Blog/PE diligence

GTM Due Diligence for PE Buyers

GTM due diligence tests whether a company's market, product, people, systems and cash can actually deliver the growth the deal assumes. Commercial diligence tells you the market is attractive. This tells you what it costs, and how long it takes, to get at it. The findings sort into fix it during the hold, reprice the deal, or walk away.

Companies get bought on growth stories the business couldn't pull off if everybody worked weekends for a year.

You can have a sane thesis, a good management team and a product customers like, and still have no idea what it takes to double revenue. Maybe the market is smaller than the deck says, or the reps spend half their week on stuff somebody should have automated three years ago. Or the product falls over the first time an enterprise security team looks at it. Each of those costs a different amount to fix, and some of them should change what you pay.

What is GTM due diligence, and how is it different from commercial diligence?

Commercial due diligence looks at a target's market, competitive position, customers and growth prospects, and tries to answer whether the business can hit the performance the deal assumes. Good commercial work already gets into route to market. Third Bridge's commercial diligence checklist, for example, asks what drives sales cycle length and conversion, and tells deal teams to look at ramp time if the model assumes more salespeople produce proportionally more growth.

What I add is the view of someone who has built sales, marketing and RevOps systems and knows where they tend to break. I want to know how the revenue engine works today, what it would cost and how long it would take to change it, and whether the people, process, product, data and cash exist to do that. Those findings should feed back into the financial and commercial work, and the other way around.

How big is the market this company can actually reach?

TAM is the number every deck leads with, and it's usually bullshit until somebody filters it. What matters for a growth plan is the part of the market you can reach: accounts that fit the product, feel the pain, have budget, and can be contacted at a cost the business can afford.

At Cyft, we started with around 40,000 MSPs as our addressable market. Then we filtered for companies with enough technicians, enough technical and AI sophistication, and a real chance of getting value out of the product, and the plausible universe dropped to roughly 5,000. It was a very different company than the 40,000 number described, and that matters a lot if you already paid for the 40,000 version.

Suppose you reach 30% penetration on both. On 40,000 accounts that's 12,000 customers. On 5,000 it's 1,500. At a $20,000 ACV, the smaller market tops out around $30 million a year. If the deal needs more than that, the company needs another segment, another product or different pricing, or you need to pay less for it.

A real market map is more than a logo count. For each segment I want fit, the pain or exposure, a budget proxy, who owns the budget and who sits on the buying committee, whether the product can actually serve them, a channel that reaches them, the sales cycle, the implementation and support load, expected conversion, CAC and how much capital it takes. Someone with experience can override a score. I'm fine with that, as long as they write down why and then test it. Killing a segment on vibes is how you drop a good market. The investor should also see the raw data and the assumptions, not only the conclusion.

Most of the headline market isn't reachable

Cyft's headline market narrows from approximately 40,000 MSPs through enough technicians, AI readiness and customer type to approximately 5,000 with real fit.CYFT / MARKET FITThe headline marketnarrows with qualification~40,000 MSPsEnough techniciansAI-readyRightcustomertype~5,000 real fit

Most of the headline market was never reachable.

What does it cost to reach the market, not just size it?

At AirOps, where I used to work, we built account universes the way I'd build one for a fund: start wide, then narrow. We pulled from tools like Apollo, ZoomInfo and Clay and cut by employee count and revenue, for example more than 2,000 employees when we were testing upmarket. Then we layered on Ahrefs traffic, marketing and RevOps headcount, org structure and who held buying authority. For budget we estimated from public company filings, G&A and marketing spend, industry benchmarks for SEO and ads, and marketing headcount as a rough proxy.

Airlines kept coming up in that work. They had huge web traffic, big marketing teams, programmatic SEO spend, and a lot of exposure to AI answers eating clicks from search. For context, a Pew Research Center study of 900 U.S. adults' Google browsing in March 2025 found people clicked a traditional result link in 8% of visits when an AI summary showed up, versus 15% when it didn't. On paper airlines looked great. We still hesitated, because selling to an airline is nothing like selling to a startup.

You'd be paying for expensive account-based campaigns, flying people out to pitch in person, and sitting through procurement for six, nine or twelve months, probably with enterprise sellers who already know the industry. The vertical sometimes got set aside on judgment, not on an assumption anybody wrote down and tested, even when the metrics looked favorable.

What I do know is that going after them would have changed how much cash the plan needed and when it showed up. At a venture-backed company that can mean another fundraise before any of it pays off. At a PE-backed company it means delayed cash flow, more capital at risk and a different hold-period picture. Market attractiveness and the cost of getting to that market are two separate questions, and I'd score them separately.

AirOps ran into a related problem. Some enterprise customers struggled to operate sophisticated SEO workflows responsibly over time, and bad practices could hurt their SEO results. The company moved toward a more services-supported offering to help customers execute safely and effectively.

It does show you something to look for in any product that needs expert hands to work well. If customers need that much help, what happens to delivery cost, support load, gross margin, time to value and churn, and how much of the revenue is software versus services? You want those answers before you underwrite software margins.

Selling to a startup versus an airline
The startup lane moves faster: shorter sales cycle, light procurement, lighter touch and cash sooner. The airline enterprise lane moves slower: longer sales cycle, heavy procurement, account-based campaigns, in-person pitches and cash later, requiring more capital.SALES MOTIONSame market appeal.Different cost to reach it.Startup buyerShorter cycle / Light procurementLighter touch / Cash soonerAirline enterpriseLonger cycle / Heavy procurementAccount-based campaignsIn-person pitchesCash later / More capital

Can the product support the growth plan?

If the thesis needs the company to move upmarket, look at what the product can actually support. A company that sells fine to small customers might not have the SSO, MFA, integrations, admin controls or service capability an enterprise buyer expects. That's engineering time, implementation work and money, and it belongs in the model.

For product-led companies it goes further, because activation, adoption, expansion and retention are the GTM motion. If people sign up and never get to value, throwing more top-of-funnel spend at it just makes the problem more expensive.

Should you hire more salespeople before fixing the revenue engine?

When you buy a lower-middle-market business, you're usually buying something that works at a small scale and betting you can make it a lot bigger. Before hiring five more salespeople, I'd ask whether the company has the people, processes and product in place to scale.

My answer to the hiring question is simple. Don't hire five more people if you don't have a RevOps function. Seriously, don't fucking do it. I see this constantly: a company adds sellers without funding RevOps, and then blames the reps when results are bad. The system was broken before they showed up.

Minimum useful RevOps looks like this: CRM deal stages that match the real sales process, call recording with some visibility for coaching, usable contacts, LinkedIn profiles and buying-committee data, funnel instrumentation you trust, workflow automation, clean handoffs from sales to customer success, and a named owner for systems and data. You don't need a full-time admin at every company. You do need somebody accountable.

Here's a CRM pattern I like. When a new opportunity gets created, the system researches and enriches the likely buying committee and flags missing personas while the deal is still open. Say you're selling a $50,000-a-month advertising package and nobody has talked to the CMO. I want the system to catch that before the deal dies. When a deal closes, the customer success manager should get the deal history, stakeholders, promises made and relevant product usage without anyone rebuilding it by hand.

Now say you've got ten reps and only three hit quota. I'd start with the three. Go see what they do differently: tools they bought with their own money, personal spreadsheets, deeper account research, MEDDPICC-style qualification, multithreading, properly scoped proofs of concept. Figure out what can be copied and automated, support the rest of the team, and judge them over a quarter or two, adjusted for how long your sales cycle runs. Some people may still need to go. I'm skeptical that coaching reliably turns B players into A players.

Is quota attainment the same as revenue quality?

No. Somebody can hit quota and still bring in customers who aren't profitable or don't stick around. Trace each customer back to the rep who originated it, then look at churn, expansion, implementation burden, gross margin and what happened after the sale. If hunters and farmers, or a separate customer success team, own different parts of the journey, keep acquisition and post-sales accountability separate. Compare cohorts by seller, segment and channel.

Channels need the same treatment. A good inbound or content lead might close faster and retain better than an expensive outbound lead, and an account-based program might cut the number of touches it takes to get into an account. Before you conclude anything, figure out whether the spend is poorly targeted, the creative is weak, or the campaign automation and attribution are broken.

How should a company run AI experiments without losing control?

I'm a believer in decentralized AI experimentation as long as somebody in RevOps can see what's being built and governs it. Let technically capable sellers and SDRs try Claude Code, Codex and automation tools. Some will waste a week rebuilding a dashboard that already exists. Others will find a workflow that produces pipeline. The central team's job is to tell the difference and make the good experiments repeatable.

At AirOps I gave SDRs access to tools including HubSpot APIs, Lemlist and Email Bison. Some of it produced meetings and pipeline and some of it burned time. Watching what happened let me discourage the waste and standardize the winners.

AirOps also built an AEO readiness scorecard that gave prospects useful research whether they ever bought or not. I ran broad outbound and account-based campaigns centrally, and the SDRs used the scorecard for tighter territory-level campaigns in Email Bison. That showed us which messages and targeting worked, and we could fold it back into the bigger motion.

Build the governance in from the start. Nobody gets unrestricted production credentials. Each tool gets the least access it needs, high-impact actions need approval, and you log what the agents do. The OWASP Gen AI Security Project says the same thing under Excessive Agency. Data governance, privacy and contractual obligations are real constraints too.

The architecture doesn't need to be one giant AI platform. Generally the CRM holds sales records, an outbound tool like Email Bison runs campaigns, a database like Supabase can consolidate activity, something like Metabase can report coverage across accounts, territories and the market, and a GitHub repo can hold custom automation, documentation and shared context. None of those is the single source of truth. Context lives in several governed systems.

What share of the target market did you reach this quarter?

One goal we worked toward at AirOps was reaching roughly 20% of the addressable market each quarter through email, advertising and other channels. You can only do that if you have an actual account universe and activity records. Most companies can't answer a much simpler question: what percentage of our target market did we touch this month?

Be picky about what counts. Attempted contact, delivered outreach or impressions, real engagement, qualified pipeline and revenue are all different things. The KPI I like for investors is revenue and qualified pipeline per unit of market coverage, next to what that coverage cost. You can't measure that without a data foundation under it.

How do you underwrite the growth engine, not just the forecast?

We can't model anything without a baseline. Before somebody tells me how they'll double revenue, I want to know exactly how revenue gets generated today. That means ARR or revenue, customer count, ACV, gross margin, acquisition spend by channel, sales cycle length, win rates, quota attainment, rep productivity, churn and retention, territory coverage, how much of the market they reach, and how clean the data is.

Then model the alternatives: add sellers, make the current team more productive, automate research, fund RevOps, improve conversion, shift marketing spend, enter a new segment, fix retention and expansion. For each one, show ARR, CAC, payback, cash needs, EBITDA contribution and timing, with sensitivities. Keep observed values, estimates and assumptions separate.

Suppose a software company has $10 million in ARR and a sponsor that wants to double it during the hold. Management says the answer is five more salespeople. At $20,000 a month per fully loaded seller, that's about $1.2 million a year before ramp time. Maybe management is right. Before spending it, I'd ask what else that money could do. What if the current sellers got another ten hours a week of actual selling because we automated research and admin? What if better buying-committee intelligence lifted win rates? What if we moved some of the outbound budget into channels that bring in customers with shorter cycles and better retention?

Don't assume one option beats the other. $50,000 a month in ads costs about as much as two and a half of those sellers. Which one wins depends on qualified pipeline, conversion, ACV, retention, gross margin, CAC, payback, time to close, seller capacity and cash.

Be careful with the formulas. Bookings aren't ARR, and bookings minus seller cost isn't EBITDA. You have to account for timing, gross margin, churn, services cost and acquisition expense. CAC payback in months is CAC divided by monthly new gross profit. With a hypothetical $120,000 CAC and a customer paying $60,000 a year at a 75% gross margin, that's 120,000 divided by 3,750, or 32 months.

I'd run three lenses: status quo, efficiency and infrastructure improvements, and aggressive growth investment. The deliverable is a recommended scenario plus a transparent model operators can poke at. A slider isn't evidence that an improvement can happen, so every assumption should tie to a credible intervention and its cost. The model should stay alive through the hold, not die as a static spreadsheet.

Fix it, reprice it, or walk away?

PE firms often buy fixable problems on purpose, so the job is sorting findings, not scaring anybody. A sales team that needs better systems can usually improve. These are the findings that should worry you:

1. The qualified market can't support the growth the exit case needs.

2. The product can't serve the segments the plan needs without a slow, expensive transformation.

3. Customer acquisition economics stay unattractive even after reasonable operational improvements.

I'd turn every material finding into an action, a cost, a timeline and a risk. Some belong in the first hundred days. Some change the three-to-five-year plan. Some are bad enough that the deal doesn't work at the proposed valuation, and then you reprice or walk. A broken CRM is generally fixable. A broken growth thesis might not be.

Fix, reprice, or walk

Material finding, then cost and timing, then thesis sensitivity. Branches lead to fix it in the hold, reprice the deal, or walk away.THESIS SENSITIVITYWhat does the findingmean for the deal?Material findingCost and timingThesis sensitivityFix it in the holdReprice the dealWalk away

Each material finding gets a cost and a timeline, then gets tested against how sensitive the thesis is to it.

The GTM diligence framework at a glance

Each row is one workstream, in the order I'd run them.

GTM due diligence framework: nine workstreams
WorkstreamQuestionEvidenceIf it fails
1. Market reality and buying capacityHow big is the qualified, reachable market, and can those buyers pay?Bottom-up account universe with raw data, budget proxies and stated assumptionsThe exit case needs another segment, product or price, or a lower purchase price
2. Cost to reach the marketWhat do the sales cycle, channel and capital requirements look like in the segments the plan needs?Sales-cycle data, CAC by channel, procurement and implementation requirementsThe growth plan needs more capital or more time than the model assumes
3. Product and segment readinessCan the product serve the customers the plan targets?Security and integration gaps, activation and retention dataEngineering cost and delay go into the model, or the thesis changes
4. Sales and marketing economicsWhat does a customer cost, and how fast does it pay back?CAC, payback, win rates, quota attainment, pipeline by sourceReallocate spend or reprice
5. People, process and RevOpsCan this team execute at twice the size?Deal stages versus the real process, call recording, data quality, ownership of systemsFund RevOps before adding headcount
6. Revenue qualityDo the customers the team wins stay and expand?Cohorts by rep, segment and channel; churn, margin, implementation burdenQuota numbers overstate the real engine
7. AI and data foundationsCan the company measure and govern its own experiments?Market-coverage reporting, tool access policies, audit logs, data governanceAI spend becomes theater, or a risk
8. Financial scenariosWhat do realistic paths cost and produce?A baseline, three scenarios, sensitivities, assumptions tied to interventionsThe forecast is a hope, not a plan
9. DecisionFix, reprice or walk away?Each finding with an action, cost, timeline and riskThe decision gets made on instinct

Diligence workflow

Five workstreams in order: Market and buyer access, Product readiness, People/process/data, Financial model, Prioritized decisions.GTM DILIGENCEBuild the evidence.Then make the decision.Market and buyeraccessProduct readinessPeople/process/dataFinancial modelPrioritized decisionsEach workstream feeds the next

Each workstream feeds the next. The financial model is where the findings turn into a price.

What questions should a deal team ask?

These are the questions I'd put in front of management, and then check the answers against data.

1. How many accounts are in your qualified market, how did you build the list, and who can show me the raw data?

2. What percentage of that market did you contact this quarter, and what share engaged, became pipeline and closed?

3. Which segments does the growth plan need, and what does the sales cycle look like in each?

4. What does it cost, in dollars and months, to get a customer in the segments you're planning to enter?

5. Which parts of the product would break or need to be built to serve those customers?

6. Do deal stages in the CRM match how you actually sell, and who owns the CRM and its data?

7. Are calls recorded, and are next steps and stakeholders captured without manual work?

8. Of your reps, who hits quota, and what do they do differently from those who don't?

9. For customers acquired in the last two years, what are churn, expansion and gross margin by originating rep, segment and channel?

10. What happens after a deal closes, and who is accountable for onboarding and retention?

11. Who is experimenting with AI tools, what access do they have, and who reviews what they build?

12. What is your baseline today: ARR, customers, ACV, gross margin, CAC by channel, win rate, cycle length and retention?

13. If we add five sellers, what is the fully loaded cost, ramp time and expected output, and what else could that money do?

14. If the market, the product or the acquisition economics don't support the plan, which of those can we fix, and at what cost?

What does GTM due diligence need to deliver?

What I find useful at the end is a defensible map and tiering of the market, a view of how the revenue engine really runs, an assessment of the people and systems needed to scale it, a data and AI readiness check, quantified scenarios with sensitivities, and a prioritized plan for the first hundred days and the rest of the hold. Each piece should connect a specific operational change to the investment case.

Where Apero fits

Apero does forward deployed GTM engineering for PE and VC backed companies. We build GTM systems, evaluate sales and marketing operations and implement automation, so we come at diligence as people who build the thing we're evaluating.

We also do GTM due diligence for funds, before an acquisition or during the hold. It complements commercial and financial diligence.

If you have a deal in diligence or a portfolio company you're worried about, get in touch, or see how we work with private equity portfolios. I wrote about de-risking founder-led sales separately.

Sources

Third Bridge, Commercial due diligence checklist (published 15 Oct 2026), thirdbridge.com. Used for the scope of commercial diligence.

Pew Research Center, Google users are less likely to click on links when an AI summary appears in the results (22 July 2025), pewresearch.org. Used for the 8% versus 15% click comparison.

OWASP Gen AI Security Project, LLM06:2025 Excessive Agency, genai.owasp.org. Used for least-privilege and approval controls.

Questions

What is GTM due diligence in private equity?

It tests whether a company's market, product, people, systems and cash can deliver the growth in the investment thesis. It looks at what it costs and how long it takes to reach the market, how the revenue engine runs today, and whether the plan is realistic, then sorts findings into fix, reprice or walk away.

How is GTM diligence different from commercial due diligence?

Commercial diligence asks whether the market is attractive and whether the business can win in it. GTM diligence adds a practitioner's view of how the revenue engine operates, what it would cost and how long it would take to scale it, and whether the people, processes, product and data are in place. The two should share findings.

Why does a company's real market shrink after qualification?

TAM counts everyone who could buy. The reachable market counts only accounts that fit, feel the problem, have budget and can be contacted affordably.

Should a portfolio company hire more salespeople to hit its growth plan?

Not before checking the revenue engine. Without RevOps, clean deal stages, call recording, usable data and automation, new sellers inherit a broken system. Study what the top performers do, put someone in charge of RevOps, and compare automation, marketing channels and conversion fixes against headcount.

What are the possible outcomes of a GTM diligence review?

Fix it, reprice it, or walk away. A messy CRM or a team that needs better systems is usually fixable during the hold. If the qualified market can't support the exit case, the product needs a slow transformation, or acquisition economics stay poor after reasonable improvements, the price or the thesis has to change.

Does Apero do GTM due diligence?

Yes. Apero does forward deployed GTM engineering for PE and VC backed companies, and also does GTM due diligence for funds before an acquisition or during the hold. The work covers a market map, a baseline of the revenue engine, a read on people, process, product and data, and scenarios that tie findings to price.

More questions

Let's talk shop.

We try to give away as much GTM alpha as we can on a first call. Tell us what you're thinking about, we'd love to help.