Radical Personas is a synthetic user research tool from Austria. It puts websites, ads, emails, and product ideas in front of AI-modeled synthetic personas — and returns Scores, reasoned feedback, and a PDF report in about 20 minutes. Every persona carries eight psychological layers — biography, Big Five (Schmitt 2007), Kahneman biases, value orientations, Hofstede’s cultural dimensions, and more — so you see what bothers Franz about your headline, and why. With a source for every layer.
The same prompt, sent to ChatGPT and to Radical Personas: “How would Franz Hinterberger, 48, owner of a metalworking business in Upper Austria, review this landing page?” Both responses below are real, translated from the German originals.
“First impression: too much marketing speak. ‘Synergies’ and ‘scaling’ in the first paragraph — that makes me suspicious. The contact details belong higher up, not hidden in the footer. Stock photos: interchangeable.”
A solid output. But: slightly different on every run, no reproducible Scores, no bias attribution, no conflict map against other personas — and no speech register that tells you Franz is from Upper Austria.
“First screen, and I see no price and no phone number. If they don’t want me to ask, then I won’t ask. Click, gone. The stock photos look like they came straight out of a PowerPoint template. If real people were standing there next to their own work, I’d have stayed.”
ChatGPT gives you a plausible answer. Radical Personas gives you Franz — with his Upper Austrian speech register, a reproducible Score per criterion, bias attribution, and a conflict map against the other personas in the same run. In the meeting, that’s the difference between “interesting input” and “a basis for a decision.”
A landing page, an Instagram Reel, a logo draft, a sales email. Eight predefined object types, one upload form. User journeys across multiple pages: up to 5 pages from Pro, up to 10 from Business, up to 20 from Enterprise. The wizard translates your file into the format our personas can work with.
Two to twenty Personas per Review — no fixed limit on the Enterprise plan (fair use). From our Persona Library, or with your own layer configurations. Every Persona comes with its eight layers built in — its biography, its biases, its Life-World, its language sensitivity. You don’t have to explain any of that. The Persona brings it along.
A scorecard per criterion, every Persona’s comments verbatim, an AI summary with conflict detection. Exported as a PDF, formatted for stakeholders. No dashboards you can’t show anyone. A document you attach to an email.
“The headline talks to me like a management consultant. My clients are small agencies — the word ‘scaling’ loses them instantly.”
Instead of a list of attributes, you get a human — with the eight layers that make him who he is. This is the building block every Radical Persona is made of.
Technical school (HTL) in Wels, metalworking, took over the business from his uncle. 22 employees, 4 of them apprentices. Married, two teenagers. Guild master at the Upper Austrian chamber of commerce. Local councilman in Thalheim.
Biographical grounding · curated in-house
Conscientiousness 88/100, openness 32/100. Values: a handshake that counts, hard work, family. Fears: skilled-labor shortage, bureaucracy, no successor. Motivations: handing over the business, training apprentices, regional impact.
Big Five · Schmitt et al., 2007 (n=17,837, 56 nations)
Status quo bias, high: “Why change what works?” Anchoring effect, medium: the first price named dominates the negotiation. In-group bias, high: trusts a recommendation from Maier at the guild more than three external consultants.
Kahneman & Tversky, 1979 · Prospect Theory
Baseline slightly positive (valence 0.55), calm (arousal 0.45). High stress tolerance. Negative trigger: new EU regulation for metal processing. Positive trigger: an apprentice passing the journeyman’s exam.
Russell Circumplex Affect Model
Hofstede: high uncertainty avoidance, high long-term orientation, moderate power distance. Life-World: The Stewards (Standards & Bearing). Value cluster: groundedness, local clubs and associations, the regional economy.
Hofstede, 2010 · our own Life-World classification (AT/DE)
Tech affinity 35/100. Competent with CAD and CNC controls in the shop; privately, the smartphone is for calls and WhatsApp only. Media diet: the regional paper in print, radio, commercial TV. No social media. Decisions: rational — with gut feeling when people are involved.
Behavioral patterns · curated in-house
Has been putting off digitalization for three years — “orders come first.” Underestimates what his own online presence does for winning new business. Trusts recommendations from his network blindly, questions online information too little.
Anti-pattern registry · curated in-house
Upper Austrian dialect with business partners, standard German with clients from outside the region. Low tolerance for anglicisms, except for technical terms. Austrian markers like “Jänner” and “heuer.” Standard-German greetings like “Guten Tag” feel foreign to him — he prefers the Austrian “Grüß Gott.”
Language sensitivity profile · curated in-house
Eight layers. One Persona. No shortcuts.
These eight layers don’t work in isolation. How they weight and contradict one another — that is the heart of Radical Personas.
You’ve seen what a Persona is made of. Now find out what it says about your landing page.
All of it included from €29/month (€19 introductory price until July 31, 2026).
The Big Five are validated across 56 nations, n=17,837 (Schmitt et al., 2007). Kahneman’s work is peer-reviewed. Hofstede’s dimensions have been measured empirically since the 1970s. Values research (Schwartz) is broadly peer-reviewed. All of it is available to you — in papers, in books, in APIs.
Our compositional logic decides how these layers interact in a specific review context. Which layers factor in differently for a logo than for a cold-outreach email. How conflicts between emotional state and cultural context are resolved by rules. How eight layers stay coherent at the same time without collapsing into averages.
This orchestration is the heart of Radical Personas. What you see is the output: a report with Scores, quotes, and a synthesis.
All of it included from €29/month.
A logo isn’t reviewed the way an email is. Each object type brings different layers to the foreground — not because we swap them out, but because the compositional logic reweights them.
Incl. landing pages: product pages, campaign destinations with a CTA, user journeys of up to five pages from Pro, up to ten from Business, up to twenty from Enterprise.
Dominant layers: Biography + Behavior + Language
LinkedIn, Meta, TikTok, YouTube. Image, video, copy.
Dominant layers: Biases + Emotional State
Newsletters, sales sequences, transactional emails.
Dominant layers: Language + Biases
Brand marks, wordmarks, combination marks.
Dominant layers: Cultural Context + Language
Instagram, LinkedIn, TikTok, Facebook, X — incl. preview mockups.
Dominant layers: Aesthetics (Behavior) + Emotional State
Product screenshots, app views, feature mockups.
Dominant layers: Behavior + Anti-Patterns
Proposals, quotes, and sales documents — as a PDF or as text.
One question, put to several Personas at once — a Sentiment Snapshot with the majority and the dissenting voices at a glance.
No tool onboarding, no setup day. Three concrete situations from an ordinary workday — showing where Radical Personas steps in between draft and send.
The letter is written. Before it goes out: the moment where “I understand” turns into “I’m looking for alternatives.”
The Persona reads it through the eyes of a customer who has been with you for three years — and marks the line where loyalty tips, and the phrasing that keeps it.
The ad is ready to go. You want to know: does it reach the skilled-trade families in the region?
A Persona with a regional biography, Life-World “The Preservers,” and low tech affinity reads along. It tells you whether the language feels too urban, or whether the list of benefits misses what actually matters.
It goes live next week. You want one last check — but not a usability test with real participants.
Three to five Personas from your target group review it in parallel. About 20 minutes later, you have Scores, criticism in each Persona’s own words, and a conflict map of where they contradict each other — as a PDF.
“Franz Hinterberger (48) scored the headline 3.0/10. His reasoning: ‘The word synergies makes me suspicious. When I’m looking for a metalworker to build me something, I don’t talk about synergies. I talk about a door, a measurement, and a delivery date.’”
Activated layers: 2, 3, 8 (Psychology, Biases, Language)
A scorecard per criterion. Every Persona’s comments verbatim. An AI summary with a conflict map between Personas. Formatted for stakeholders, not for dashboards.
Every Persona stays available. Ask for the reasoning behind a Score, have variants weighed, go into detail — in character, with memory of the entire Review. Example conversation, translated from German.
The scientific foundation (Big Five, Kahneman, values research, Hofstede) is peer-reviewed and fully documented on the science page. Go to the science page →
From €29/month · Free plan, no credit card · Cancel anytime See all plans →
A free plan to try it out. No credit card, no minimum term, no onboarding call. Your first report in about 20 minutes.
Most people start on their own. Larger teams talk to us — both work.
Free plan to try it out, no credit card. First results in about 20 minutes. Cancel anytime.
Custom personas for your industry, pilot projects, a compliance review (GDPR, Swiss FADP), integration into your existing workflows. 30 minutes with the founders — no sales team.
DE
EN