A founder called me last spring, three weeks after what he described as the best investor meeting of his life. Revenue had tripled year on year. Two logos most people in the room would recognise. The partner had stayed forty minutes past the slot and asked for the data room on the way out. Then nothing. One polite email asking about churn definitions, answered the same day, and silence.
He wanted to know what to write next. My question was different: what had he not told them? He paused, then said the thing he had not said in the meeting. Sixty-one percent of revenue came from one customer, and that customer’s contract was a twelve month pilot that had been extended twice on email rather than renegotiated. He had not lied. He had simply built a deck where that fact did not appear, and he had assumed that if nobody asked, it was not material.
It was material. It was the only thing that was material. And by the time they found it on their own, in a spreadsheet he had handed over without really reading it himself, the conversation had changed from “should we invest” to “what else is he not telling us.”
Traction opens the door. It does not close the deal.
There is a belief, almost universal among first-time founders raising institutional money, that investors are in the business of buying upside. You show growth, you show a market, you show a team that can execute, and the capital follows in proportion to the size of the dream. It is a natural belief because it describes exactly what happens in the first meeting. The first meeting is about upside. That is why the partner stayed forty minutes over.
Everything after the first meeting is a different exercise. It is a hunt. Not a hostile one, usually, but a systematic search for the reason not to proceed. An institutional investor will write perhaps eight to twelve cheques from a fund over several years and will see several thousand companies to do it. The constraint is not finding good companies. The constraint is avoiding the ones that look good and are not, because a blow-up costs the fund money, costs the partner internal credibility, and costs two years of board time that could have gone somewhere else.
I learned this most clearly from the wrong side of the table. At HSBC in Hong Kong I sat inside innovation budgets, assessing partners and vendors, some of them venture-backed companies pitching us as a channel. The analysis I ran was not “how big could this be.” It was “what is the probability this partner cannot deliver, and what happens to me if they cannot.” Different vocabulary, identical instinct. A large institution and a venture fund both discount the upside case heavily and price the downside case with precision, because the upside is a story and the downside is a liability.
So when you walk out of a strong first meeting and assume the hard part is done, you have it exactly inverted. You have passed the easy filter. You are now entering the part of the process that actually decides the outcome, and you are entering it with whatever paperwork, structure and candour you happened to accumulate while you were busy building the product.
The deals I have watched die almost never died on the numbers
In twenty years of building across Europe and Asia, as a founder raising, as an executive assessing counterparties, and now as a CEO in climate tech where the capital cycles are long and the diligence is technical, I can count on one hand the rounds I have seen collapse because the metrics were disappointing. Disappointing metrics kill you before the first meeting. They do not get you to week six.
What kills you at week six is something structural that nobody can explain in a single sentence.
Revenue concentration is the most common. Not the concentration itself, which is normal at an early stage and which any reasonable investor will accept, but the absence of an explanation for it. “Our largest account is 61% of revenue, here is the renewal mechanism, here is the pipeline that takes it to 30% within four quarters, here is what happens to our runway if they leave” is a conversation. The same fact discovered by an analyst in a spreadsheet during week five is a red flag with no owner.
Technical architecture is the second. In consumer hardware, building across China with the Jean-Michel Jarre venture, we grew one product into a line of eight, and the question that came back from every serious commercial partner was not about design or demand. It was about whether the thing could be manufactured at volume without the quality curve collapsing. Software has an equivalent. An investor sends a technical adviser, the adviser asks how the system behaves at ten times current load, and your CTO either has an answer with numbers in it or the round quietly changes shape. The answer does not need to be impressive. It needs to exist.
The third killer is the co-founder arrangement that was never papered. Early teams operate on trust, which is correct and efficient, right up to the moment somebody outside the trust circle has to underwrite it. A co-founder who left eighteen months ago with a verbal agreement about their equity, a founding engineer whose IP assignment was never signed, a vesting schedule that everybody discussed and nobody executed. These are all fixable. None of them are fixable in week six without cost, delay and a visible loss of confidence.
And then there is structure. I have built and operated entities across Europe, Hong Kong and mainland China, including leading digital transformation for a private healthcare chain expanding across the mainland, and I will say this plainly: cross-border structures that made sense for operational or tax reasons at the time become a diligence problem the moment an investor’s lawyer tries to draw the ownership chain on one page and cannot. If your IP sits in one jurisdiction, your revenue is booked in another, and the operating team is employed by a third, you need the single-page diagram and the rationale ready before anyone asks. Otherwise you are asking a partner to defend a structure they do not understand to an investment committee that will like it even less.
Why silence past four weeks is almost always a no
Founders read silence as deliberation. It is usually the opposite.
Here is the mechanic. An investor who completes a diligence workstream and finds nothing unexpected has every incentive to move quickly, because good companies are competitive and speed is one of the few things a fund can offer that costs nothing. Interest accelerates. The questions get more specific, more commercial, more forward-looking. You start hearing about portfolio introductions and board composition. Momentum is the signal.
An investor who has found something does not usually tell you. They park it. They go quiet while they decide whether it is worth the internal effort of getting comfortable, and in most cases the answer is no, because there are eleven other companies in the pipeline without that problem. The polite email arrives six weeks later saying it is too early for them.
My working rule is four weeks. If a process has gone quiet for four weeks after an engaged first meeting and a data room handover, something was found. The useful move at that point is not another follow-up asking if they have had a chance to review. It is a direct email naming the most likely problem yourself: “I suspect the customer concentration is the sticking point. Here is what we are doing about it.” Sometimes that reopens the conversation. More often it does not, but it tells you what to fix before the next one, which is worth more than the first investor was.
Build the file six months before you need it
The fix is unglamorous and it is almost entirely about timing.
Six months before you intend to raise, assemble the diligence file as if a term sheet were already signed and you were in confirmatory diligence. Cap table with every instrument reconciled, including the handshake deals. Signed IP assignments from everyone who has touched the codebase or the product design. Customer contracts, with the renewal terms and the ones operating on expired or extended paper flagged in a cover note. Revenue by customer by month for twenty-four months, with your own commentary on concentration. A one-page entity diagram with the reason each entity exists. A technical architecture document written for a competent outsider, not for your own engineers. Employment contracts for the whole team. Any regulatory dependency, in writing, with its current status.
Doing this six months out costs you perhaps three weeks of work and some uncomfortable conversations. Doing it in week six of a live process costs you the round, because every gap becomes a negotiating point and every delay reads as disorganisation.
The second half of the fix is behavioural, and it is harder. Volunteer the ugly facts in meeting two. Not meeting one, where you are still earning the right to a conversation, but meeting two, when there is enough interest to absorb bad news. Name the concentration. Name the technical debt and the timeline to clear it. Name the co-founder who left and how it was settled. Name the jurisdiction that complicates things.
Founders resist this because it feels like handing over ammunition. In practice it does the opposite. A problem you disclose is a problem you have priced, contained and attached a plan to, and it moves the conversation to whether the plan is credible. The identical problem discovered in week six is not about the problem at all. It is about you, and about what else might be sitting in the file that nobody has found yet. That question has no good answer, and it is the one that ends rounds.
Traction gets you in the room. What you have already written down, and what you are willing to say before you are asked, decides whether you leave with money.
I write from twenty years of building businesses between Europe and Asia. If your company is facing this, start a conversation.