A buy-side QoE looks backward. The deal needs to look forward too, and that is what commercial diligence does. Across the nine commercial DD engagements I sat on between 2023 and 2025 — SaaS, B2B services, contracted-services, and one multi-channel consumer-products process — the same three workstreams produced almost all of the deal-altering findings: customer concentration walked deeper than the headline number, churn examined on cohorts rather than the blended rate, and pipeline interrogated against the seller's own conversion history. The work is less commoditised than the QoE. Fewer providers do it well, the deliverable varies more by firm, and it is the work that most often surfaces the deal-killing risk that the QoE never finds. In about 40% of the engagements I worked, a cohort or pipeline finding moved the deal-value calculation more than any single QoE adjustment did. Here is the methodology, anchored to what Bain, McKinsey, OC&C, and L.E.K. publish on commercial DD, and how each finding usually lands in the purchase agreement.
01 Customer concentration — six lenses, not one
The headline number — top-5 customer concentration as a percent of revenue — is the easy part. Almost every CIM presents it. The commercial diligence walks deeper. Top-5% of revenue is a single lens; it is not the diagnostic. In our practice we run six lenses against the customer base and look at where they disagree. The places where the lenses diverge are where the real concentration risk lives.
The first lens is the conventional one — top-1, top-3, top-5, top-10, and top-20 share of revenue, plotted on a trailing 3-to-5-year basis. The level matters less than the trajectory. A 32% top-5 holding flat for four years is a different business from a 32% top-5 that was 22% three years ago. The second lens is concentration by gross-margin contribution, not revenue. Big customers are often discounted to win; we have seen top-3 customers contribute 38% of revenue and only 22% of gross margin. The diligence cares about cash-generation concentration, which is the gross-margin view, not the revenue view.
The third lens is the Herfindahl-Hirschman Index on customer revenue shares — the sum of squared revenue percentages, expressed as a single number. HHI captures tail behaviour the top-N view cannot. A business with 300 logos where 60% of revenue sits in 7 accounts has an HHI that flags the structural concentration even when top-5 looks acceptable. Borrowing from the antitrust regulatory thresholds, HHI above roughly 1,800-2,000 is the band where concentration risk is material and should be priced. The fourth lens is segment-level concentration — by industry vertical, geography, channel, and product line. We routinely write the diagnostic line: "top-10 logos = 32% of ARR, but 81% of revenue is concentrated in one end-market." That sentence is the finding, not the 32%.
The fifth lens is the contracted-versus-transactional split. Re-compute concentration on contracted revenue only. A 20% top-customer share on a contracted multi-year MSA is not the same risk as a 20% top-customer share on a T&M project relationship. The CIM rarely makes this split clean; the diligence does. The sixth lens is customer-grouping of related entities. Sellers present the customer ledger by billing entity. Buyers (and we, on the buyer's behalf) roll subsidiaries up to the corporate parent. "Customer A" plus "Customer B" turning out to be the same Fortune-500 parent has produced step-change concentration findings on roughly a third of the engagements I have advised on. It is mechanical work and it changes the conclusion.
Customer concentration is not a number. It is a relationship-quality assessment that the number is a proxy for. The diligence is which lens you use, and whether the lenses agree.
02 The relationship behind the number
Concentration analysis without relationship-quality work is half the diligence. Once the six lenses produce the picture, the walk goes deeper on the top 10-20 customers individually. For each, the commercial DD assembles a one-page factbook covering tenure and history, product breadth, integration depth, pricing and contribution margin, dependency on individual relationships on either side, and the contract terms that frame the relationship. The factbooks are what get read at the investment-committee meeting, not the dashboard.
Tenure and history is the simplest credibility test. A top-customer relationship that has been in place for nine years, has expanded through three buying centres, and has no notable disputes on file is a very different risk from a 14-month-old relationship that has expanded fast because the seller did a custom build-out. Both look the same in the top-5 line. Product breadth and integration depth is the second credibility test. Is the relationship one product or several? Are there embedded integrations, custom workflows, or operational dependencies on the customer side that raise switching costs, or is the seller substitutable? L.E.K. and OC&C both put workflow embeddedness at the centre of their CDD playbooks, and we apply the same lens.
The pricing and margin walk by top account is where the deal-altering findings sometimes surface. We have seen "trophy logos" that the seller marketed as evidence of category leadership turn out to be heavily discounted, loss-making at gross margin, and being retained as marquee references rather than economic accounts. The dependency-on-individuals test is the next one: how many top relationships are owned one-person-deep by the founder, the head of sales, or a single account director, versus institutionally managed by a customer-success team, a key-account organisation, or a quarterly-business-review cadence? Founder-owned and rainmaker-owned relationships compound with the sales-team-retention risk we cover in section five.
The legal overlay is the last piece, and the most often skipped on rushed processes. For each of the top 10-20 customers we map term and renewal mechanics (auto-renew vs explicit, evergreen vs fixed-term), termination rights for cause and for convenience with notice periods, minimum commitments and take-or-pay clauses, exclusivity, change-of-control provisions, and SLA penalty regimes. The change-of-control clauses are deal-specific risk — a customer with the right to renegotiate or terminate at change-of-control is a customer whose concentration number is conditional on the close itself surviving. LawFlex and the standard commercial-DD playbooks insist on this overlay and we have seen it move purchase price more than once.
03 Churn — what cohorts show that the blended rate does not
For subscription, recurring-revenue, contracted-services, or any business with renewing customers, the churn analysis is the heart of the commercial diligence. The blended churn rate the seller presents is the starting point. The cohort analysis is the work. The methodology is well-established across the SaaS-investor community — Bessemer, OpenView, ChartMogul, and the strategy-firm CDD practices use broadly the same playbook — and is more diagnostic than the blended number in roughly four out of every ten engagements I have worked.
The cohort walk works by acquisition vintage. Group customers by the month or quarter they first paid. Track each cohort by life-month — M0, M1, M2 — through the available history. Build both a logo retention table (count of customers active in month N divided by the cohort size at M0) and a revenue retention table (MRR or ARR from the cohort at month N divided by starting MRR at M0). Compute gross dollar retention (excluding expansion) and net dollar retention (including expansion) on the revenue cohort. The two tables produce the curves the diligence reads against.
Pattern one — the hero vintage masking deterioration
The most common cohort finding is a blended churn rate that hides a meaningful gap between old vintages and new ones. A 10% blended annual logo churn with 2021 cohorts at 90% retention at M12 and 82% at M24, 2022 cohorts at 86% at M12 and 74% at M24, and 2023 cohorts already down to 80% at M12, is a business whose forward-looking churn is materially worse than the blended rate suggests. The 10% is averaging great old cohorts against weaker new ones. The forward valuation case has to anchor on recent-cohort behaviour, which is materially different.
Pattern two — channel quality masked by mix
Cohort retention split by acquisition channel almost always shows variance. Referral and partner-sourced cohorts retain at the top of the band — frequently 90%+ at M12 with NRR well above 100%. Outbound-sourced cohorts run a step lower. Paid-search-sourced cohorts in mid-market SaaS routinely run 70% logo retention at M12 with NRR flat or below 100%. If the seller has scaled paid-search bookings in the most recent two years, the blended view will lag the channel-quality deterioration until the new cohorts age into the average. The diligence surfaces this and re-builds the forward forecast with channel-weighted retention assumptions.
Pattern three — customer-size segments
Blended retention can also mask a stable enterprise core under heavy SMB churn (or vice versa). A 15% blended annual churn might be 30% in the SMB tier, 12% in the mid-market tier, and 6% in the enterprise tier. If the buyer's investment thesis is upmarket, the enterprise retention is the relevant number and 15% blended understates the strategic durability of the base.
Two more diagnostics matter. The first is gross-versus-net retention together — a big gap (GRR 85%, NRR 120%) is a leaky bucket with strong upsell, which is fine if the upsell motion is scalable and not concentrated in two accounts. A small gap (GRR 95%, NRR 102%) is a stable base with limited expansion runway. Both are credible businesses; they have different growth assumptions. The second is the customer-interview workstream. Where time and budget allow, we run 8-15 phone interviews with current, recently-onboarded, and churned customers. We ask about renewal drivers, perceived switching costs, alternative-supplier consideration, and any planned changes. The qualitative work either reinforces the cohort curves or contradicts them — both are useful. L.E.K., Adience, and Bain all centre customer interviews in their CDD methodology, and we have not yet finished a commercial DD where the calls did not change at least one paragraph in the final report.
04 Pipeline credibility — three tests in increasing order of discomfort
The seller's management presentation includes a pipeline and a forward plan. The commercial diligence assesses whether the pipeline is real. The work begins with a full export of the CRM pipeline for the trailing 12-24 months — opportunity ID, creation date, stage history, owner, segment, deal size, expected close date — reconciled to the GL's closed-won bookings line. The reconciliation step is the first diagnostic; CRM data that does not tie cleanly to the GL is itself a finding before any analytics happen.
Once the data is clean, three tests run in increasing order of how uncomfortable they get for the seller, and in increasing order of how diagnostic they are.
- 01 Test 1 — probability-weighted pipeline value against actual historical conversion. Compute stage-by-stage conversion rates from the trailing 24 months by segment, channel, and rep. Re-weight the current forward pipeline using those empirical conversion rates. Compare to the seller's presented forecast. A seller forecasting 32% conversion of late-stage pipeline when their trailing-24 actual is 19% has a credibility problem the diligence has to name. We routinely see gaps of 8-15 percentage points between management forecast and empirical conversion; the gap is the pipeline haircut.
- 02 Test 2 — pipeline composition and shape. Is the pipeline concentrated in one or two large opportunities (high variance, low credibility) or distributed across many smaller opportunities (lower variance, higher credibility)? Are there aged opportunities sitting in late stage for more than twice the median sales cycle (red flag — likely zombies)? Are the close dates being repeatedly pushed quarter-over-quarter (red flag — sales-team slippage that the forecast has not absorbed)? Are deals single-threaded on one stakeholder on the buyer side, or multi-threaded across three or more (single-threaded is materially less credible). The composition view tells you the variance around the forecast even where the central case looks supportable.
- 03 Test 3 — customer-reference calls on pipeline opportunities. Where the data room permits and the seller is willing to facilitate, we call a sample of named buyers in the pipeline and ask about actual purchase intent and timing. This is the most diagnostic test and the most uncomfortable to ask for. Some sellers refuse; the refusal itself is information. When the calls do happen, the conversion rate on called-buyer opportunities re-weights the forward forecast and is the cleanest single piece of evidence we get on pipeline credibility.
Pipeline composition tells you the variance around the forecast. Customer-reference calls tell you whether the forecast is real. The third test is uncomfortable and the most diagnostic.
05 Sales-team retention as a pipeline-credibility factor
The pipeline is, in practice, owned by individual salespeople. The relationships behind the pipeline live in their phones, their LinkedIn, and their muscle memory. If the top three salespeople by pipeline value are not retained through close and the early integration period, the pipeline credibility drops materially regardless of how clean the CRM looks. The commercial DD surfaces this risk explicitly and quantifies it.
The quantitative work is well-documented. A Penn State / Fortune-500 study on sales-representative departures put the annual revenue loss at 13-18% on accounts whose rep transitions without a planned, industry-experienced reassignment. New hires are systematically less effective than incumbent reps at retaining accounts in the transition period, and successor reps with high individual performance histories do not consistently reduce the loss — fit with the customer profile matters more than headline performance. Mercer post-acquisition workforce data, often cited in the M&A-advisory literature, puts the share of critical talent that leaves within 18-24 months post-deal at roughly 40%. Sales leadership is disproportionately represented in that 40%.
On the pipeline side, the modelling we apply to inherited at-risk pipeline is: conversion probability drops 10-20 percentage points on opportunities where the incumbent rep departs without a planned successor; sales-cycle length extends 10-25% on those same opportunities; and where more than 30% of the forward pipeline is concentrated in one rep, the buyer should either secure that rep through a binding retention agreement before close or haircut the pipeline by 20-50% in the underwriting case. Building Radar and the broader 2025-2026 sales-operations commentary frame the 30%-on-one-rep threshold as the standard alarm bell, and we have used it as the diligence trigger on every engagement since 2024.
The diligence also walks process-versus-heroics. A pipeline generated through a defined ideal-customer-profile, repeatable signal-based outbound motion, multi-threaded stakeholder mapping, and disciplined CRM hygiene is fundamentally more credible than a pipeline generated by one or two relationship-driven principals working their networks. The first is acquirable; the second is the people, not the asset. Strategy-firm commercial-integration research — across Bain, McKinsey, BCG, and HBR — has been consistent for several cycles that proactive sales-force retention and 90-to-120-day comp-plan alignment correlates with a 15-30 percentage-point higher probability of meeting revenue-synergy targets versus deals that treat the sales organisation as an afterthought.
06 What each finding usually does to the deal structure
Commercial DD findings rarely stop a deal. They re-price it, and they re-shape the consideration mix. Three patterns recur.
Concentration findings typically result in earnouts, escrow, or specific reps and warranties keyed to the top-customer relationship surviving through a defined post-close period. A top-1 customer at 22% of revenue with a change-of-control clause in their MSA is the textbook case where the seller gets 70-85% of headline consideration at close and the balance keyed to that customer being retained for 12 or 24 months. The earnout sizing is not arbitrary — it is sized to approximate the deal-value loss the buyer would suffer if the customer walked, discounted for the probability of departure that the diligence flagged.
Churn findings — particularly recent-cohort deterioration not visible in the blended rate — typically result in working-capital peg adjustments (lower than the seller would otherwise have negotiated) and revenue-retention earnouts. We see ARR-based earnouts in subscription-software deals tied to gross-dollar-retention thresholds at 12 and 24 months, with seller consideration scaling against actual retention versus the diligence-implied curve. The mechanism transfers the cohort risk from buyer to seller in the period where the seller still has influence over the renewal motion.
Pipeline-credibility and sales-team findings most often result in deferred consideration tied to actual pipeline conversion, plus buyer-funded retention pools for named key sales staff. Retention bonus sizing in the deals we have seen runs 40-60% of base salary for the CRO or VP Sales, 20-80% of base for principal AEs and rainmakers (50-100% where one individual controls a disproportionate share of pipeline), and 10-30% across the broader sales team tiered by quota size and tenure. The timing is almost never fully at close — common structures are 50% at close and 50% at 6-12 months post-close, or 25-33% at close with the balance vesting over 12-24 months. The retention bonuses are typically seller-funded for the pre-close portion and buyer-funded for the post-close tranches, with the post-close tranche sometimes substituted for buyer equity or a long-term-incentive plan as part of the broader sales-leadership integration.
Each finding becomes a structural element in the deal, not a stop-the-deal signal. The structural elements compound. The equity payable at close drops accordingly. Sellers who arrive at the process with the diligence work pre-done — clean cohort curves in the data room, top-10 customer factbooks ready, named retention agreements already signed with key reps — clear at meaningfully higher effective valuations because the buyer has fewer reasons to defer consideration. This is the same preparedness-premium logic we wrote about in the working-capital piece at /blog/working-capital-peg-defense-operator-view/ and in the buy-side QoE work at /blog/buy-side-qoe-multi-site-healthcare/ and /blog/buy-side-qoe-dtc-cpg-what-it-finds/. The dollar value of being prepared is consistent across commercial DD, QoE, and the working-capital negotiation.
07 Sector nuances — same spine, different emphasis
The three workstreams travel across sectors but the emphasis shifts.
B2B services — agencies, consulting, IT services, managed services
The mix of project, fixed-fee, retainer, and managed-services revenue is the first split, and concentration looks different on the recurring-retainer subset than on the project subset. Win-rate on re-competes and RFPs becomes a leading indicator of forward churn. Multi-threading on the customer side is critical — agencies and services firms with a single sponsor on a $2M account are routinely re-priced when that sponsor leaves. Human-capital concentration on the seller side compounds the customer concentration: a principal consultant or partner who owns the top-3 accounts is a key-person risk that has to be retention-papered or priced into the consideration.
SaaS, subscription, recurring data services
The cohort retention work dominates here. GRR and NRR by segment, channel, and plan are non-negotiable; switching costs measured by integration depth and embedded workflows are the secondary credibility test; usage metrics (seats, API calls, daily active users) become leading indicators of churn-to-be. The customer-reference call workstream is often easiest in SaaS because the buyer pool is structured and named, and it is most diagnostic.
Contracted services — BPO, facilities, logistics, maintenance
Contract structure is the diligence centre. Tenure, remaining term, renewal mechanics, termination-for-convenience clauses, SLA penalties, and indexation to inflation are all priced in. The backlog and contracted-revenue-coverage ratio against the cost base is the operational-credibility test. Contract re-tender economics matter — what does the margin look like if a top contract is re-bid against incumbent assumptions versus open competition.
Product, distribution, and consumer-products operating companies
Retailer and channel concentration replaces logo concentration. Category role on shelf (must-have versus nice-to-have) and exposure to retailer private-label strategy are the qualitative overlays. Volume-versus-price dynamics, contribution margin by account, exposure to promotional intensity, slotting fees, and rebate obligations all feed into the same six-lens concentration framework but with retail-specific data. The buy-side QoE piece on consumer-brand deals at /blog/buy-side-qoe-dtc-cpg-what-it-finds/ pairs with this section — the QoE finds the historical reserve-adequacy gaps, and the commercial DD finds the forward-looking concentration and channel-credibility risk. Both run for a complete consumer-brand diligence.
08 Questions before the commercial DD opens
Three questions for sellers and one for buyers, sequenced as the practical checklist we run on every engagement.
- For sellers: is the cohort churn analysis available in the data room by channel, product, customer-size tier, and acquisition vintage — and does it reconcile to the GL?
- For sellers: is the pipeline presented with probability weights that match the seller's actual historical conversion rates by stage, segment, and rep, with the reconciliation table included in the management presentation?
- For sellers: are named retention agreements signed and dated with the top three sales producers and the top two customer-success leaders before the process opens, sized in line with market practice (40-100% of base, 50/50 between close and 6-12 months out)?
- For buyers: what is the budget for customer-reference calls and pipeline-buyer interviews during commercial DD, and is the seller prepared to facilitate them — including the named buyers whose deals materially impact the forward forecast? The hospitality and healthcare diligence frameworks at /blog/buy-side-qoe-hospitality-groups/ and /blog/buy-side-qoe-multi-site-healthcare/ both treat reference calls as a budget line, not an optional extra. Commercial DD is the same.
Frequently asked questions
How does commercial due diligence differ from quality of earnings (QoE)?
What customer concentration thresholds should trigger commercial DD concern?
When does cohort retention analysis reveal something blended churn does not?
How do commercial DD providers test whether a sales pipeline is real?
How does sales-team retention affect deal value and pipeline credibility?
How do commercial DD findings flow into the purchase agreement?
What is the typical timeline and scope of a mid-market commercial DD?
Sample: 9 commercial DD engagements 2023–2025 across SaaS, B2B services, contracted services, and one multi-channel consumer-products process. Findings aggregated; no individual transaction or party identified.
Methodology framing: Bain & Company PE Due Diligence practice; L.E.K. Consulting Commercial Due Diligence practice; OC&C Strategy "value-added" commercial DD; Big-4 (PwC/EY/KPMG) standard CDD frameworks; Umbrex commercial DD practitioner guidance.
Cohort retention methodology: ChartMogul GRR/NRR definitions; Bessemer Cloud Index benchmarks; OpenView PLG retention research; Stripe and Maxio SaaS cohort guides; Headline Deepdive and ConsultEFC 2026 cohort frameworks.
Sales-team retention impact: Pennsylvania State University study on sales-representative departures (13-18% annual sales loss on affected accounts); Mercer M&A workforce continuity research (40% critical-talent attrition within 18-24 months post-deal); LSA Global high-attrition cost analysis; MBO Ventures, TKO Miller, Walden M&A, and Class VI Partners stay-bonus structures.
Full source list at content-pipeline/research/commercial-dd-customer-concentration-churn/sources.md in the Putra & Co content pipeline.