The Trusted Network Is the First Instrument, Not a Shortcut

I The claim

Most founders treat their network as a shortcut: call the people who already like you, and once a few of them have said yes, write “early traction” on a slide. We use the same people for a different job. They are the first measuring instrument a founder owns, and like any instrument they have a known error, which in their case is that they read high.

We use them anyway and we use them first, because a reading that is wrong in a direction we know about is still worth more than no reading.

II Why warm comes first

Kyle Poyar and Maja Voje surveyed 300 software go-to-market leaders in September 2026 for Growth Unhinged, and 62% of them called network and introductions a core channel. SEO got 51% and LinkedIn posting by founders and employees got 50%. AI search, the subject of every conference talk this year, got 39%.

The usual reading is that warm introductions convert well, which is true and is the less useful thing about them. What we want from a trusted contact is that they answer within a day, that they will say the headline confused them, and that they will get on a call and let us watch them read the page. A cold ad hands back a click rate two weeks and some money later, and the rate cannot say why.

While an offer is still a guess we would rather have a fast, candid answer from a few people than a large one from many. So the network gets the test landing page as soon as it is built, before any of it goes out cold and well before there is a product to show. We are trying to learn which claim people repeat back to us, which line they skip and where they stop reading.

III A ceiling, labelled as one

Kromatic’s guide to landing page smoke tests puts the problem flatly: traffic from your own list or followers “measures how much your existing fans like you, not whether a market exists.” It says to treat any warm-audience result as “a generous upper bound.” We agree, and we suspect most teams nod at this and put the warm number in the deck anyway.

A warm conversion comes with help a stranger never gets. The person knows your face and may owe you a favour, and they read past the weak sentence because they assume you meant something sensible, an assumption no cold visitor will make for you.

This makes a warm failure the most valuable result the network can return. If people who like us will not take the offer it is dead, and we found out for the price of a few coffees. A warm pass tells us much less, only that the offer can work under the best conditions it will ever see.

Our rule is clerical and we hold to it. Warm and cold results never share a column, and the pass/fail number, which is agreed in writing on the first day, applies to the cold result only. The warm number gets written down with the word “ceiling” next to it. We expect the cold number to come in lower, and finding out how much lower is what the ad budget is for.

IV What to ask, and what not to

“What do you think?” is the worst question available and the one everybody asks, and friends are too polite to answer it. “Would you buy this?” is nearly as bad, since a yes costs them nothing. We ask instead about what they understood and what they have already done.

• Read the page for twenty seconds, close it, and tell us what we sell.
• When did you last have this problem, and what did you do about it?
• What did that cost you?
• What would you stop using if you used this?
• Who has this problem worse than you, and will you introduce us?

The last question is the one we score. Praise costs nothing to give, whereas an introduction spends some of the giver’s reputation, so people only make one when they believe the thing is worth somebody else’s time. We count the introductions offered and ignore the praise.

We also don’t ask for general feedback. “Any thoughts?” gets us font opinions from someone who was never going to buy.

V Watching without leading

Then we watch someone use the page. We give them a task and no tour (“You need to do X. Go.”) and after that the technique is to shut up, which is harder than it sounds when the thing on screen is yours.

When they stall we count to ten before saying anything, and when they ask what a button does we ask what they expected it to do. What they did and what they said go in separate columns, and where the two disagree we believe what they did. An agent can transcribe the session and tag where each person stopped, so the record does not depend on anyone’s memory of it. Our memory of our own session is always flattering.

We don’t explain anything during the session. Every explanation we give out loud is one the page was supposed to give, and the cold visitor will not have us sitting beside them. If we catch ourselves saying “oh, that just means,” we’ve found the next edit.

VI The founder’s account

Between the trusted few and the cold many there is a middle ring of people who know of you. The founder’s own account reaches them, while the company page mostly reaches employees and a recruiter.

The evidence here is thin. RevBoss, an agency that sells founder-led marketing, published benchmarks from its own client panel in which founder accounts earned 2.1 times the median impressions of company pages. That compares 1,122 founder posts with 80 company-page posts from three pages, and RevBoss itself calls it directional. The Growth Unhinged survey gives firmer ground on a neighbouring point: founder-led sales was the motion for 44% of respondents under $1M ARR and for 26% in the $1M to $10M band. It gives no figure above that.

Our reason for preferring the founder’s account does not depend on reach in any case. People reply to a person, and a reply from someone who has the problem, written in their own words, tells us something that a like on a company page (usually a colleague being supportive) does not. So the founder posts the claim in the offer’s exact words and we read who answers and what they call the problem. That audience is lukewarm, and its results get a column of their own.

Then we buy strangers. By that point the page, the wording and the objections have all been through people who told us the truth, and what none of them could tell us is whether any of it survives without us in the room, which is the part that costs money.