Why I Built UserSay
I had dashboards for everything — signups, churn, conversion. I had all the numbers. None of them told me what to build next.
I build AI tools. Multiple products, each solving a different problem. Together they serve over two million users across 140 countries.
I have analytics. I know how many people sign up each day, how many convert, how many leave. I can tell you that 96% of free users never come back. I can tell you that the #1 cancel reason is "missing features."
But I can't tell you which features. I can't tell you who these people are, what they were trying to do, or what went through their mind when they decided to leave. I have all the numbers. I have none of the stories.
The coffee chat I lost
When I started, I had 50 users. I knew a lot of them. I'd DM someone who just signed up: "Hey, what are you building?" They'd tell me everything. I'd learn that one user was a teacher making lesson plans, another was a consultant preparing client decks, another was a student who just needed one quick output for a thesis.
Every product decision I made came from those conversations.
Then I grew. Thousands of users, then tens of thousands. I got dashboards. I got NPS surveys. I got cancel reason dropdowns. I got everything a "data-driven" founder is supposed to have.
And I understood my users less than ever.
The dashboard says conversion is 2.3%. Is that a pricing problem? A product problem? An onboarding problem? The data doesn't say. It just sits there, looking back at me.
The question nobody answers
I realized there's a gap that no tool fills.
Analytics tell you what users do. Click here, drop off there, convert at this rate.
Customer support tells you what's broken. Bug reports, feature requests, angry emails.
NPS surveys tell you a number. Seven out of ten. What does that even mean?
None of them answer the question that actually matters: Why?
Why did this user sign up? Why did they stop coming back? Why won't they pay? Why did they choose us over the other ten options? What are they actually trying to accomplish in their life, and where does our product fit?
These are the questions that change roadmaps. And the only way to answer them is to sit down with a user and have a real conversation.
But I run multiple products, solo. I don't have a PM. I don't have a user researcher. I can't schedule 50 coffee chats a month.
So I built one that could.
What UserSay actually is
UserSay is simple. It lives inside your product. When a user hits a meaningful moment — they've used up their free credits, they've been paying for a week, they're about to cancel — it starts a conversation.
Not a survey. Not five stars. Not a dropdown. A real conversation.
"Hey — what were you working on when you ran out of credits?"
The user talks for three minutes. They tell you who they are, what they do, why they're here, what's missing. They feel heard. You get a story.
The next morning, you don't open a dashboard. You read this:
"I'm a freelance consultant. I use your tool to make diagrams for client presentations — about 3 a week. The free plan is perfect for my volume, but I won't pay $12/month because I don't use it every day. If you had a pay-per-use option, I'd buy it immediately."
One paragraph. And now you know more about your pricing strategy than a month of A/B testing could tell you.
What I believe
I believe the best product decisions come from stories, not scores.
I believe every SaaS product talks to its users constantly — onboarding emails, push notifications, feature announcements, surveys — but almost none of them listen.
I believe the "talk to your users" advice is the most important and least followed principle in building products. Not because founders don't want to — because it doesn't scale.
I believe AI changes that. For the first time, you can have a real, thoughtful, adaptive conversation with every user who matters — automatically, at the right moment, in their language — for less than a penny per conversation.
I believe your product deserves a listening layer. Not just an analytics layer, not just a support layer — a layer that hears what your users have to say, and turns their stories into your next product decision.
That's what I'm building.
— Sarah
Solo founder & builder of UserSay