A key account management plan that changes next Tuesday
A key account management plan on one page: share of wallet with its source, the time budget most plans omit, and ninety day actions, not objectives.
ReadBDGL / Insights / Entrepreneurship
Thirty people said the problem was real and the thing still did not sell. That is the normal outcome of research done well, because research measures interest and interest is free. The signals that cost the buyer something are the only ones that carry information, and there are five of them.
The sequence is familiar enough to be a genre. Somebody spots a problem in an industry they know, talks to thirty people in it, hears thirty versions of yes that is a real problem, builds for nine months, launches, and sells almost nothing.
Nobody lied in those thirty conversations. The problem probably was real. What the conversations could not tell you is whether it was expensive enough, to a person with a budget, to displace whatever they are doing about it today. That is a different question and it has a different test.
Customer research is worth doing and most founders do too little of it. It tells you what the workflow actually looks like, where the cost sits, what language the buyer uses and which of your assumptions were invented. None of that is the same as evidence that somebody will pay.
The reason is structural rather than about honesty. In an interview, agreeing with you is free and disagreeing with you is awkward. The respondent carries no cost either way, so the answer carries very little information about what they would do when there is a cost. Research tells you whether you have understood the problem. It cannot tell you whether you have found a market.
Running the research half properly is its own discipline, including how many conversations is actually enough and the question habit that gets honest answers, and that is covered separately in B2B customer discovery interviews. Treat this piece as the step after it rather than a replacement for it.
It helps to be precise about the base rate, because the numbers that circulate here are usually wrong in a specific direction. The US Bureau of Labor Statistics tracks every private sector establishment from the quarter it opens, through its Business Employment Dynamics programme, and publishes the survival series as a table rather than a headline.
Reading the survival table for the most recent cohort with a full five years of history, establishments that opened in the year to March 2021, the path is this: 79.1 per cent were still open after one year, 67.0 per cent after two, 59.1 per cent after three, 52.3 per cent after four and 47.2 per cent after five.
Two things follow, and the second matters more. The first is that the popular claim of a ninety per cent failure rate is not what the official series shows; slightly more than half survive four years. The second is that this data cannot tell you anything about your idea. It counts establishments opening and closing, it is American, it covers every industry from restaurants to logistics, and it says nothing about why any individual business closed. It is a floor for calibration, not evidence about a market.
Which is the whole point. There is no published dataset that will validate your idea. The only evidence that exists about your market has to be generated by you, before you build, by asking somebody to do something that costs them something.
Every validation conversation eventually reduces to one request, and the wording of it decides whether the answer means anything.
The question that gets nothing is a version of "would you use this?" The question that gets something is a version of "will you commit to this now, on these terms, before it exists?" The second is uncomfortable to ask, which is exactly why most founders skip it and spend nine months instead.
What you are looking for is not enthusiasm. It is a buyer willing to accept a cost: money, a signature, political capital inside their own company, or a slot on their calendar they have to defend to somebody. Each of those costs a different amount and therefore proves a different amount.
These are not alternatives to each other so much as a ladder. Each rung costs the buyer more and tells you more. Most ideas die somewhere between the third and the fourth, and dying there is cheap.
| Form | What it costs the buyer | What it actually proves |
|---|---|---|
| Written problem statement in their words | Twenty minutes | The problem exists. Nothing about budget. |
| Introduction to their own colleague who holds the budget | Internal credibility | They believe it enough to be seen asking. |
| Signed letter of intent with a named price | A signature, no money | The price is not absurd to them. |
| Paid pilot or deposit before delivery | Real money, early | A budget exists and they can release it. |
| Design partner agreement with their time committed | Money and named staff hours | It is a priority this quarter, not in principle. |
The second row is the one founders undervalue. In B2B the person who loves your idea frequently cannot buy it, and the test of whether they are a real route in is whether they will spend their own standing introducing you upward. A contact who talks to you monthly but will not make that introduction has told you something precise.
The fourth row is where most of the information is, and it is worth being careful about how it is structured, because a pilot priced wrongly proves the opposite of what you wanted. The trade-offs are set out in paid pilot versus free trial.
There is a legitimate version of this and a version that burns the market you are trying to enter. The difference is disclosure, and it is not a grey area.
The honest version: say plainly that it does not exist yet, say what you are asking them to commit to and when you will deliver, name what happens to their money if you do not, and put it in writing. Buyers have seen this before and a surprising number will participate, because an early customer who shapes the product gets something real out of it.
The version to avoid is demonstrating a thing as though it is built, taking money against an implied delivery date you have not committed to, and sorting out the detail later. That converts better in the short run and it is how a founder acquires a reputation in a small market before they have a product.
A practical middle path is to make the thing concrete without building it. A clickable prototype settles a surprising number of arguments that a document cannot, including disagreements between stakeholders who both thought they agreed, and our colleagues at Rivl set out what it does and does not settle in a clickable prototype before development. Showing a screen somebody can click is also a far more honest basis for asking for a deposit than describing a vision.
Define the failure condition before you start, in writing, with a number and a date. Something like: if six named organisations in this segment are approached properly and none will sign a letter of intent at this price within eight weeks, the idea does not proceed in this form.
Without that line written down in advance, every result reads as encouraging. Three polite maybes and a reschedule feel like momentum, and momentum is what funds nine months of building.
And when the answer does come back no, read which no it was, because they are not equivalent:
If the conclusion is that the idea works but you cannot reach enough of the right people to test it, that is a distribution problem rather than a validation problem, and the sequence for solving it is in how to get your first ten B2B clients.
Pre-commitment is not proof of a business. Five signed letters of intent prove that five buyers found the price plausible on the day. They do not prove the market is large enough, that you can deliver at a margin, or that the second ten are as easy as the first five.
Some things genuinely cannot be sold before they exist. Regulated products, anything requiring a security review before procurement will engage, and categories where the buyer cannot evaluate the claim without seeing it running. In those cases the honest answer is that validation is partial and the first build is a bet, so size the bet accordingly rather than pretending the test was conclusive.
The survival figures above are American. We used that series because it is published, consistent and open. We have found no equivalent longitudinal series for Egypt or the Gulf, and we will not extrapolate one.
And a test run badly is worse than no test. Six conversations with contacts who like you, at a price you have not thought about, will produce six encouraging answers and no information. The discipline is in choosing names you have no relationship with and naming the price out loud.
Send us the segment you are aiming at, the price you have in mind, and the six organisations you would approach first. We will come back with which of the five commitments to ask for, the wording that gets a straight answer, and the failure condition worth writing down before you start.
Book a Free Consultation