A go-to-market diagnosis rests on four sources, in this order: lost deals, customers who succeed, pipeline data, and salespeople in the act. The order matters, because lost deals say what won ones leave out. The deliverable fits on one page and names a single cause, not a list of observations. Budget three weeks of listening and one of arbitration, not ninety days of audit.
“Run an audit of our go-to-market.” The request often arrives in those words, and the misunderstanding is already in them. An audit describes a state. A diagnosis names a cause. The first produces a document everyone approves and nobody applies, the second produces an uncomfortable decision.
The difference lies in the method, and above all in the order in which the sources are opened. Here is the order we follow, and why it is not interchangeable.
Why a workshop is not enough
The temptation is to gather the team for two days and map the customer journey on a whiteboard. That is useful for alignment, and insufficient for diagnosis, for a simple reason: an internal workshop produces only what the team already knows. It surfaces disagreements, which has value, but it brings in no new information.
And the cause of a stalling go-to-market almost always sits in a shared blind spot. Something everyone has taken for granted for so long that nobody states it any more. A workshop does not go looking for it, it circles around it.
Source one: lost deals, first
This is the most useful inversion of the usual method, which starts with happy customers because they are easier to reach and more pleasant to listen to.
Lost deals say what won ones leave out. A customer who signed narrates their purchase after the fact, rationalising it, and willingly credits you with qualities that carried no weight. A prospect who did not sign remembers precisely what stopped them, and has no reason to spare you.
Look for three categories: those who chose a competitor, those who chose to do nothing, and those who vanished without deciding. The third is the most instructive and the least called. In most B2B markets the main competitor is not another vendor, it is inaction.
Five conversations are enough for a pattern to appear. They are uncomfortable to ask for, and that is precisely why they never happen.
Source two: customers who succeed, not customers who are happy
Second source, and mind the distinction: you are not looking for satisfaction, you are looking for results. A happy customer who gets nothing measurable will leave at the next budget review.
Take those who renew, expand and refer, and look for what they had in common before they even met you: size, maturity, a dedicated team, a triggering event. That core reveals your real target, which often diverges from the stated one, and it is the heart of the work described in why an ICP that is too broad kills growth.
Ask them too for the words they would use to describe you to a peer. Those are your best messaging elements, and they almost never resemble your homepage, as we explain in positioning: say who you serve.
Source three: data, to measure, not to discover
Third only, and the rank matters. Data says how much, rarely why. Opened first, it steers the diagnosis towards what is measured rather than towards what is important.
Four numbers carry most of it: stage-by-stage conversion, cycle length by segment, acquisition cost by channel, and retention in revenue. The fourth is most often missing from sales dashboards, although it determines whether the growth on display is real or merely replacing what is being lost.
Sometimes the data is unusable. That is a diagnostic result in itself: it means decisions are made on intuition, and the first correction is a foundation, not a sales action. We describe that work in building your revenue operations foundation.
Source four: salespeople in the act
The last source, and the most neglected: listening to real sales conversations, recorded, not recounted. Ten calls, five that closed and five that did not.
What you are looking for is not individual performance. It is the gap between the official message and the message that actually comes out of your reps’ mouths. That gap is almost always large, and it is informative both ways: sometimes the field is right and the official message is theoretical, sometimes the message is good and was never transmitted.
Once documented, that gap is the raw material of a usable playbook, a subject we cover in a sales playbook that survives your best closers.
The deliverable: one page, one cause
This is where most diagnoses fail. After three weeks of listening you have twenty observations, all true, and the temptation is to deliver all of them. The result is a document the executive approves and files, because it leaves the arbitration to them.
A useful diagnosis arbitrates for you. It names one dominant cause, shows why the other observations follow from it, and puts a number on it. “Our target is too broad” is not a diagnosis. “Half of sales capacity goes to a segment worth 12 % of revenue and 40 % of churn, around 180,000 of capacity a year” is one.
It then proposes a first move, dated, with a named owner. Not a twelve-month plan: the next ninety-day move, about which you will be able to say at the end whether it worked.
What ninety days actually means
The diagnosis itself takes four to eight weeks depending on how available people are. The rest of the period is for executing the first correction and measuring its effect.
That distinction matters, because a diagnosis that runs three months without changing anything is a disguised audit, and because the only way to validate a cause is to act on it. If the move corrects the symptoms, the cause was right. If not, you have eliminated a hypothesis and you go back to whichever source was listened to least.
It is also what separates strategy from execution, and the point where most plans get lost, as we write in strategy vs execution.
Where to start this week
Commission nothing. Take the list of your last ten lost deals and call five of them. Not a questionnaire, not a form: a fifteen-minute call opening with “we are not trying to sell you anything, we are trying to understand what tipped the decision”.
By the third conversation you will know whether a pattern is forming. It is the cheapest part of the diagnosis and the most valuable, and it is almost always the part nobody has done.
Frequently asked questions
- How long does a go-to-market diagnosis take?
- Four to eight weeks for the diagnosis itself, including two to three weeks of interviews and analysis of real data. The ninety days in the title cover the diagnosis plus the first corrective move: a diagnosis that runs three months without changing anything is an audit, not a diagnosis.
- Where do you start with little data?
- With interviews, which need no tooling. Ten conversations, three of them on lost deals, produce more usable material than six months of poorly maintained dashboards. Missing data is itself a diagnostic result: it tells you that decisions are being made on intuition.
- Do you need an outside view to diagnose your GTM?
- Not necessarily, but you need someone who did not build the system being examined. Internal teams know the ground better than any outsider, and they share one blind spot: what seems obvious to them is never questioned, and that is often where the cause sits.
- What is the deliverable of a good diagnosis?
- One page that names a single cause, puts a number on it, and proposes the first corrective move with an owner and a date. A forty-page document listing twenty unranked observations produces no decision: it hands the arbitration back to whoever commissioned it.