One-Time Revenue Share: How Much of Last Year's Turnover Returns on Its Own?

In the first week of an investment review, once a three-year income statement series is on the table, the question the reviewing party puts is almost invariably the same, and it almost invariably arrives unrehearsed: how much of last year's turnover would have arrived this year without any renewed selling effort. The finance director on the company side typically answers not with a figure but with a narrative — client loyalty, repeat orders, the durability of the relationship. What has been asked, however, concerns not the quality of the relationship but the composition of the revenue, and a composition question can only be answered if the company's own recording discipline took a classification decision at the point each revenue item was booked. Where no such decision was taken, the answer is produced later in the review, without the company's participation, on the counterparty's own assumption.

This gap arises from the internal logic of the accounting framework rather than from neglect. A uniform chart of accounts is well suited to segmenting revenue by client, product line, geography and tax regime, and poorly suited to segmenting it by expectation of recurrence, recurrence being not an attribute observable at the moment of the transaction but a judgment about the future. The same line item, the same invoice amount to the same client, may represent installation work one year and maintenance the next, and the accounting entry posts both to the same revenue account. The only natural moment at which the classification can be made is not the invoice date but the point of order acceptance, since the contract term, the presence or absence of a renewal mechanism, and whether the engagement is an asset delivery or a continuity undertaking are all known at that point. Left unrecorded, this information dissipates within months and can only be reconstructed by opening contract files one by one — the slowest available exercise, required precisely in the week when the review calendar is tightest.

Even where the classification has been established, a second mechanism quietly erodes the ratio: the location of classification authority. Where the decision is left inside the sales organisation — that is, where the team selling the work also determines whether the work counts as recurring — the classification drifts over time in the shape of the commission plan. In a period carrying a recurring-revenue target, borderline engagements are booked to the recurring side; in a period carrying a volume target, to the project side. Each individual determination appears defensible, yet in aggregate the ratio reflects the incentive calendar rather than the economic nature of the work. Although the drift may look modest in magnitude, its valuation effect is not linear, since the spread between the multiple applied to recurring revenue and the multiple applied to project revenue exceeds a full turn in most sectors.

The third and most widespread confusion is the collapse of the distinction between a recurring client and recurring revenue. A company that has taken project work from the same corporate account for three consecutive years treats that revenue, in its internal narrative, as loyal and therefore dependable; what recurs, however, is the purchasing behaviour rather than the revenue, and behind that behaviour sits no contractual provision binding the client for the following year. The practical test of the distinction is straightforward, and the reviewing party invariably applies it: when the procurement manager on the other side changes, or when the client's own budget cycle tightens, does the revenue continue of its own accord or must it be won again. In the latter case the revenue, however regular it appears in the ledger, is one-time revenue that happened to repeat — and repetition is not commitment.

The picture separates further once the measurement dimension is introduced. Where the ratio is calculated only once, at year end and ahead of the audit, the resulting figure is a presentation item rather than a performance indicator, having influenced no management decision at any stage. Measurement becomes meaningful when the ratio is tracked monthly or quarterly alongside the direction of shift in revenue composition, since what proves decisive in valuation is the trajectory of the ratio as much as its level — often more so. A one-time share rising for two consecutive years indicates deteriorating revenue quality behind an expanding turnover figure, and that pattern is typically detected by the reviewing party before the growth line is even discussed.

The institutional cost surfaces through several channels simultaneously. The first and most direct is the multiple base: where the distinction is undefined, the counterparty determines the recurring share by its own estimate, and under conditions of information asymmetry that estimate is calibrated conservatively — which is to say, against the company. The second is deal structure; in files where revenue quality remains uncertain, the price gap is frequently closed not through negotiation but through an earn-out, whose trigger is constructed precisely on the item the company was unable to evidence. The third is the scope of representations and warranties, since where no statement on revenue composition can be given, the escrow percentage is raised and the escrow period extended. Because these three channels reinforce one another, the absence of a single classification rule generates cost not in one line of the transaction economics but across three headings at once.

A parallel mechanism operates on the credit side. In sizing debt capacity against EBITDA, a credit committee strips out non-recurring items to arrive at a normalised figure; where the company has not performed that exercise in advance through its own recording discipline, the stripping is performed with the committee's own margin of prudence, and the covenant headings — the leverage ceiling and the cash flow cover ratio in particular — are calibrated against that narrowed base. The absence of the classification is therefore priced as a valuation discount in a sale process and as a tighter covenant package in a financing, and in both cases the cost sits an order of magnitude above the cost of having built the classification in the first place.

The ownership dimension is the weakest link in most companies, because the one-time revenue share is not an item that finance, sales or operations owns alone; it sits at the intersection of the three, and for exactly that reason belongs to none of them. The symptom of that vacancy is that any question about how a borderline engagement should be classified escalates to the founder or the general manager — the rule living in a person's judgment rather than in a document. This configuration operates quickly and consistently while the company remains small; for the reviewing party, however, it means the definition of revenue quality is founder-dependent, and where founder dependence extends as far as the definition of revenue itself, no assumption can be made that the picture remains comparable in the post-founder period.

Structural intervention is built through decision architecture rather than awareness, and it has four components. The first is that the classification rule be written and testable: the conditions under which revenue may be treated as recurring — minimum contract term, presence of a renewal mechanism, unilateral termination provisions, uninterrupted delivery of the service — are defined jointly and in terms leaving no room for interpretation. The second is the timing of the decision: classification is performed at order acceptance, ahead of revenue recognition, and any subsequent reclassification is made subject to separate approval. The third is ownership; the rule is held within the finance function, while borderline cases are adjudicated by a standing forum in which finance and sales sit together. The fourth is the record: every altered classification is entered, with its rationale, into a decision log, since what the reviewing party seeks is the traceability of how the ratio was produced as much as the ratio itself.

BEIREK builds this intervention by beginning not with the calculation of the ratio but with testing the classification rule against the company's existing contract inventory; read afresh under the rule, contract by contract, the portfolio reveals where the rule remains ambiguous and in which business line it produces a systematic drift. A fixed review cadence is then operated for adjudicating borderline cases — typically monthly, embedded within the closing calendar, with participants and authority defined in advance — and the output of that session becomes a decision log carrying the one-time revenue share for each period together with its reasoning. What is presented at the review table is consequently a chain of records rather than a calculation, and a chain of records, unlike the figure for a single period, substantially removes the counterparty's grounds for applying its own margin of prudence.

The only genuine test of whether this structure is sustainable independently of the founder is not that the rule exists in writing but that a contested case can be resolved with the founder absent from the room. Where finance and sales can classify a borderline contract together, against a defined rule and within a reasonable period, the definition of revenue quality has been institutionalised; where that test is not met, the resulting ratio — however accurately computed — is an outcome whose reproducibility has not been demonstrated. What is priced in valuation is precisely this: not the number itself, but the demonstrable capacity of the company to produce the number again.

The one-time revenue share is, in the end, less an income statement line than an indicator of how honest a record a company keeps of the nature of its own revenue. An investor who finds the ratio high may well accept it where the nature of the business requires it, recurring revenue being no reasonable expectation from a project business; what is not accepted is that the ratio is unknown. The operative question, accordingly, is not what proportion of turnover will repeat, but whether the company can answer that question ahead of the review, by looking at its own records.