The Selge Manifesto
Stop guessing. Start asking.
A short letter to anyone who has ever made a website decision without knowing why.
March 2026
Dear growth-minded person,
We want to say something that most software companies won't: the problem isn't that you lack data. You probably have more data than you can read. Analytics dashboards. Heatmaps. Session recordings. Funnel reports. You have everything except the one thing that tells you why.
You know your conversion rate dropped 12% last month. You do not know what caused it. You know visitors spend 45 seconds on your pricing page. You do not know whether that means they're reading carefully or they're confused and looking for an exit.
Before we tell you what we built, we want to sit with you in three familiar moments. We suspect you'll recognise them.
Scene I
A Tuesday afternoon
Your conversion rate is flat. Has been for two months. The team runs a retro. Someone says the pricing page “just feels off.” Everyone nods. You spend three weeks redesigning it based on collective instinct, a competitor you've been watching, and a blog post from 2021. It launches. The conversion rate is still flat.
Three weeks of work. You still don't know why visitors weren't converting. You just have a different-looking page they're not converting on.
This is the most common story in SaaS web teams. Not because the people involved lack skill. But because opinions fill the vacuum that data leaves. When you don't know what visitors are thinking, everyone's guess sounds equally valid - and the loudest voice usually wins.
Scene II
Six weeks later
You hire an agency to run user research. They recruit 8 participants, run moderated sessions over Zoom, synthesise findings into a 40-slide deck. Six weeks pass. The deck arrives. The headline finding: “Users find the value proposition on the pricing page unclear.”
$18,000. Six weeks. The answer was always one question away from anyone who landed on that page.
Traditional user research has its place. For deep product questions, for exploring unknown unknowns, for understanding behaviour in context - it is genuinely valuable. But for “why aren't visitors converting on this page?” you do not need six weeks and a Zoom panel. You need to ask the people who are already there.
Scene III
End of quarter
Someone proposes an A/B test. You check your traffic numbers. 800 visitors a month to the pricing page. To reach significance on a 5% conversion lift, you'd need to run the test for 14 months. The roadmap can't wait 14 months. So you ship the variant anyway, call it at 6 weeks with no significance, and argue about the data for two more weeks.
A/B testing is a precision tool. Most SaaS companies don't have the traffic to use it correctly. They run tests anyway, reach no conclusions, and wonder why nothing improves.
A different approach
What if you just asked them?
Not in a research lab. Not in a spreadsheet. On your website, in the moment they're forming their opinion - when the answer is still honest and the context is still real.
What changes
From guessing to knowing
The conversation happens while it's still relevant.
A visitor who didn't sign up is the most valuable person you can talk to. In an exit survey, they'll tell you exactly why - because they just decided. Three days later they can't remember your site. Two weeks later they're using your competitor. The window is now.
On-site surveys get 3-5x higher response rates than email follow-upsYou get answers, not data. There's a difference.
Analytics tells you that 68% of visitors leave the pricing page after 40 seconds. An on-site survey tells you: 'I couldn't figure out which plan I needed.' One requires interpretation. The other requires action. The latter is faster and cheaper.
50 responses is enough to find a recurring pattern. You don't need 5,000.The question matters as much as the answer.
Most survey tools hand you a blank form. You figure out what to ask. But a bad question produces useless data - and you won't know it's useless until you've waited three weeks for responses. Every Selge template was built from real conversion work: questions that have been tested, refined, and proven to surface information that changes what you build next.
The right question at the right moment gets directional answers from 20 responsesFrom question to change in a week, not a quarter.
Ask on Monday. Read responses Thursday. Rewrite the copy Friday. Ship it next week. That's the loop that compounds. Not hypothesis, design, build, test, wait for significance, argue about p-values. Just: ask, listen, act.
Average time to first actionable insight with Selge: 72 hoursWhat we believe
The principles behind Selge
Asking beats guessing
Every time. In every situation. A direct answer from a real visitor beats any amount of indirect data interpretation.
Context is everything
A survey on the page where the decision happens is worth more than any follow-up email, panel, or focus group. Real context. Real moment. Real answer.
The question is the product
Most tools give you a widget. We give you the expertise to know what to ask. That distinction is why the templates exist.
Small is enough
You don't need 10,000 responses to see a pattern. You need 30 people to say 'I didn't understand the pricing' before it becomes a thing you fix.
Focused tools outlast bloated ones
Selge does one thing. On-site micro-surveys. We won't add heatmaps, session recordings, or an AI-powered everything suite. We'll just get very good at this.
Speed compounds
Teams that close the question-to-action loop in days improve faster than those running quarterly research cycles. The gap widens over time.
Selge is Estonian. It means clear. That's the whole idea: getting clear on what your visitors actually think, what stopped them from signing up, and what one change would make them stay.
The tool exists because after 15 years of A/B tests, heatmaps, and open-text responses, the most valuable data always came from the same place: asking a direct question to the person who was right there.
If those three scenes felt familiar - the gut-feel redesign, the expensive agency deck, the underpowered A/B test - then this is where they end.
With respect for your time,
Margus Veeber
Founder, Selge
No credit card. No sales call. Just the tool.