Commercial Due Diligence: What Buyers Read

Why buyers ask for the raw data instead of your deck

The old sequence was simple: you built an investment story and presented it. Today the request is to look at the underlying data directly. That moves the first impression to a place you cannot influence on the day of the meeting — into whatever your company has been recording over the past several years.

Mark Miller co-founded Carlsquare, an M&A boutique for B2B software with offices in Berlin, London and San Francisco. Manuel Hartmann spoke with him for the LinkedIn newsletter Diary Of A CRO and sums up Miller's observation like this: put the PowerPoint away, give me access to your customer cube, I will look myself. The cube is simply the full set of raw records about your customers — cohorts, expansions, downgrades, cancellations, retention.

From the LinkedIn newsletter Diary Of A CRO by Manuel Hartmann.

The uncomfortable consequence for a CEO or CRO is this. If one segment of new customers reliably leaves after twelve months, that fact sits in the records regardless of how good your narrative is. You now shape the first impression through the reality you built, not through the slide you are building.

What Net Revenue Retention actually tells a buyer

Most founders treat retention as a reporting line. A buyer treats it as the answer to one question: does this business grow on its own, or only for as long as someone keeps feeding it new logos? Get that wrong and every growth projection in the room loses its footing, because the engine behind it is unproven.

Two numbers do the work, and they are often confused. The first asks how much revenue stays when nothing is added — that is Gross Revenue Retention. The second asks how much stays once expansions with the same customers are included, which is Net Revenue Retention. A strong net figure can hide a weak gross one: you lose many small accounts while growing a handful of large ones. That is not automatically bad, but it is a decision, and you want to explain it before somebody interprets it for you.

Underneath both sits the cohort — customers grouped by the month they started, tracked separately. A single blended average hides the thing a buyer is actually hunting for: that last summer's intake churned out at twelve months while everything else held.

Three answers to the same question is the real finding

The dangerous situation is not a weak number. It is three people in the same company giving three different answers to the same question. At that point the debate is no longer about performance but about credibility — and the definition gets handed to you from outside, at a moment you did not choose.

Hartmann describes exactly how that sounds in the same newsletter: asked for Net Revenue Retention, one company produces "I think around 100 percent", then "92", then "110". Over time it harmonises. The only real question is whether you drive that harmonisation or whether a diligence process forces it on you.

The most honest observation in that issue cuts against any framework worship. Miller and Hartmann both recall a founder who, asked about his Net Revenue Retention, asked back what that even was — and then named an outstanding figure. No benchmark, no model, simply because he had helped customers who came back and asked what he would solve next. Customer impact beats vocabulary.

An unusual number is a question, not a diagnosis

A value that stands out gets read as a verdict, and two expensive reactions follow. Either somebody rushes to repair the number instead of its cause, or the team panics about something that was never a problem. Both happen under observation, which is the worst possible moment for either.

Miller uses an image that captures the role of data precisely: the customer cube is the blood test, not the doctor. The values often show immediately that something is off. But a blood test does not diagnose, it asks the next question. Are you drinking too much? Moving too little? What sits behind the unusual value gets resolved in conversation.

Our view: we build revenue systems so that every metric carries its own follow-up question. A retention figure without the cohort underneath it is an alarm with no address — you know something is wrong, but not which intake, which segment, or from which month onward. Those three details decide whether a finding turns into action or into a meeting.

Strong retention is built, not inherited

Plenty of founders treat a loyal base the way they treat weather: pleasant when it happens. That posture costs real money, because it points the budget at new logos while the faster lever sits inside the existing base — with people who already know you and whom you have already helped.

The six months are the point of that figure. It is not a by-product of five patient years but the result of a deliberate motion aimed at the accounts that carry most of the revenue. The same pattern shows in the work with Perspective: 5x More Deals and >2x Growth From Existing B2B Customers with HubSpot.

Our view: retention is a sales discipline, not a condition. So we start with the twenty accounts that carry your revenue rather than with a blended average. The average tells you how the company is doing; the top twenty tell you what a buyer will conclude about the quality of your customer relationships.

Where those numbers come from, and why it is not a two-week project

The data almost always exists somewhere. The problem is that it does not exist in a form you can query: a contract value sits in a document, a cancellation in an inbox, an expansion in a spreadsheet. Every analysis then turns into archaeology, precisely when there is least time for it.

For a retention calculation to work without reconstruction, four things have to live in the same system as fields rather than prose: the customer's starting month, the current contract value, every change to that value with its date and direction, and the reason behind a cancellation. In our projects that system is HubSpot. It is where the deal, the contract value and the lifecycle stage converge, which makes it the one place a cohort can be assembled on demand instead of pieced together from email.

This is deliberately not a feature tour. What matters is not which reports HubSpot ships, but whether your team maintains those four pieces at all. A workflow that writes the date and the direction on every contract change is dull, and it saves you weeks later.

If that sounds like a multi-quarter effort: Nestermind reached 2x Pipeline Visibility and Faster Sales Cycles on HubSpot within 4 weeks with us. Making a system readable is craft with a known sequence. What takes years is the history inside it — which is why you start now rather than when the first buyer writes.

For the same logic applied to a funding round, see our piece on sales due diligence and the CRM: that one covers the five data points an investor pulls from the CRM, this one covers the customer base a buyer reads before an exit.

Where the numbers stop, the deal begins

There is a moment when every figure is on the table and the thing still wobbles. Someone asks the question data cannot answer: may I talk to your customers? Anyone who responds by hastily launching a survey produces something that smells like a sales process — and it gets read exactly that way.

The alternative costs nothing but consistency: be in genuine, structured contact with your customers all the time, not only by accident. What do you like about us? What is missing? Do that continuously and you have no artificial questions to ask when a buyer knocks, you have a development to describe. Hartmann quotes a sentence that sounds nothing like a commissioned reference: "We did not see automation as urgent, customers wanted it, we shipped it, they are happy."

Then there is a gap almost nobody staffs. Hartmann names it directly: what is usually missing is a chief adoption officer. There are owners for marketing, sales, success and product, but rarely anyone accountable for whether what was bought actually gets used and then delivers impact. His view: software is nearly always three times too expensive for the adoption it really achieves, and at the same time three times too cheap for the impact it could deliver.

Our view: trust is one of the few advantages that is hard to copy, which is why we treat customer references as part of the revenue system rather than as a marketing task. A network of customers who talk openly and at length is not something a competitor assembles after launching tomorrow. It is demanding, and that is probably the point.

The reassuring half of this picture: weaknesses are not disqualifying. A financial investor is looking for upside, meaning places where value can be added. A team that acquires well and neglects the base is not a flaw, it is a plan. What remains decisive is the very human question of whether someone will hand you their core business for the next three to five years.

What to do differently over the next twelve months

There is no date on which preparation begins. There is only the question of whether your revenue system could let someone look inside today. Answer that with no, and what you lose is not weeks of work but the years of history that cannot be created retroactively.

The rest is posture: run your company as if a buyer had read access to your system tomorrow, because that is what they ask for first today. More on how we build and operate such a system is on our Revenue Enablement page.