There is a story people tell when they stop building companies and start funding them: that having built things makes you better at backing them. It is half true, and the half that is false costs founders money.
I have done both now. Three companies of my own. Four organizations as an executive, at Honor, Lyft, Grammarly and Pitch. Then a year as EIR at Peak in Amsterdam, 2023 into 2024, running product reviews across roughly thirty companies, most of them at seed. And Noord, from 2024: an AI Design Lab that occasionally invests in startups, and operates inside the companies we work with. So let me be specific about which parts of that history carry over and which parts actively mislead, because the general claim helps nobody.
Three things the history lets you see
Whether a metric was built to flatter. Anyone who has run a review cycle has watched someone define the number they will later be judged by. You develop an ear for it. A founder counts active users and a session counts as a user. A retention figure arrives with no window attached. I ask about the definition rather than the value, every time, and the answer is almost never dishonest. It is usually a number the house has always counted that way, and nobody has had reason to test it. What I learn is whether this team argues with itself.
Whether a team can ship or is describing a plan to ship. These sound similar in a meeting and are not remotely the same. One question separates them. What went out in the last two weeks. People who build answer with something specific and usually show you, often something small and unglamorous. People who have outsourced their own product judgment answer with a roadmap. I was on the wrong side of that question for years. I could describe what we were building. I could not have opened it and used it.
What a hire will actually cost. Not salary. A founder says they will hire a head of design in the spring, and I can see the six months that follow. The search. The ramp. Everybody's expectations reorganized around a person who does not yet have context. Sometimes the right advice is that the role is a year premature and the founder should learn the tools well enough to write the spec themselves. I have given that advice three times this year to founders asking whether to add a second designer, and said no every time, for different reasons.
Three ways it makes you worse
Now the part operators do not put in their fund's positioning.
You look for companies that rhyme with your own. My history is a marketplace, a mobility company, a writing tool and a presentation tool, mostly consumer, mostly with a strong design surface. I read those companies easily and I am therefore more confident about them than my track record justifies. That failure is not really about backing too many of them. What goes wrong is elsewhere. In the categories where my instinct still hands me an opinion, I am less useful and slower to admit I do not know.
You over-index on craft. An operator who ran design will discount a team whose product looks assembled from off-the-shelf components. Sometimes that is the right read and the interface is telling you what the founder thinks the hard part is. Sometimes the interface is ugly because they are shipping four times a week to customers who are paying, and the polish is the correct thing to defer. I have been wrong this way more than any other. It is the most expensive mistake available to someone with my background, because it looks like taste.
You want to fix rather than fund. This one is almost physical. You sit with a team and you can see the four things you would do in their first month. The urge to take the wheel gets strong. Those have very different answers and only one of them is your job.
It turns the investment question into a different one: whether you would enjoy running this company. Only one of those is your job.
What an investor can actually do
A standard list of what a seed investor adds: marketing, recruiting, mentorship. Having been on the receiving end of all three, I think the middle one carries most of the weight and the other two are often theater.
Marketing help at seed usually means an introduction to a journalist and a logo on a website (both real, neither decisive). Mentorship is real and is also what a founder's peers provide better, more honestly, and without a position in the company.
Recruiting is different, because money is rarely what's holding a seed company back. What holds them back is that four people have to become nine without any of the five being a mistake. A founder doing that for the first time has no network deep enough. And no way to calibrate a senior candidate they have never managed the equivalent of.
Here is the concrete version, which I ran for a year and would run again. Keep a live list of strong operators who are between things. Not a database of everyone. Thirty or forty people you have actually worked with or reviewed. A note on what each is good at, what environment they need, and what they want next. That third field is the one everybody skips. Match it against the hiring plans of companies you are close to. It decays in about six months, which is why almost nobody maintains one.
Two things make it work. First, your best former colleagues are the supply. The strongest people I ever managed left to start companies or take founding seats. Knowing where they are is an asset. Second is closing. An investor can make the final call to a candidate deciding between two offers, and say something a founder cannot say about their own company without it sounding like a pitch. That call is the single most valuable thing I have done for a portfolio company, and it takes twenty minutes.
I also ran small dinners for engineering, product and design leaders in a couple of European cities, invite-only, eight to twelve people, one guest worth showing up for. Its stated purpose was community. What I actually got out of them was knowing who was good and who was moving, six months before either was public.
The honest summary
I do think operating gives you a reliable read on execution and an unreliable read on markets. It makes you good at the question of whether these people can build the thing, and no better than anyone else at whether the thing is worth building. The investors I have learned most from know which half they are, and bring in someone for the other half instead of pretending to have it.
What I would tell a founder: use your operator investor for the team, the hiring and the product decisions. Do not weight their market opinion above your own. They have seen four companies from the inside at close range. You are looking at yours every day.
I've founded three companies and led design and product at Honor, Lyft, Grammarly and Pitch. I spent 2023 and 2024 as an entrepreneur in residence at Peak in Amsterdam. Now I run Noord, an AI Design Lab that occasionally invests in startups. No portfolio company or individual above is identifiable, deliberately.
© 2026 Renato Valdés-Olmos