Signal Over Metrics (Pre-PMF Decision Rule)
You cannot optimize your way to product-market fit — hunt for pull, not conversion lifts.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 92%
Butler's deliberately contrarian position: at early stage, your numbers are too small to tell you anything, so metrics-driven decision-making is a trap. Instead you look for qualitative signal — visceral pull from users — and trust intuition. The concrete test Figma used before leaving stealth was whether prospects physically grabbed the laptop out of Dylan Field's hands during a demo.
Origin
Claire Butler's stated 'controversial point of view' from running Figma's pre-launch go-to-market in 2015, when the company had only a handful of teams using the product full time.
Core principles
- 01You cannot optimize your way to product-market fit — a 5% email lift tells you nothing about whether the thing works.
- 02Signal beats metrics when n is small: a couple of people who love it outranks any conversion improvement.
- 03Product-market fit is felt as pull — someone trying to take the product from your hands.
- 04The same intuition-trust that is the most stressful part of early stage is the thing you need most to succeed.
How to run it
- 1
Set the milestone ladder, not a metrics dashboard
Sequence the goals: get one company to use your product; get them to keep using it; get a second; get someone to pay. Each is a discrete, observable milestone.
Pro tip Step one — literally one company using it full time — is genuinely hard and deserves full attention.
Watch out Do not substitute funnel metrics for these milestones; at low volume they are noise.
- 2
Demo relentlessly and watch the reaction, not the outcome
Go to companies and demo the product in person. Ignore whether they buy; watch the emotional reaction to the tool itself.
Pro tip Ask for feedback, never sell — the feedback conversation is what actually converts them.
Watch out Plenty of people will simply not care. That is fine; you are looking for the outliers who do.
- 3
Apply the pull test
Look for visceral pull — users taking the laptop out of your hands to play with it themselves. That reaction, not a metric, is your go signal.
Pro tip Combine with a hard floor: a small number of teams using it full time, plus excited non-users who want to try.
- 4
Get scrappy about who you show it to
Use any channel to reach early users — DM influencers for feedback, slide into DMs of a friend's ex-boyfriend, talk to people in a Lyft. Anything to get more demos and more signal.
- 5
Protect the intuition-driven model as you scale
As the company grows, leaders will demand ROI proof for the bottom-up motion. You need an executive who believes in signal and can hold the line while metrics catch up.
Pro tip Butler names this as the single most important prerequisite for the whole GTM model working.
Watch out It is easy for a flood of new, metrics-trained hires to break the thing that was working.
In the wild
Figma's core differentiator — multiplayer — was still a year from shipping, and even Notion's early team told Butler they couldn't launch without it. But designers in demos were physically taking the laptop out of Dylan's hands to play, a few teams were using it full-time, and the team was demoralized by three years of silence.
→ Figma launched at the end of 2015 without multiplayer, on the strength of qualitative pull signal alone, and used the following year of user feedback to build toward it.
Butler noticed a Microsoft employee had signed up for Figma, recognised them as her friend's ex-boyfriend, and DM'd them to get the first Microsoft conversation and feedback.
→ That contact led into the Xamarin team becoming patient zero, and Microsoft eventually becoming a self-generated enterprise deal.
Common mistakes
Chasing conversion optimization pre-PMF
A 5% lift on an email or landing page tells you nothing fundamental about whether the product works. Optimization is a post-PMF discipline.
Waiting for the perfect product to launch
Figma launched without its headline differentiator. Staying in stealth longer would have cost a year of feedback and destroyed team momentum.
Is it for you?
Best for
Founders and early GTM hires at pre-product-market-fit startups where user counts are too small for statistical inference
Not ideal for
Scaled products with real volume, where metrics and optimization are exactly the right tool
From the transcript
“you can't optimize your way to product Market fit”
“I think metrics are really hard and signal is actually way more important like can you get a couple people who'd love it”
“they would like take the laptop out of Dylan's hands when he would start showing it because they wanted to play with it”
“the most important thing there is that you have an executive and a leader who believes in”
From the episode
An inside look at Figma’s unique GTM motion
Claire Butler (first GTM hire)