Commercial Due Diligence: What It Covers, and What It Can't See

The Anatomy of a Full CDD Scope

A full commercial due diligence engagement is a set of workstreams, not a single question. Five appear in almost every scope: market size and growth, customer research, competitive landscape, pricing and value proposition, and a view on the revenue plan. Larger scopes add channel economics, regulatory outlook, or expansion cases for new geographies and segments.

Each workstream carries a distinct burden. The market work establishes how big the opportunity is and which way it is moving. Customer research tests whether the people paying for the product would keep paying, and why. The competitive work maps who else is in the room and how the ground between them is shifting. Pricing work asks whether the company sets its prices or accepts them. The forecast review ties all of it back to the numbers the deal is being underwritten on.

The workstreams feed each other. Interview themes sharpen the sizing segments. Win-loss patterns populate the competitive map. Pricing evidence stress-tests the forecast. A scope that runs them as parallel silos produces five separate answers instead of one view.

Scope comes in sizes. A red-flag review runs one to two weeks and answers a single question: is there a reason to stop. A full scope runs three to six weeks, staffed by a small team, and produces a report the advisory firm signs. Sell-side processes add a third variant, vendor commercial due diligence, commissioned by the seller and shared with every bidder. Useful, but written to survive scrutiny rather than to invite it.

When scoping an engagement, three requests earn their place in the letter:

• A named minimum of churned-customer interviews, not just current references.
• A bottom-up market build, not a cited figure from an industry report.
• An explicit view on the revenue plan, line by line, rather than a general market blessing.

The scope letter is the cheapest place to fix a CDD engagement. Everything after it is negotiation.

Market Sizing Done Properly

Three letters carry most market-sizing conversations. TAM is everyone who could conceivably buy the category. SAM is the portion the company's actual product and geography can serve. SOM is what it can realistically win in the planning window. Only the last number belongs anywhere near an underwriting case, and it is the one least often shown.

The common abuse is top-down arithmetic. Take a headline figure from an industry report, apply a hopeful share, call it opportunity. The number is unfalsifiable and usually enormous, which is why it survives. Serious sizing starts from the bottom: count the buyers, segment them by size and need, attach a realistic price to each segment, and add it up. The bottom-up build is slower and smaller. It is also the one that can be argued with, which is the point.

Triangulate before trusting. Three legs: the top-down clip, the bottom-up build, and a competitor check that sums what the known players actually earn today. If the three land within a reasonable band of each other, the size is probably understood. If they disagree by an order of magnitude, nobody knows, and the report should say so.

Watch the category definition as closely as the arithmetic. Widening the definition is the quiet way to inflate a market: a payroll product becomes a workforce platform, and the addressable figure triples without a single new buyer appearing. The definition that matters is the one buyers use when they allocate budget, not the one the industry report used to group vendors.

Two further habits separate careful work from decorative work:

• Growth matters more than size. A modest market growing steadily beats a vast one that is flat, and the growth rate deserves its own evidence, not an extrapolated line.
• Every sizing hides assumptions about price and adoption. Ask what would have to be true for the number to hold, then check whether any of it is true yet.

A market figure without a method attached is a slide, not a finding.

Customer Interviews That Earn Their Keep

Voice-of-customer work is the spine of commercial due diligence, and the part most often done politely rather than well. For a mid-market B2B target, a typical program runs fifteen to thirty depth interviews. The count matters less than the mix.

Four groups belong in the sample. Current customers, weighted toward the accounts that carry the revenue. Churned customers, the hardest to reach and the most informative. Prospects who evaluated the product and chose someone else. And channel partners or resellers, who see the company from the side. A sample built only from happy current customers answers a question nobody asked.

Who supplies the names decides what the interviews can find. A management-provided reference list is curated by definition. The working fix is to request the full customer list and draw the sample independently, with churned accounts drawn at the rate they actually occur.

Sequencing helps too. Run the first handful of interviews early, revise the guide, then complete the program. The opening conversations usually rewrite the questions.

What to ask matters as much as whom. Satisfaction questions produce satisfaction answers. Questions anchored in real decisions produce evidence:

• Walk through the last renewal decision. Who was involved, and what almost changed it.
• Which alternatives were evaluated, and how far did each one get.
• If the price rose ten percent tomorrow, what would actually happen.
• Which parts of the product were bought but are no longer used.

Listen for tense. A customer who says the product still runs their billing and one who says it used to are describing two different companies. Listen for the unprompted competitor mention, the feature nobody opens, the champion who has left. Fifteen honest hours of this reliably outperform a hundred survey responses, because surveys collect opinion at scale and interviews collect specifics one at a time. Specifics are what a deal decision needs.

Competitive Dynamics and the Moat Question

A competitive assessment has two halves: who is in the market today, and which way the ground is moving. The first half is a mapping exercise. Direct competitors, adjacent substitutes, and the two rivals every deck forgets: building it in-house, and doing nothing. In many categories the do-nothing option wins more deals than any named vendor.

Win-loss reality grounds the map. Claimed differentiation lives in the deck; realized differentiation lives in the deals the company actually won and lost, and in the reasons buyers give unprompted. Two questions sharpen it quickly: which named competitor took the most revenue from the company last year, and which one the company took the most from.

The second half is harder and worth more. A snapshot of market share says little; the trajectory says a lot. Who is winning the deals that came up for grabs this year. Where is new funding flowing. Which competitor is hiring salespeople in this segment. A leader losing three points a year and a challenger gaining them sit on the same chart and support opposite theses.

Moat claims deserve specific tests, because every deck makes one:

• Switching costs. Measured in migration months and integration depth, not in sentiment. If customers say leaving would be painful but churned accounts left in a weekend, there is no moat.
• Network and data advantages. If real, they show up in retention curves and win rates. If they only show up in the narrative, they are a story.
• Cost position. Verifiable against gross margins, not asserted.
• Distribution lock-in. Contracts, shelf space, embedded integrations. Check the terms, not the logo slide.

The cleanest single probe: where does price competition show up. A company with a moat loses deals on fit. A company without one loses deals on price, and its discount ledger will say so long before management does. Differentiation that never appears in pricing or retention is positioning. Positioning is not a moat.

Cohorts, Churn, and What Retention Really Says

Retention is the closest thing commercial work has to a verdict. Customers who have used the product and choose to keep paying are casting an informed vote. The workstream's job is to read that vote correctly, which is harder than it sounds.

Start with the distinctions. Logo churn counts departed customers; revenue churn counts departed money; the two diverge whenever accounts differ in size. Gross revenue retention measures what was kept before any expansion. Net revenue retention adds upsell back in and can look healthy while the base quietly erodes. All four numbers describe the same company differently, so ask for all four.

Then insist on cohorts. A blended churn rate averages the loyal customers of five years ago with the wobbling ones of last quarter and hides the difference. Cohort curves, each acquisition year tracked separately, show whether recent vintages behave like earlier ones. The shape matters most. Curves that flatten indicate a durable core; curves that keep sloping indicate a leaky one. Where the curve flattens sets what a customer is worth, and no lifetime-value arithmetic is more reliable than that flattening point.

Cut retention by segment before averaging it. Enterprise and small-business cohorts often behave like two different companies under one brand, and the blend flatters whichever story needs flattering. Retention by acquisition channel repays the same treatment. Customers who arrived through a partner rarely behave like the ones who arrived through outbound.

Three abuses recur often enough to check for by name:

• Annual contracts defer churn. A cohort sold twelve months ago has not yet had its first chance to leave.
• Downgrades classified as retention. A customer paying a third of last year's fee is retained in logo terms only.
• New-logo growth masking base decay. Fast acquisition can hide rising churn for years.

The layer-cake chart, revenue by acquisition cohort stacked over time, makes all three visible on one page. If management cannot produce it, that is itself a finding.

The Pricing Power Test

Pricing power is the most honest summary of a commercial position. A company that can raise prices without losing customers is delivering more value than it charges for. One that cannot is renting its revenue. Commercial due diligence should establish which of the two is on the table.

The evidence is behavioral, and most of it already exists. Start with realized price, not list price. List prices are marketing; the invoice ledger is fact. Then look at four things:

• The last price increase. When it happened, how much of it stuck, and who left. A company that has never taken one is untested, which is not the same as strong.
• Discount variance. If similar customers pay widely different prices for the same product, price is negotiated deal by deal, and the power sits with the buyer or with individual reps, not with the company.
• Contract escalators. Whether the annual uplift written into contracts is enforced or quietly waived at renewal.
• Win-loss by price point. Whether deals are lost on price, and at what threshold.

Realization also reads as a trend. A slow slide in average realized price per unit, masked by volume growth, is pricing weakness compounding quietly, and it shows in the ledger years before it shows in the narrative.

Interviews add the forward view. The renewal-decision question and the ten-percent question, asked of real customers, are rough instruments but directionally useful. Stated willingness to pay always exceeds revealed willingness, so weight what customers did over what they say.

The test that matters most sits in the deal itself. If the underwriting case assumes future price increases, ask when the company last proved it could take one. A thesis that depends on unproven pricing power is a bet on an event with no precedent, and the report should call it that in plain terms.

Who Commissions the Work, and When

Most commercial due diligence is commissioned by a buyer with a deal already in hand. Private equity funds are the heaviest users, typically after a letter of intent is signed and an exclusivity window has opened. Corporate acquirers commission it for larger or less familiar targets. Lenders occasionally require it before underwriting debt. Growth and minority investors commission lighter versions, weighted toward market and retention. Sellers commission the vendor variant to smooth a competitive process.

Timing follows deal mechanics. The full scope usually runs inside exclusivity, in parallel with the financial and legal workstreams, on a three-to-six-week clock. Some buyers run a short red-flag scope earlier, before the letter of intent, to avoid spending exclusivity on a market that could have been ruled out from the outside. On proprietary deals with no competing bidders, the work can start earlier and breathe more.

Who does the work spans a wide range: large strategy consultancies, specialist boutiques focused on a sector, and in-house deal teams for smaller checks. Fees scale with scope, and scope should scale with the size of the check and the novelty of the market. A fund buying its fifth company in a category it knows may need two weeks of confirmation. A corporate entering an adjacent market it has never sold into needs the full instrument.

One structural honesty check applies regardless of who is hired. By the time a full scope is commissioned, the buyer has usually spent months forming a view and some money defending it. The brief can quietly become confirm, not test. The tell is whether the scope contains a genuine kill question, a finding that would end the deal if it came back true. Scopes written without one tend to return the answer they were designed to return.

Reading a CDD Report Like a Skeptic

A commercial due diligence report is an argument, and arguments are best read from the evidence backward. Start in the appendix. The methodology pages state how many interviews were done, who the respondents were, and where the market numbers came from. Five minutes there calibrates everything the executive summary is about to claim.

A short interrogation list covers most of the risk:

• How many churned customers were actually interviewed. Not planned; interviewed.
• Who supplied the interview sample, the advisor or the management team.
• Which market-sizing leg carried the final number, and what happens to the deal case if the bottom-up figure is used instead.
• Where does the report disagree with management. Every honest engagement finds at least one disagreement. A report that confirms every management claim was either scoped to confirm or was not listening.
• What would have to be true for the growth case to hold, and how much of it is evidenced rather than assumed.

The forecast section deserves its own pass. Note which lines the advisor adjusted and which passed through untouched. An untouched line is not an endorsed line; it may simply have fallen outside the scope.

Language carries signal too. Phrases like directionally consistent, management believes, and unable to independently verify are load-bearing hedges. They mark the exact places where the advisor declined to stand behind a number. Read them as coordinates, not filler.

Color-coded summaries deserve one specific habit: read the ambers. Red findings were argued over and survived. Green findings were easy. The amber items are where judgment got negotiated between draft and final, and they repay a direct conversation with the team that wrote them. One question closes that conversation well: what almost changed your mind. The answer is usually worth more than the section it refers to.

Where Commercial Due Diligence Goes Wrong

The failure modes of commercial due diligence are consistent enough to list, which also makes them checkable.

• Confirmation scope. The work is commissioned after conviction has formed, and the brief rewards validation. The fix is structural: a kill question in the scope letter, and a standing invitation to bring bad news early.
• Curated samples. Interviews drawn from a management reference list measure the company's best relationships, not its market. Draw the sample independently or discount the findings accordingly.
• Top-down numbers surviving into underwriting. The headline TAM travels from the report into the deal memo while the caveats stay behind. Carry the bottom-up figure forward instead.
• Static competition. The landscape is mapped as of today, with no view on trajectory, funding, or entry. Six months later the map is wrong.
• Opinion collected as evidence. Interviews that ask how customers feel rather than what they did produce warm noise. Anchor every question in a decision that actually happened.
• Compression. Exclusivity clocks shrink the interview program first, because interviews are the slowest part. The part that gets cut is the part that was the point.
• Reconciled forecasts. The revenue view gets bent until it supports the price already agreed. When the evidence and the thesis disagree, the report should say so, not split the difference.
• The filed report. Findings that never reach the value-creation plan. The churned-customer themes, the pricing headroom, the competitive threats: these are the first hundred days' agenda, written before close, and often left unread after it.

None of these argue against the workstream. Most are cheap to prevent at scoping and expensive to discover at the investment committee. They argue for reading the work the way the work reads companies: on behavior, not on claims.