (2019): 11 Irrtümer über Kennzahlen. Mit den richtigen Erkenntnissen führen. Springer Gabler, Wiesbaden.