First-Time Visitor
📱 mobileModerate tech literacy, explores product-understanding-first before pricing.
“I'll figure out what this does before I care what it costs.”
For SaaS teams about to spend money on traffic
Soultab sends five psychographically distinct AI personas through your live website — reading, clicking, getting impatient, hunting for pricing — and returns the top friction findings with evidence. In minutes. Before the ad spend.
Early access — the founder runs every scan personally right now, and you'll get the report by email. No signup, no card, no bots crawling without permission.
Moderate tech literacy, explores product-understanding-first before pricing.
“I'll figure out what this does before I care what it costs.”
High conscientiousness, runs a trust-verification checklist before judgment.
“Show me who you are before I show you my email.”
Time-poor, outcome-first, impatient. Evaluates your CTA almost immediately.
“You have one screen. Go.”
Technical depth, heads straight for docs and specifics.
“Marketing adjectives are lorem ipsum to me. Where's the how?”
Low tech literacy, purchase-decision focus, pricing-first navigation.
“If I can't find the price, the price is 'too much.'”
They are not real people, and we never pretend they are. They're a repeatable panel — the same five archetypes, every run, so you can change your page and measure what changed.
A live, public SaaS site. Yours, presumably.
Each persona gets a real browser session and walks your actual site — navigating, reading, reacting in character. No screenshots-only shortcuts; they click.
Every step, every reaction, every “wait, what does this actually do?” is captured with the page it happened on.
A Soultab Score across five dimensions — comprehension, intent, trust, discoverability, friction — a verdict band, and the top findings, each with the persona's own words and a recommendation framed as a hypothesis to test.
Findings are hypotheses, not facts — the fast first pass before real user research, not a replacement for it. We label confidence and publish methodology, because a truth-telling tool that overclaims would be a very short-lived joke.
Illustrative example — not a customer result.
“I scrolled the whole page and opened the menu. I never found what this costs, so I assumed enterprise-sales and left.”
Hypothesis worth testing: A visible pricing link in primary nav recovers price-first evaluators. Test with real visitors.
Two of five personas, same wall, independently. That's the kind of convergence that's worth your attention — and we tell you when it's only one persona's opinion, too.
They're inside the same bubble you are.
You'll learn, but tuition is billed to your ad account. And analytics knows where they left, never why.
Excellent, in three weeks, for four figures.
They look at a picture of your page. They never click anything. In a 2023 test, Baymard measured GPT-4 screenshot-only UX audits at ~20% accuracy with ~80% false positives (source).
Five personas, real browsers, your live site, receipts, minutes.
Before this page shipped, we fed it to Soultab. The Skeptical Evaluator went looking for reasons not to trust us; the Busy Professional gave the hero one screen. What they flagged, we fixed — and what we didn't fix, we're honest about.
See this page's own Soultab report →You just shipped a new landing page. Or a repositioning. Or your first paid campaign starts Monday. You have a quiet suspicion that nobody outside the building actually understands what the product does — and every person you could ask is either on your payroll or being polite.
Soultab is calibrated for PLG and self-serve SaaS landing pages — deliberately narrow, so the panel's instincts (pricing-first behavior, trial expectations, docs-hunting) match how SaaS actually gets evaluated. Not e-commerce, not content sites, not yet.
Give us the URL. The founder will personally run the panel and send your report — early access, free while we calibrate, in exchange for one thing: you tell us which findings were news to you and which were noise. That feedback is how the instrument gets sharper.
We scan with permission, respect robots.txt, and never touch non-public pages. Your report is yours — we don't publish it, benchmark you against competitors, or cold-email your team.