First Ten

The problem in detail

With too few users for any statistical signal, these founders have no trustworthy way to set a price or choose what to build next. They copy competitor pricing pages, treat the loudest feature requests as votes, and start then abandon formal methods like Van Westendorp or RICE because those assume sample sizes they'll never have. The result is months of undercharging, attracting low-value users, and shipping features that dilute the paid core instead of strengthening it.

Proposed product

A guided small-sample decision kit. The founder imports or types in their customer list and open feature requests. First Ten generates a short, scripted 15-minute conversation guide per customer (willingness-to-pay anchoring questions, the job they hired the product for, what they'd stop paying for), sends the booking links, and gives the founder a one-page capture form to fill in during each call. Once four or five conversations are logged, it produces two outputs: a recommended price band with the exact customer quotes that justify each end of it, and a ranked feature shortlist where each request is scored by whether the people asking are paying, retained, and in the core segment — not by how many times it was asked. Requests that only low-value or churned users want are explicitly flagged as 'do not build'. Re-run quarterly to track whether the price band is moving up.

Why now

The evidence shows no sharp recent change: small-sample pricing research is a long-standing problem that consultants already sell as a service and how-to guides already describe. What is new is only the packaging — DemandMaven-style live willingness-to-pay interviews and Savio-style lifecycle segmentation of feedback exist as published methods, but the tooling that operationalises them (for example Canny's ARR-weighted request scoring) is built for teams with revenue-data integrations, not founders with a spreadsheet of twelve customers.

Smallest useful version

A simple web app: paste in your customer list and your feature request backlog, get personalised interview scripts and capture forms, and after five logged calls receive a price recommendation with supporting quotes plus a build/don't-build verdict on each backlog item. Sold as a one-off £49 'pricing sprint' to founders recruited from indie hacker communities. Success means founders actually complete five calls and change their price or kill at least one planned feature within two weeks.

Potential ways to charge

• Free script generator — £0: interview script and capture forms for up to three customers, no analysis (free tier typical for comparable tools; earns the paid step by proving the scripts are usable)
• Pricing sprint — £49 one-off: full script set, capture forms for up to fifteen customers, price band with supporting quotes, and build/don't-build verdicts on the backlog (anchored to the £49 figure in the admin's MVP brief)
• Quarterly re-run — £15 per month: repeat sprints each quarter with price-band movement tracked over time and a comparison against the previous round (typical for comparable solo-founder tools; no price in the evidence)
• Done-with-you sprint — £400 one-off: the founder does the calls, a human reviews the transcripts and writes the price recommendation (positioned well below the consultant fees referenced in the evidence, whose actual figures are not recorded)

Routes to early customers

• Indie builder communities where solo founders post launches and revenue updates — approach founders who have just announced their first handful of paying customers
• Founders publicly asking 'how should I price this?' in SaaS and bootstrapper forums, offering to run the sprint with them personally
• Newsletters and guides in the pricing-advice space (the kind of how-to content the search evidence found dominating this topic) as guest posts or template giveaways
• Consultants who sell small-sample pricing engagements, as a cheap downstream referral for founders who cannot afford their fees

Main risks

• The riskiest assumption stands untested: founders may want an answer without talking to anyone, and five booked calls is real unglamorous work that most will not complete.
• The segment with the problem is the segment with the least budget — the evidence shows a paid consultant market exists, but for funded B2B SaaS, not 5–30-customer indie founders.
• No cross-class corroboration was found that this segment pays for tooling at all; they may read a free guide and do it in a Google Doc.
• The feature-prioritisation half is already commoditised: Canny weights requests by ARR and segment, Savio segments by lifecycle stage, so that differentiator is weak even if those tools suit larger teams.
• Both halves of the method are published best practice, not new insight, so the product is packaging — easy for a competitor or a well-shared template to replicate.

Questions to test first

• Sell the sprint before building it: offer twenty founders a £49 manual version delivered over email, and count how many pay and how many actually complete five calls.
• Measure call completion specifically — of those who pay, how many log five conversations within two weeks is the single number that decides whether this is a product or a wish.
• Ask ten founders who have run such calls whether they would have trusted and acted on a price band drawn from five conversations, or wanted more.
• Compare the output against a free template: give half the test group just the scripts in a doc and see whether the generated analysis changes the decision they make.
• Probe whether founders would pay again at quarter two, or whether the price question is answered once and forgotten.

Evidence and sources

The sources this assessment is based on. Evidence level: Limited, counted from these sources as described in how ideas are assessed. Findings are what a source shows; anything the research only inferred is marked as an inference.

Research findings

• Supportsconsulting firm page

  Shows: paid-service-substitute

  DemandMaven's pricing guide describes running Van Westendorp-style questions in live pricing interviews rather than surveys, noting that getting people to commit to a number in real time is revealing even when exact numbers are unreliable. This is observed evidence that the core First Ten workflow (scripted WTP interviews instead of statistically-powered surveys) is already an established paid consulting method — supporting the problem framing but showing the method itself is not novel, only its packaging as self-serve software for sub-30-customer founders.

• Contextconsulting directory/marketing page

  Shows: competitive-landscape

  A published roundup of SaaS pricing consultants founders hire, with one firm explicitly positioning its output as high-fidelity work with medium-high confidence intervals suited to early-stage companies lacking analyst teams. Observed: a paid human-services market exists for exactly this decision (small-sample pricing research). Supports willingness to pay for the outcome, but also means First Ten competes on price/convenience against consultants whose fees most 5–30-customer founders likely cannot afford — a market-potential risk (the segment with the problem is the segment with the least budget).

• SupportsGitHub issue

  Shows: issue-tracker-artifact

  A public GitHub issue tasks someone with drafting an interview script to validate pain points and willingness to pay among early-stage SaaS founders. Observed artefact of someone treating 'write the WTP interview script' as a discrete unit of work, which is the exact deliverable First Ten automates. Weak as demand evidence: it is a single low-traffic repo and appears to be auto-generated task output, not a customer complaint or paid engagement.

• Weakens the casevendor product page

  Shows: competitor-capability

  Canny markets weighting requests by impact rather than volume alone, showing ARR tied to a request, the segment it came from, and whether it blocks a deal. This directly overlaps First Ten's differentiator (score requests by whether askers are paying/retained/core rather than by vote count), so the feature-prioritisation half of the idea is already commoditised by incumbents — though those tools are built for teams with revenue-data integrations, not founders with a spreadsheet of 12 customers.

• Weakens the casevendor blog

  Shows: competitor-capability

  Savio's prioritisation method recommends segmenting feedback by customer lifecycle stage — trial users, paying customers, churned customers — when deciding what to build. Observed confirmation that 'who asked' weighting is accepted best practice; First Ten's 'do not build' flag for churned/low-value askers is a packaging of an existing published method rather than a new insight.

Where the problem was reported

Public posts in which people described this problem, grouped into one problem before research began.

• A 20-Minute Weekly Metrics Review When You Have Under 100 Users · dev.to · 4 October 2026
• How I Decide Which Feature Requests to Ship (and Which to Politely Refuse) · dev.to · 1 October 2026
• How I Price a SaaS When I Have Almost No Data · dev.to · 29 September 2026

Looked for, not found

• The requested cross-class corroboration was not found: no product review, support complaint, search-volume data, or job/freelance posting surfaced showing founders paying for small-sample pricing-and-prioritisation work specifically at the 5–30 customer stage. Search returned mainly how-to content and consultant marketing aimed at funded B2B SaaS. Whether the sub-30-customer indie segment actually pays for tooling (versus reading a free guide and doing it in a doc) is therefore unverified; demand at that stage remains a hypothesis.