Steering by the figures.

The accounts say what happened; they do not say what happens if

Every spring, an executive receives a set of accounts. He leafs through them, finds the result, and does not find what he is looking for: can I hire, from what turnover does the year hold, what happens if I lose this client. There is nothing abnormal in that. An income statement is a reporting document, designed for the tax authority, the bank and the shareholders' meeting. It answers exactly the question of what happened; it answers none of the questions an executive really asks.

Yet these questions require no new data. They require a different reading of the same figures. That reading is what the firm practises in every engagement that touches finances, and what it has set down in a method handbook and in an instrument, StratBoard. The firm does not keep the accounts, does not audit them and does not replace the accountant; it reads them from the point of view of the person who has to decide, and turns that reading into decisions.

What makes these decisions hard

The first difficulty is that the separation that governs everything appears in no accounting document: which of these costs fall when activity falls, and which do not move. Two companies with identical results and different structures have neither the same break-even point, nor the same resistance to a bad quarter, nor the same capacity to hire. That split depends on the trade, cannot be guessed, and is almost never done.

The second is that the accounts are exact and misleading at the same time. An executive who pays himself below market, rent paid to his own property company, a spouse who keeps the books on Saturdays: three invisible gaps, stable from one year to the next, which flatter the result. A company can show ten profitable years in a row while being structurally loss-making once its real costs are reconstructed.

The third is the confusion between profitability and cash. The more a company grows, the tighter its cash becomes, even as its result improves; that is the road by which profitable companies go bankrupt. The number of days between the moment one pays and the moment one is paid is the reading executives most often confuse with profitability.

The situations the firm handles

A margin eroding with no visible cause. Turnover holds, the result falls, and nobody can say where. The firm separates the costs that follow activity from those that do not, establishes the break-even point and the margin of safety, and traces the erosion back to the line items that carry it. The answer rarely lies in an industry average; it lies in one or two lines of your own accounts.

Dependence on one client or supplier. A forty per cent concentration on three clients says nothing as long as one does not know where the contracts stand. The firm crosses concentration with contractual deadlines, then calculates what the departure of the first client would take from the result: not its turnover, but the margin it carries. An executive can form that sentence in ten minutes once the figures are read; he will never form it by reading his accounts.

A pricing policy to revise. A price discount produces no saving in costs: every franc conceded comes straight off the result. The firm puts a figure on what a price negotiation is really worth, and moves the discussion to where it can be won, on scope rather than on price.

A budget that has to commit. A hire, an investment, an opening. The firm measures what every additional franc of fixed cost requires in revenue to be covered, what the decision does to the margin of safety, and on what condition it remains tenable if the first client leaves at the end of its contract.

A management dashboard to build or rebuild. The executive's dashboard is not the controller's. It fits on one sheet, eight blocks, and it is fed with a closed financial year and a spreadsheet. The firm builds it with the executive and the accountant, with restatements written down, sourced and kept from one year to the next, so that the company can compare itself with itself over time.

Seven instruments, in the order they are used

Everything starts from a single gesture: separating, line by line, the costs that follow activity from those that do not, within a horizon set in advance. Once that gesture is made, five readings become available on a closed financial year, with no additional data: the break-even point, the margin of safety, the cost structure, client dependence and the cash cycle. Two further readings, seasonality and sustainable growth, require a second financial year or intra-year figures, and the firm says so rather than fabricating them.

Three restatements complete the reading and often change everything: the owner's remuneration brought to market rate, family labour valued, rent imputed. Each is written down with its value, its source and its year, and kept, because a restatement recalculated on a whim destroys the only comparison that counts, that of the company with itself.

And one rule that applies to any report you commission from anyone: a report that produces a figure for every question asked is suspect. An indicator produced from missing data is not approximate, it is false, and its falseness is invisible because it carries decimals like the others. A report that declares certain indicators unavailable, naming the missing data, behaves correctly.

The mistakes the firm sees most often

Classifying as variable everything one would like to be able to adjust. An open-ended contract with three months' notice does not fall within a six-month horizon, whatever one thinks. The resulting break-even point is optimistic, and the confidence unjustified.

Seeking precision to the percentage point. A sixty-forty split applied consistently is worth more than a fifty-eight-forty-two split recalculated every year; what counts is not the accuracy of the split but its stability over time.

Comparing with an industry average of unknown provenance. Accessible benchmarks often come from other countries, mix unrelated companies, and state neither their year nor their sample. A reference without provenance is not data, it is a numerical rumour.

Reading the result as a measure of resilience. Two companies showing the same profit do not run the same risk; at equal result, one can lose a tenth of its activity before tipping over, the other a quarter.

Expecting from the accountant a reading that was never commissioned. The accountant is committed to accuracy, compliance and deadlines; he generally knows the answers, but nobody asked him to put figures on them. Three questions a year are enough to change the relationship, and the firm provides them.

Deliverable and duration

A reading of the accounts as management questions, the decisions it calls for, and the indicator to track for each; the eight-block template, reproducible in a spreadsheet, stays with the company. Two to four weeks, then a regular review if you wish, at a fixed fee set in the written proposal. The first thirty-minute conversation is not charged.

What you can do right now

The full method is public and free: the firm's handbook, "Décider sur des chiffres, pas sur des impressions", is available for download in the Documents section› (in French), with a complete case that can be redone line by line. Take the list of your costs from last year and ask each line a single question: if my activity fell by thirty per cent for six months, would this item fall by thirty per cent? The percentage that comes out, the share of your costs that does not fall, is the one figure an executive should know by heart. If you have never calculated it, that is where a first conversation with the firm begins.