You can vibe-code the product. You can't vibe-code the evidence
Fine Science tells you whether your AI-built product meets a real user need, and what to fix first if it doesn't. Before you build more, launch or raise.
Dr Nick Fine - PhD HCI - Founder of UX Psychology - 20 years in digital consumer behaviour
"Three onboarding failures we'd been shipping past for six months. Every finding was specific, traceable and immediately actionable. We knew exactly what to fix and in what order."
Series A founder - SaaS
"We were two weeks from launch. A critical positioning failure caught before it shipped. The no-go verdict was uncomfortable. It was correct."
Product lead - B2B platform
"No agency padding. No 40-slide deck. A clear diagnosis, a priority stack and a verdict we could defend to the board. Exactly what we needed."
Digital Director - Enterprise
Is this you?
Builders, founders, agencies - anyone with a product that isn't working
Product teams prototyping with AI
Your PMs can build a working prototype in an afternoon. That makes it easy to fund the wrong thing by Friday. Prototype Review tells you which user needs the prototype actually serves, which are assumed, and where it lands on the Fine Science scale: Structurally Sound, Needs Surgery or Rebuild Required.
You shipped fast. Before you spend on launch or a raise, find out whether it meets a real need and what is stopping people using it. Start with a Rapid Diagnosis, or a Teardown if you are pre-revenue.
The vibe-coding goldrush is real.The failures are too
// Captured insight
"We spent a fortune with the agency and it isn't selling." Fine Science hears this every week. The product works. The problem is it was never built around what users actually need.
User needs still have to map to product function. Most vibe-coded products fail because nobody traced what the user actually needs to what the product actually does. Fine Science closes that gap before you waste another quarter scaling the wrong thing.
// THE COST OF SPEED
A weekend build is cheap. Long-term failure is not. AI collapsed the cost of shipping, not the cost of misreading demand. You can vibe-code fast. You cannot vibe-code the evidence that would have told you the product was wrong.
Code auditors check whether it works. Fine Science checks whether anyone needs it.
// WHY NOT JUST USE AN AI TOOL?
AI Audit Tools
Scans heuristics
Scores interfaces
Follows checklists
Fine Science
Feels user confidence rising or collapsing in real usage
Spots the gap between the founder message and what the product actually communicates
Matches cross-industry patterns and names the failure mode quickly
What we do
Ways to work with Fine Science, priced up front
Every service has a defined output and a stated price, so you know exactly what you're buying before you commit.
// With your permission, teardowns may be published (anonymised on request)
20-minute recorded video walk-through of your live product
One-page verdict on the Fine Science scale: what works, what doesn't, what to do next
Async, no call needed
// Service 02
Rapid Diagnosis
Fast expert diagnosis before you launch, spend on growth or raise
A cold walk-through of your product by a senior practitioner with 20 years of pattern recognition. No report. No deck. A structured 60-minute verbal debrief covering the three to five most significant problems observed, with an honest assessment of whether they are surface-level fixes or structural failures. Hard-scoped, not a taster. One walk-through, one debrief, one page of findings. It exists for founders who need a credible outside eye before committing to a full engagement.
The £1,500 is credited in full against Laboratory Analysis if commissioned within 30 days. This is not a taster. It is a risk-free first step into the engagement.
Written summary of three to five headline findings
One-page findings document - all headline observations, the honest recommendation and next steps, delivered same day as the debrief
Honest recommendation on the Fine Science scale: fix it yourself (Structurally Sound), Laboratory Analysis needed (Needs Surgery) or rebuild (Rebuild Required)
// Service 03
Prototype Review
For product teams deciding whether an AI-built prototype deserves a funded build
£4,500
5-7 working days
+£3,000
Option: Fine Science runs the five user sessions and reports back
// The gate between "we have a prototype" and "we fund the build"
Need-Feature-Benefit map of the prototype: which user needs it serves, which are evidenced and which are assumed
Verdict on the Fine Science scale: Build (Structurally Sound), Test first (Needs Surgery) or Stop (Rebuild Required)
Five-user test plan targeting the riskiest assumptions
60-minute debrief with the team
// Service 04
Laboratory Analysis
A verdict on what's failing in your live product, why, and what to fix first. Works for agency builds and AI builds alike
A structured expert evaluation that maps every feature to the need it serves and the benefit it claims to deliver. Every finding is grounded in observable behaviour, not assumption. Covers onboarding integrity, positioning accuracy, journey failure points, need-feature alignment and retention signals. Works for live products and pre-launch builds. Concludes with a clear verdict and a sprint-ready plan your team can act on immediately.
The £1,500 Rapid Diagnosis fee is credited in full against Laboratory Analysis if commissioned within 30 days.
// Every finding is traceable to a specific observable behaviour - not a heuristic checklist
Laboratory Report (10-20 pages)
Full NFB mapping - every feature against the need it serves
Priority Stack - ordered by severity, sprint-ready
Verdict on the Fine Science scale: Structurally Sound, Needs Surgery or Rebuild Required
Go/no-go statement where pre-launch
60-minute debrief call
45 days async follow-up
// Service 05
Fine Science On Call
Rigorous product thinking, available when you need it
A standing engagement for teams who want priority access to senior diagnostic input without commissioning a full project each time. One monthly review call, up to three product reviews per month. Each review is a structured evaluation of one product, feature area or problem statement, delivered as a written findings note. A monthly findings memo and priority escalation when something critical surfaces. Quarterly summary report included.
Up to 4 Prototype Reviews per month (expert review; user sessions priced separately)
Monthly 60-minute team session
Priority escalation channel
Quarterly summary report
// Training
Learn the method Live, with Dr Nick Fine
Fine Science teaches the same diagnostic method it uses on client work. Live cohorts for individuals, in-house sessions for teams.
// Live cohort course
Build the Right Thing with AI
You can prototype in an afternoon. Find out, in four weeks, whether it deserves a quarter
A four-week live course for product managers, designers and founders who prototype with AI. Take one of your own prototypes through a needs check and a five-user test, and leave with a verdict on the Fine Science scale: Structurally Sound, Needs Surgery or Rebuild Required.
Week 1: Needs before features. Map every feature to the user need it serves
Week 2: The prototype as a question. Find your riskiest assumptions and plan a five-user test
Week 3: Watching, not asking. Run the sessions and read behaviour, not opinion
Week 4: The verdict. Make the call and present the case to your team
Founding cohort
£695Standard price £995
4 x 90-minute live sessions - weekly office hours - 20 seats maximum - recordings and templates included
The same method, run in-house for your product team, using your own prototypes
The Fine Science method for product managers who prototype with AI. How to check a prototype against real user needs before it goes to engineering. Your actual prototypes are the subject.
Basics
The diagnostic framework and how to apply it
£2,500
Full day, remote or on-site, max 12 participants
Integrated
The framework embedded into your team's existing workflow
// Your team leaves able to run a version of this themselves
Case study
A diagnosis in practice
One anonymised example of the kind of problem Fine Science is brought in to diagnose and what changed once the underlying issue was made visible.
Problem
The client had launched, traffic was arriving and sign-up was happening, but almost nobody reached the first moment of value. Internally, the debate had split between fixing the messaging, redesigning onboarding or rebuilding the feature set.
Finding
The failure was structural rather than cosmetic. New users were being asked for commitment before the product had established credibility, so the journey created doubt before it created momentum.
Outcome
The team paused launch spend, rewrote the first-run journey and relaunched two weeks later with measurably better activation.
How Fine Science works
Structured Not guesswork
Fine Science is expert opinion - but it's expert opinion applied through a repeatable diagnostic framework, not a vibe check. Every finding is traced to a specific observation or behavioural pattern. The methodology is transparent. The reasoning is visible. You can disagree with the verdict, but you can see exactly how it was reached.
// Dr Nick Fine
01
Intake - context without contamination
A brief is taken, precise questions are asked, then the product is worked through independently before any conclusions are formed. The client's view of what works is noted and set aside.
02
Independent immersion
The product is experienced as a new user would experience it. Every friction point, dead end and unmet expectation is documented without reference to the client brief.
03
Diagnostic analysis
Each feature is mapped against the need it serves and the benefit it claims to deliver. Root causes are identified - not just symptoms.
04
Finding synthesis and prioritisation
Findings are grouped by severity and type. Every recommendation is specific and actionable. "Improve onboarding" is not a recommendation. Changing a specific step is.
05
Verdict and delivery
A clear verdict is stated. The report is written in plain English - no jargon, no padding. Then a debrief call where the client can ask hard questions and get straight answers.
// How is this different from a UX audit?
A UX audit
Scores the interface against a checklist
Identifies heuristic violations
Produces a list of recommendations
Tells you what looks wrong
Fine Science diagnosis
Uses the product as a new user would
Identifies root causes, not surface symptoms
Delivers a verdict you can act on
Tells you why it is broken and what to do first
A heuristic checklist can tell you a button is hard to find. It cannot tell you that users are abandoning because the product asked for commitment before it established credibility. That is the difference.
The verdict
Fine Science gives verdicts, not suggestions
Every diagnostic ends in an unambiguous verdict on this one scale. Not a list of things to consider. A professional opinion you can act on.
Structurally Sound
The product has problems but they're fixable without fundamental rethinking. The priority stack tells you what to do and in what order.
Needs Surgery
One or more structural failures need to be addressed before surface-level fixes will have any effect. Specific surgery required before you scale.
Rebuild Required
Fundamental need-feature misalignment that can't be resolved incrementally. Rare - but when warranted, Fine Science will say so honestly.
Earlier work: enterprise UX and research
Coca‑Cola
UX strategy
Jaguar Land Rover
Research programme
Bentley
UX strategy
GSK
Research programme
Novartis
Research programme
EY
UX design
Virgin Media
Conversion rate optimisation
Camelot - The National Lottery
Conversion rate optimisation
SABMiller
UX design
Shell
Research programme
Home Office and Cabinet Office
Research programme
NHS Blood & Transplant
Research programme
// Selected from 20 years of agency, client-side and government engagements
About Nick Fine
20 years of pattern recognition One discipline
01
Pattern recognition across industries and contextsTwenty years of walking through products cold: • B2B and B2C platforms • Internal tools • Consumer apps • Regulated environments
02
Brutally honestThe market is oversupplied with people who tell clients what they want to hear. Fine Science isn't one of them.
03
Client-side and agency experience• Series A-C scale-ups • Enterprise platform teams • Regulated and compliance-heavy environments • Consultancy product work
Client-side taught consequences and long-term thinking. Agency taught speed and client variety. The diagnostic method draws on both.
04
eCommerce and conversion rate optimisation specialist• Designing and running multivariate tests • Analysing conversion funnels • Diagnosing drop-off points
The difference between testing randomly and testing with a hypothesis grounded in behavioural analysis. CRO isn't guesswork - it's structured experimentation with a diagnostic foundation.
05
Founder of UX PsychologyNick coined the term UX Psychology in 2018. Nobody in UX was treating Psychology as a discipline in its own right. He published Introduction to UX Psychology on YouTube in March 2018 - before the term existed anywhere. It has since entered global use.
He had been practising it since 2009. The academic foundation is a PhD in HCI from Brunel University and an MSc from University College London. Published research covers behavioural analysis and personalised interface design. The training matters because it's why every finding is structured, traceable and defensible - not just observed.
06
Speaker and published commentatorRated in the top 5% of speakers by audience feedback. Keynotes and workshops at international UX and product conferences across Europe. The Signal to Noise keynote - on separating genuine diagnostic insight from AI-generated noise - delivered at three international conferences in 2024.
11 podcast appearances including Brave UX, One Knight in Product and Human Tech - where UX Psychology was first named publicly in March 2018. Regular contributor to industry debate on the future of UX, the limits of AI tooling and the case for evidence-based practice.
A talk that shows how Fine Science thinks - recorded live
"Most product failures aren't technical. They're behavioural. The product works fine - it just doesn't work the way real people think."
Email Nick
The laboratory is open
Tell us what the product is and what's wrong with it. Fine Science will confirm which service fits and whether it's the right match for the problem.
BasedHertfordshire, UK - remote-first
TurnaroundMost engagements within 2 weeks of confirmation
Current capacityTaking a limited number of engagements each month
MSA and custom invoicing available for enterprise engagements. Data handling terms on request.
Dr Nick Fine - PhD HCI - Founder of UX Psychology - 20 years in digital consumer behaviour
