Market research in an Innovator Founder Visa submission is one of the easiest sections to fail at. Applicants often submit copy-pasted market size tables from industry reports, call it research, and wonder why the assessment marked the viability pillar low.
Both endorsing bodies that have spoken publicly are explicit that desk research is not enough. Richard Harrison at Innovator International:
Not just Google, let's get real people who want your solution and are prepared to pay for it.
Source: innovatorinternational.com.
If you're serious about starting a business, why would you go to a country you don't know with an idea that you haven't tested? If you're telling me that you're doing that, I don't believe that you're a genuine entrepreneur with a genuine opportunity.
Scott Horton at Envestors reinforces the same principle through the innovation pillar's four probing questions, which specifically ask whether the applicant has researched the UK market rather than the global market. Research that doesn't anchor to the UK doesn't count.
What "UK-anchored" means
UK-anchored research has three characteristics that generic market research lacks:
- UK-resident or UK-based respondents. Interviewees who live, work, or buy in the UK. A survey of 500 people globally with 40 UK respondents is less useful than a survey of 60 UK respondents.
- UK-specific findings. The research reveals something about UK behaviour, UK regulation, UK market structure, or UK pricing. Not just global industry trends with the word "UK" inserted.
- UK-actionable conclusions. The findings inform UK-specific product decisions — which segment, which channel, which price point in the UK market.
A good heuristic: if your market research section could be lifted entirely and applied to the US, German, or Australian markets without changing anything, it isn't UK-anchored.
Desk research: what it's for and what it isn't
Desk research (reading industry reports, government statistics, competitor websites, published market analyses) is legitimate as background context. It sizes the market, names the players, establishes the trajectory. For an assessor, it shows the applicant has done basic due diligence.
What desk research cannot do:
- Validate demand for a specific product
- Establish willingness to pay at a specific price
- Reveal customer objections and purchase barriers
- Test the founder's specific value proposition
Assessors read desk research quickly. They're looking for red flags (numbers that don't add up, sources that aren't credible) but they're not treating it as evidence of product-market fit. That evidence has to come from primary research.
Primary research: the forms that count
Primary research means you, personally, engaged with real UK customers. Several forms count, in rough order of credibility.
Structured interviews
A minimum of 20-30 interviews with UK buyers or users of the category you're entering. Each interview should have:
- A named respondent (anonymised in the public document if needed, but named in your records)
- A documented role and organisation
- A consistent interview protocol (same questions across respondents)
- Written notes or recorded transcripts
Presenting the methodology ("we interviewed 24 UK-based procurement directors at companies between 100 and 1,000 employees, using a semi-structured protocol") signals research discipline. Presenting raw unstructured conversation notes does not.
Pilots and trials
Running a pilot of your product with actual UK customers is a form of primary research. Each pilot generates usage data, feedback, and demonstrated willingness to engage. See letters of intent vs paying customers for how pilots rank as evidence.
Surveys with methodology
Quantitative surveys are evidence if they are methodologically credible: sample size appropriate for the claims, UK respondent filtering, validated question design. Unvalidated SurveyMonkey results with 50 respondents from your LinkedIn network aren't research; they're vibes.
Industry events and field observation
Attendance at UK trade events, observational research in UK retail environments, field interviews with UK operators. These count if documented — with dates, locations, observations, and notes.
Know exactly where your application stands.
Get your free AI assessment in 90 seconds.
Get your assessmentHow to structure the research section
A strong market research section in a submission has five components, in this order:
1. UK market size and trajectory
Desk research. Sourced, specific, UK-focused. 3-5 paragraphs with numbers and citations.
2. Primary research methodology
Named and explained. "We conducted 34 structured interviews with UK-based [target segment] between [dates]. The interview protocol focused on [three areas]. Interviews ranged 30-60 minutes." Signals research discipline.
3. Findings from primary research
The substantive output. What did you learn that changed your view of the market? Specific quotes (anonymised if needed), specific findings, specific segmentation insights. This section is the most valuable because it reveals the applicant's genuine engagement with UK customers.
4. Willingness-to-pay evidence
One of the most important — and most often-missing — findings. Did you ask UK customers what they would pay? What did they say? Did any give specific numbers? This connects directly to the financial forecast.
5. Implications for the business
How the research shaped the product, the target segment, the pricing, the go-to-market plan. Research without implications is research that wasn't actually used. Assessors want to see the research driving decisions.
Common failure patterns
Numbers without sources. "The UK market is £4.2bn." Where does that come from? Cite it. If the source is "estimated based on our analysis," say so explicitly.
Industry reports as primary research. A paid-for Gartner or IDC report is desk research, not primary research. Treating it as primary is a common error.
"I've worked in this industry for 10 years." Experience is viability evidence, not research evidence. Your prior experience is why your research is credible — it isn't a substitute for doing research.
Generic pain-point language. "UK customers struggle with [generic problem]." Without specific evidence that you asked specific UK customers about the specific pain, this is asserted research, not conducted research.
Sample sizes too small to infer. "We spoke to five people." Five conversations can generate hypotheses but can't validate demand for a product. At least 20-30 interviews are needed before you can make demand claims.
The AI-written research trap
A specific failure mode that has emerged recently: market research sections written or heavily drafted by AI tools. Richard Harrison:
We said straight away it's an AI plan, it's not genuine from the founder.
AI can summarise existing reports. AI cannot go and interview 30 UK procurement directors. A market research section that reads as competent generic analysis but lacks specific named respondents, specific quotes, and specific dates is a red flag for AI generation or ghost-writing.
See ghost-written ideas and buzzword traps for the broader pattern.
Cross-body note
Both Envestors and Innovator International reject desk research as primary validation evidence. Envestors' four innovation questions test whether the applicant has specifically researched the UK market. Innovator International's viability questions test for credible demand evidence. See the three endorsing bodies compared.
External context
The Home Office Innovator Founder Visa guidance requires endorsers to assess whether the applicant's plan is viable — and viability rests on credible evidence of market demand. Market research standards in the UK typically follow the Market Research Society code of conduct for methodology credibility.
Key takeaways
- Desk research is background, not evidence. Primary research with UK respondents is what counts.
- Minimum 20-30 structured interviews is the practical floor for demand validation.
- UK-anchored research means UK respondents, UK-specific findings, UK-actionable conclusions.
- Include methodology explicitly, not just conclusions — signals research discipline.
- AI-generated research sections are detectable and rejected. The research has to be the founder's actual work.
- market-research
- uk-market
- primary-research
- validation
- desk-research
