In an investment committee presentation, the diversity of the customer base is defended in almost the same way every time: the share of the largest account is disclosed, the combined share of the top five is added, and if both figures sit comfortably low the subject is treated as closed. Across the same table, the party running the review performs a different calculation, grouping customers not one by one but according to the source of the budget that pays them; and the picture that emerges from that grouping is frequently more concentrated than the top-five rule suggested. Where fifty customers all sell into the same downstream sector, depend on the same public incentive programme, or draw on the same year's capital budget, the portfolio holds one risk cut into fifty pieces rather than fifty separate exposures. The gap between those two calculations is rarely news to the sales organisation; because it has never been recorded formally anywhere inside the company, however, it becomes visible only when an outside desk sits down and asks.
The mechanism beneath that gap is the selection of the wrong counting unit. As a company grows the number of accounts naturally rises, and a rising count reassures everyone looking from inside; the carrier of risk, though, is not the customer but the condition that generates the customer's demand. Accounts tied to a common condition behave in correlated rather than statistically independent fashion — when one defers an order, the probability that the others defer rises with it, since the reason for the deferral lies not in the relationship with the supplier but in an external variable to which all of them are exposed simultaneously. Segment diversity is precisely the breaking of that correlation: buyer groups whose demand cycles turn at different times, whose purchases are funded from different budget lines, and whose activity falls under different regulatory regimes. Where management reads the count and the investor reads the correlation, the resulting difference is settled at the negotiating table rather than in the reporting pack.
A second mechanism is the absence of any internal segment definition at all. In most mid-sized companies the customer breakdown lives as an account ledger in the accounting system and as personal knowledge on the commercial side; who belongs to which segment, where the boundary of a segment is drawn, and on what basis a customer's reclassification is recorded appear in no written form. A diversity claim built on that footing does not clear the existence threshold, since what exists is a verbal description rather than a verifiable structure. The reviewing party does not adopt that description; it constructs its own segment definition, redistributes the income statement accordingly, and where the resulting split diverges from the narrative, it books the entire divergence to the risk side. Who performs the definition is therefore not a technical question but one with a direct price consequence.
Documentation constitutes a separate threshold from definition, and it tends to break earlier. A segment taxonomy may have been built once, used in a single board pack, and never refreshed; or the taxonomy may exist as a field in the CRM while the discipline of populating that field is left to the individual account manager. Any inconsistency between the segment split placed in the data room and the raw extract pulled from the CRM puts not only that line item but the company's general reporting reliability under question. An inconsistency found in one item typically triggers deeper sampling across the others, and that widening extends the closing timeline; a longer timeline, in turn, erodes negotiating leverage on the sell side, generally at the moment when it matters most.
What the implementation dimension seeks is whether the definition actually governs daily operations. Where a segment taxonomy exists but pricing, delivery priority, technical support allocation, and credit limit decisions are all taken without reference to it, the taxonomy lives only in the reporting layer. The observable signal is straightforward: if two customers sitting in two different segments receive the same payment terms, the same discount band, and the same delivery commitment, the company is using segments as name tags rather than as risk profiles. Absent an operational counterpart, the decision to enter a new segment also remains a response to inbound demand rather than a strategic act; and diversity assembled by response disperses at the same speed once the condition that produced it changes.
Measurement is the dimension most often left empty, because while most companies produce a segment-level revenue split, few produce a segment-level margin split. That asymmetry carries a quiet consequence: when a low-margin segment grows quickly, what appears in the consolidated statement is growth, and management presents that growth as evidence of diversification; over the same period, however, the product mix may have shifted, the working capital cycle lengthened, and the service burden per unit increased. Placing segment-level gross margin, collection period, return rate, and customer acquisition cost side by side in diligence tends to reveal that part of the apparent diversification is in fact margin erosion. The absence of measurement here is not merely an information gap; it indicates that management has not detected its own mix shift, and that observation is written against the management quality heading rather than the customer heading.
The ownership question is plain, and its answer usually resolves to the same place: who decides to enter a new customer segment, who is accountable for that segment's revenue target, and who takes the exit decision when the segment fails to produce the expected performance. In a substantial share of mid-sized companies all three answers are the founder, segments having been opened through the founder's personal network, sector knowledge, and negotiating weight. That is not a defect in terms of historical performance — more often it is the reason the company reached its present position. The investor, however, prices not past performance but the reproducibility of that performance; and a segment-opening capacity resident in one person is recorded directly as founder dependency in the repeatability test. Such a finding is typically carried into structure rather than into price: a key-person provision conditioning the founder's tenure, an earn-out trigger keyed to segment performance, or an extended non-compete undertaking.
Continuity sits one step beyond ownership and is tested against a single condition: whether the ability to move outside the current segments exists as a defined process within the company, or whether it is reinvented on each occasion. Having entered three new segments over three years is not, on its own, evidence of capacity; the evidence lies in whether those three entries were run through a common method — market sizing, pilot customer selection, price testing, a unit economics threshold, and a documented continue-or-stop criterion. Without a method, three entries are three coincidences, and coincidence does not scale. Where the entire post-investment growth case rests on opening new segments, the inability to demonstrate the mechanism beneath that case becomes the most direct justification for the discount applied to the plan.
The valuation consequence of the gap does not travel through a single channel. The most visible is multiple adjustment: a concentrated or correlated revenue base is priced at a lower multiple than a dispersed base of the same size, because the variance of its cash flow is higher. The second channel runs through the debt side, where credit committees translate customer concentration into covenant calibration, and where a single-customer share breaching a defined threshold generally produces additional reporting obligations or a reserve account condition. The third channel is transaction structure: an unverifiable diversity claim widens the scope of representations and warranties, raises the escrow ratio, and pushes a portion of the consideration into segment-linked performance mechanics. The combined effect of those three channels is frequently larger than any single negotiation conducted on the headline price.
BEIREK's intervention in this area begins by converting diversity from a narrative into a record. The first step is to construct the segment definition according to the company's own economics rather than by sector label — organised around the condition that triggers demand: the source of the budget, the calendar of the purchasing cycle, the applicable regulatory regime, and price sensitivity. Once that definition is in place, revenue, gross margin, collection period, acquisition cost, and service burden are redistributed across the segment split, and the redistribution is applied retrospectively across several periods, since a single-period split shows position while a series shows mix movement. Embedding the definition simultaneously in the CRM field, the quotation template, and the monthly management pack ensures that the split is an output the operation produces rather than a presentation item assembled after the fact.
The second step moves segment-opening capacity from a person to a structure. A threshold-based decision record governs each new segment: the rationale, the expected unit economics, the scope of the pilot, and the continue-or-stop criterion are written at the moment of decision rather than after the outcome is known; the segment's revenue and margin targets are assigned to a single role, and the requirement that this role not be the founder is applied deliberately. Segment-level variance becomes a standing item on the monthly management agenda, and a correlation review of the segment portfolio is run quarterly. Where that rhythm has operated across two reporting periods, what reaches the diligence table is not an assertion but a dated series of decisions; and a dated series answers the repeatability question in a way that no verbal defence can.
Customer segment diversity is, in the end, not a marketing heading but a resilience test measuring which single condition the revenue base depends upon. Knowing the share of a company's largest customer is straightforward; knowing the share of its largest shared condition is possible only where segments have been formally defined, measured, and assigned to a role. The question worth answering before sitting down at the diligence table is therefore narrow: what percentage of this company's revenue could stop at the same time, in the same year, for the same reason?