What to Look for in Marketing Due Diligence When Buying a Company

You are under LOI. The clock is the product. Exclusivity is what you paid for, and it is also the constraint: a few weeks to decide whether the growth story in the CIM belongs in the price, in the reps, or in a walk-away memo. Financial diligence tells you whether last year’s earnings were real. Marketing due diligence tells you whether next year’s customers still show up at a cost the model can live with.

I run this from the operator seat. A CIM is a sales document. A data room is a curated archive. Operator interviews are a negotiation. None of those, on their own, open the ad accounts, reconcile platform conversions to the CRM, or tell you what happens on Day 1 if the founder, the agency, or a single paid channel disappears. That is the substance of marketing diligence: the engine’s pros and cons, not a vendor list. If you want the firm shortlist, I already ranked that in best marketing due diligence companies for PE.

Marketing diligence workstreams at a glance

Use this as a scoping map, not a scorecard filled in from the CIM. “Good” is evidence you can take to IC. Red flags belong in price, structure, or a 100-day fix with a real budget. They are not automatic kill-shots.

You are under LOI. The CIM is not evidence.

Exclusivity is a window, not a verdict. In confirmatory diligence you can usually get further than a pre-LOI outside-in scan: invoices, agency SOWs, platform exports, Search Console, CRM source reports, legal folders, and (if the seller is serious) read-only access to ad accounts and analytics. You still will not see everything. Dark social, founder-sourced pipeline, and “the agency will send a deck” are not in the data room until you ask, and sometimes not then. Limited access is itself a finding. Grade every conclusion as verified in-account, reconstructed from invoices, or inferred from public signals.

Three artifacts, three jobs. The CIM is a sales document: TAM slides, “efficient CAC,” hockey-stick media plans, agency dashboards. The data room is where you look for contradictions: invoices that do not match reported spend, CRM stages that do not match “leads,” legal correspondence that never made the risk factors, contracts that do not transfer. Operator interviews tell you who actually knows the account. Ask who can log into Google Ads without a ticket, who wrote the last landing page, and what broke the last time CPCs jumped. Specifics are the function. Narrative is the title.

Marketing DD informs three decisions. Price: a growth story that is bought, concentrated, or unmeasured should not clear the same multiple as a diversified, measured engine. Reps: claims substantiation, consent, pixel/PHI posture, account ownership, and pending marketing litigation belong on the schedule. Walk-away: some engines are not fixable inside the hold at a cost the thesis can bear: a banned category, a TCPA book of business, a founder-only pipeline with no second channel and no budget to build one.

Buyers skip the cons. The CIM only prints the pros, and what looks like a feature is often a ceiling. A “high ROAS branded search program” is frequently a harvest engine with no unbranded demand behind it. A “lean team” is key-person risk. A “performance agency of record” is often the only institutional memory, sitting in someone else’s login. A “national TAM” is often a licensed-state problem. Treat every feature as a possible constraint until you have proven it is an asset.

TAM: sellers inflate it. Paid media cannot spend it.

Most TAM slides are category theater: a top-down industry number, a share, and the implication that the gap is yours. Diligence has to separate total market, serviceable market (what you can legally, operationally, and geographically serve), and obtainable demand (what you can acquire at a CAC the model funds). Those are three different numbers. Conflating them is how “we only have 2% share” becomes the most expensive sentence in the book.

That 2% line is usually a red flag, not an opportunity. If a company has been selling for years and still sits on a sliver, the constraint is rarely awareness. It is offer, channel, sales capacity, reviews, licensing, or a TAM that was never real at the price they charge. Share-gap stories assume the remaining 98% is available at the current CAC. It is not. The easy demand was already bought.

Split TAM the way an operator would. Category TAM is the consultant number, useful for commercial DD and almost useless for a media plan. Channel TAM is the inventory on Google, Meta, Amazon, affiliates, SEO, and partnerships at the targeting you can actually run. A huge category with tiny paid inventory is a small business. Paid-media TAM shrinks when CPCs are rising. If the last two years of efficiency came from cheap supply that is gone, the CIM curve is a history lesson. Ask for CPC, CPM, and conversion-rate trends by channel, not a single blended CAC. Then the rude question: if we 2× spend next year, where do the incremental customers come from, and at what CAC? If the answer is “the TAM is huge,” you do not have an answer.

Talent: title, function, and who knows the account

I care less about the org chart than about who can open the tools. Founder-led marketing can be a feature in a small company. The founder knows the customer and closes. It is a liability the moment the thesis assumes that person steps back. Write down what demand is founder-sourced, what is system-sourced, and what dies if they are on a beach in month four. Search funds and lower-middle-market PE miss this constantly; it is the empty chair in why PE staffs every function except growth.

Agency dependency is the other common shape. A competent agency can be an asset. An agency that holds the ad accounts, the pixel, the creative archive, and the only person who understands attribution is a key-person risk with an invoice. Ask who is the named operator, whether logins transfer, and the Day-1 plan if you fire them. If the “CMO” is a title on a slide and the function lives at a vendor, you are not buying a marketing team. You are buying a retainer you may not want.

Map the founder’s hours in pipeline. Material founder-sourced revenue is concentration, not culture. Look for marketing ops, not just a coordinator who traffics ads. Read SOWs for notice periods and account ownership, and whether the agency optimizes to platform ROAS or to the P&L. Ask whether the “CMO” can defend CAC, payback, and pipeline in a board pack, or only impressions. If the honest post-close plan is a senior operator two days a week, see fractional CMO for search funds and private equity, not a full-time hire the model cannot carry.

Day 1 is the test. If the founder and the agency both leave, what still runs? Campaigns in a personal Gmail, pixels on a contractor’s GTM container, a domain in the seller’s name, a Meta Business Manager that is not the company’s. Those tell you whether you own the engine. When the hold plan needs a leader in the seat, the fractional CMO market is how most LMM buyers actually staff it. Diligence should say so before close.

Fraudulent claims, lawsuits, and the liability that shows up at exit

Marketing creates contingent liability that financial diligence will not find unless someone reads the ads. FTC substantiation is the starting point: claims have to be true when they are made, with evidence in the file. Testimonials, before/after photography, “#1” and “guaranteed” language, typical-results fine print, and affiliate claims are where buyers inherit someone else’s copy. NAD challenges, state AG letters, class actions, and BBB patterns are not PR. They are a preview of what the next acquirer’s counsel will ask at your exit.

In regulated and performance-heavy categories the stack is the lawsuit. TCPA exposure from list buying, stale consent, and leads whose opt-in language does not travel to the buyer is a per-contact problem. Impaxium’s marketing compliance page publishes the statutory range at $500 to $1,500 per illegal text or call, not per campaign. HIPAA pixels on healthcare-adjacent pages have already produced $100M+ in penalties in the cases we cite there. CAN-SPAM runs up to $51,744 per non-compliant email. If the live funnel would fail that screen, you are buying a remediation project and possibly a tail.

Look in the data room and outside it. Live ads versus the substantiation file: if there is no file, that is the finding. Consent language on forms, SMS, and call tracking; purchased lists and “partners” are the usual hole. Pixel placement on pages that collect health, financial, or other sensitive data matters too. Browser-side Meta and Google tags on those pages are a known pattern; we map the enforcement in marketing compliance in regulated industries. Affiliate copy you do not control still belongs to the brand. Historical NAD, AG, class-action, and platform-policy files matter even when “we settled and moved on.” The next buyer’s diligence will be less polite than yours. Cleaning a claims mess in year four is more expensive than pricing it now, or walking.

Reviews: rating is vanity. Velocity is the conversion input.

Reviews are a CAC input, not a brand nicety. Google Business Profile, G2, Trustpilot, BBB, Reddit, physician-review sites, and NPS when it is real all change whether paid traffic converts and whether sales can close. A 4.8 with 2,000 reviews and steady velocity is a different asset than a 4.9 with 40 reviews and a six-month silence. Diligence the velocity and the mix, not the CIM screenshot.

Incentivized reviews, deleted negatives, and a profile that is pristine on Google and ugly on Reddit are the usual tells. Ask how reviews are requested and whether there is a complaint cluster the seller calls “competitors.” For multi-location and healthcare, look at location-level GBP, not the corporate average. For B2B, G2 and Reddit will tell you things customers will not say on a scheduled reference call. Thin or toxic social proof raises conversion CAC even if media efficiency looks fine in-platform. A strong, authentic review engine is one of the few durable advantages you cannot buy in a week. Price both.

Traffic concentration mix: where demand actually comes from

Build the source mix from analytics and from revenue, not from the agency’s last-click dashboard. Organic, paid, referral, email/lifecycle, marketplace, direct, and sales-sourced each need a number, a trend, and an owner. Then split branded vs non-branded. A company that looks “efficient on Google” because most converting queries are its own name does not have a paid engine. It has a harvest program sitting on brand demand. That can be a healthy business. It is a terrible scale story.

One-channel dependence is the classic kill-shot: Meta, Google, Amazon, or a single affiliate that takes the thesis with it if efficiency drifts 30% for two quarters. Geographic concentration is the quieter version: one DMA, one state, one marketplace storefront. Landing-page and domain risk is the operational version: one domain, one unowned GTM container, ads pointing at a URL the seller’s cousin registered. Reconcile GA4 to CRM to billed revenue. Separate brand paid from non-brand paid. Name the top affiliate and the contract that keeps them. Check who owns domains, pixels, audience lists, and Business Manager. Personal ownership is a close condition. Channel concentration is one of the four Diligence Sprint questions for a reason. Diversification after close is a budget line.

ROI and ROAS: last-click theater and the board pack

ROAS as presented in most CIMs is a platform number optimized by the platform. Last-click ROAS on branded search will look heroic. Prospecting will look expensive. The truth lives in blended CAC (media plus creative, people, tools, agency, and the overhead that actually belongs in acquisition) against revenue finance will sign. Channel ROAS is a diagnostic. Blended CAC, payback on gross margin, and cohort LTV are the underwriting stats.

Interrogate LTV like a quality-of-earnings add-back. Is it a cohort, or a spreadsheet that multiplies average order value by hoped-for retention? Are churn and discounting in the number? Are sales-sourced and founder-sourced customers mixed into “marketing LTV”? Modeled conversions have to be checked against observed cash. If the two cannot be reconciled, you do not have ROI. You have a story. Incrementality is the adult version: some spend harvests demand that would have converted anyway. You will not run a perfect geo-lift in a three-week exclusivity window. You can still ask whether branded search was ever turned down as a test, and whether “efficient CAC” survived a period when the founder stopped sending personal emails.

CAC trends matter more than the trailing twelve-month average. A blended number that is flat because mix shifted toward cheaper brand harvest is not stable. It is a fuse. The board pack hides vanity ROAS, incomplete denominators, and marketing credit for revenue the founder actually produced. Rebuild the waterfall. If you cannot, that is the finding. You cannot underwrite a multiple on numbers the target itself cannot reproduce. On the PE advisory page we publish the operator track record as 2 exits including a PE sale at 18× EBITDA, $46M+ in annual ad budgets managed, and 50% ROI delivered at eight-figure spend, claims to diligence like any other. The standard is the point: measurement that would survive an IC challenge.

Constraints: why “just scale it” often cannot

A clean CAC does not mean the engine can grow. Constraints are why the CIM curve dies on contact with operations: creative fatigue and a brand-guideline freeze that prevents testing; compliance review that adds three weeks to every landing page; inventory or slotting limits; a sales team already at capacity, so extra leads become extra ignore-rate; licensed geographies that cap where you can advertise; platform bans or policy strikes sitting in the ads account; tracking so broken the algorithm cannot optimize even if you wanted to spend.

This is where CIM “pros” flip. A premium brand with tight legal review is a feature for reputation and a constraint for performance. A licensed professional-services footprint is a feature for moat and a constraint for TAM. A marketplace business is a feature for distribution and a constraint for margin and account risk. Write the constraints into the model as capacity, not as culture. If tracking is broken, you do not have a scale plan. You have a measurement rebuild that has to precede spend, the same sequence a serious fractional CMO engagement would run after close.

The budget required to grow: hold versus scale

Ask for two budgets, not one. Hold is what it costs to keep current volume: media at today’s CPCs, creative refresh, tracking maintenance, people or agency. Scale is what it costs to produce the CIM curve: more media, more creative testing, measurement that can take the load, and humans who can convert the volume. Buyers underwrite the second and fund the first. That gap is where IRR goes to die.

A real growth budget is not a media line. It is media plus creative plus tracking plus people. Working capital for paid media matters in cash businesses: you spend this week for customers who pay later. If the model assumes revenue growth without a step-up in working capital, the growth story is fiction. If the honest plan requires a step-change in spend (a second channel, a new creative engine, a senior operator, a compliance rebuild) and that number blows the returns model, you do not have a value-creation plan. You have a CIM. “Keep the CIM curve” pretends historical CAC will survive 2× spend. It almost never does. Diligence should produce an honest plan: what growth you can buy at a known payback, what you cannot, and what leadership seat you need to run it.

How this changes the model

If the work is done in time, it changes three documents. Price and structure: concentration, unmeasured CAC, claims exposure, and a growth story that is mostly bought belong in valuation, earnouts, holdbacks, and specific indemnities, not in a vague “upside” paragraph. Holdco / company budget: the Year-1 marketing number should be the honest hold-plus-scale plan, including working capital, not the seller’s trailing spend plus optimism. The 100-day plan, lightly: Day 1 is account ownership, tracking integrity, and who sits in the growth seat. It is not a rebrand and not a new agency. If diligence found a broken measurement layer, that is the first workstream; spend waits. The risk register should already be the first page of the plan. This is not a separate 100-days essay; the plan is the output of this workstream.

That is why this sits beside QoE and commercial DD rather than inside them. QoE validates the past. Commercial DD asks whether a market is real. Marketing DD asks whether this company can acquire the next customer at a cost the thesis requires, and what it will cost you when the answer is “only if we rebuild.”

Frequently asked questions

What should you look for in marketing due diligence when buying a company?

Look at tracking integrity, TAM quality versus obtainable demand, talent and key-person risk, claims and marketing lawsuits, reviews, traffic concentration, true CAC/ROI, operational constraints, and the budget required to hold versus scale. Those workstreams tell you whether the growth story is durable, bought, or a walk-away.

When should marketing due diligence happen if you are already under LOI?

Marketing due diligence should run in confirmatory diligence as soon as exclusivity starts, with enough time left to change price, reps, or the decision to close. Outside-in work can start earlier, but account access, invoices, and CRM reconciliation are what make the findings underwritable.

How does marketing due diligence affect purchase price?

Marketing due diligence affects price when the CIM growth story is concentrated, unmeasured, founder-dependent, legally dirty, or more expensive to continue than the model assumes. Findings typically land as a multiple adjustment, an earnout on durable channels, a holdback for claims exposure, or a walk-away.

What is channel concentration risk in an acquisition?

Channel concentration risk is the share of new demand that would disappear if one paid platform, affiliate, geography, or founder-sourced motion broke. A healthy mix with a branded versus non-branded split is underwritable; a one-channel engine should be priced as a single point of failure.

How do you evaluate TAM in marketing diligence?

Evaluate TAM by separating category size from serviceable market and from demand you can actually buy at current CPCs and conversion rates. “We only have 2% share” is usually a red flag that remaining demand is not obtainable at the seller’s CAC, not proof of white space.

Can marketing lawsuits and FTC exposure transfer to the buyer?

Marketing lawsuits and regulatory exposure can follow the business depending on deal structure, especially TCPA consent defects, unsubstantiated claims, affiliate copy, and tracking pixels on sensitive pages. Diligence should read the live ads, forms, and pixel stack, not just the litigation schedule the seller included.

What is tracking integrity in marketing due diligence?

Tracking integrity is whether pixels, CRM, and finance tell the same story about source, conversion, and revenue, including dark social and modeled conversions that never become cash. If you cannot trust the measurement, you cannot underwrite CAC, ROAS, or the growth curve.

How much marketing budget do you need after buying a company?

You need two numbers: the budget to hold current volume and the budget to produce the growth the model assumes, covering media, creative, tracking, people, and working capital for paid media. If the CIM curve requires a step-change in spend the holdco will not fund, the growth story is not in the deal.

Diligence is a walk-away tool

The point of this work is not to decorate an IC deck with a marketing appendix. It is to decide whether you still want the company at this price once you have seen the engine. Plenty of targets are good businesses with a harvested brand, a founder who still sells, and an agency that reports well. Buy those on purpose, with a budget and a seat to professionalize growth. Do not buy them as if they were a scalable acquisition machine.

When the tracking is fiction, the TAM is a poster, the claims are a lawsuit, or the only person who knows the account is leaving, that is not a 100-day opportunity. That is a walk. The buyers who get this right treat marketing diligence the way they treat QoE: as permission to confirm, to re-trade, or to stop. If you are in the window and you need that read in deal time, that is the Diligence Sprint on our PE advisory practice. If you still need a vendor map, start with the ranking. If you need the substance, you are already in it.