These four historical examples show how early evidence can support, weaken or leave unproven a problem thesis. Each case separates the evidence a founder observed from the decision they made, the counter-evidence that remained, and the questions the case could not answer.
Stage 1 uses public evidence to test whether a problem is real, visible and urgent. It does not establish buyer budget, willingness to pay, offer fit, acquisition economics, product-market fit or commercial viability.
Evidence is not the same thing as a founder decision. A founder can act on limited evidence, but that decision does not turn the evidence into proof that a business, buyer, offer, or commercial path has been validated.
Counter-evidence stays visible because mixed results are part of the record. A decision to continue means a founder chose to keep learning or building. It does not mean the business was validated. Equally, an ordinary later business failure is not evidence that an earlier validation test said “stop.” A stop or narrow example needs a documented decision to abandon, reframe, or narrow the current direction based on the evidence available at the time.
Real sourced example - illustrative, directional continue case
Problem thesis and context: Before Buffer had a functioning product, Joel Gascoigne was exploring whether a standalone tool for queuing Twitter posts would be useful. The initial idea came from his own experience of wanting to avoid sending several tweets at once. It was a founder-originated solution hypothesis, not evidence of a defined B2B buyer or a broadly established urgent problem.[1]
Source and source date: Joel Gascoigne, Buffer founder, Idea to Paying Customers in 7 Weeks: How We Did It, 16 February 2011.[1] A contemporaneous Hacker News launch discussion from 30 November 2010 provides visible counter-evidence about the proposition.[2]
What was actually tested: Gascoigne published a non-functional two-page landing flow, shared it on Twitter, and invited reactions. The flow collected email notifications and feedback. He then inserted a pricing page before the email form to see whether visitors would select a plan before leaving an email address.[1]
Evidence: The founder reported email submissions, feedback through email and Twitter, continued completion after the pricing step, and a small number of paid-plan selections. He chose to build a functioning version after those directional signals.[1]
Counter-evidence / missing evidence: The early response was mixed. In the launch discussion, one commenter questioned the value of paying monthly for automated posting, and Gascoigne described the product as niche with polarising feedback.[2] The sources do not report a visitor denominator, submission rate, buyer segment, urgency measure, pre-set threshold, or pre-build payment.[1] [2]
Decision actually made: The founder continued. After the initial landing-page response, he added the pricing-page step. After the pricing-page response, he chose to build a functioning product.[1]
Stage 1 reading: This is a directional, mixed response to a proposed solution. It illustrates the difference between an early action that shows interest and a later action that tests a stronger signal. It does not establish that the underlying problem was broadly urgent or that a defined buyer had been identified.
Not established here: This does not establish buyer budget, willingness to pay, offer fit, acquisition economics, PMF or commercial viability.
Next test: The documented next test after the first landing page was a pricing-page step that recorded plan selections before the notification form.[1]
Monetary test cost: not reported.
Source note: A paid-plan selection was not payment. The first actual payment followed the launch of a functioning product, so it is later buyer, payment, product, and revenue evidence rather than proof from the pre-build test.[1]
Real sourced example - illustrative, narrow transaction continue case
Problem thesis and context: In 1999, Zappos founder Nick Swinmurn was investigating whether people would buy shoes online despite being unable to try them on first. His account includes a personal shopping trigger and a reported mail-order statistic, but neither establishes a broadly evidenced public problem.[3] [4]
Source and source date: Nick Swinmurn, founder interview published by Business Insider, 28 November 2011, describing the 1999 test.[3] A first-person Fortune profile published on 5 September 2012 corroborates the basic store-listing sequence.[4]
What was actually tested: Swinmurn photographed shoes in local stores, listed them on a website, and agreed to purchase any item that sold from the retailer at full price.[3]
Evidence: He later reported making “a couple of sales.” The corroborating profile describes the same outcome as “a few orders.” This is a small reported online transaction signal.[3] [4]
Counter-evidence / missing evidence: The record contains investor scepticism about whether people would buy shoes without trying them on. It also shows later supply, selection, cost, and funding constraints.[3] The sources do not report visitor volume, buyer identity, conversion, return rate, margin, repeat behaviour, a formal go threshold, or a defined public-problem evidence base.[3] [4]
Decision actually made: The company continued pursuing online shoe retail. The sources record later funding, hiring, and supplier development. They do not state that the couple of sales met a predefined go rule or alone caused those later decisions.[3]
Stage 1 reading: This example shows a small observed online-transaction signal. It does not establish whether a public problem was real, visible, or urgent, and it does not establish demand scale or commercial economics.
Not established here: This does not establish buyer budget, willingness to pay, offer fit, acquisition economics, PMF or commercial viability.
Next test: not reported.
Monetary test cost: not reported.
Source note: The founder’s commitment to buy each sold item at full price is part of the test setup. The sources do not give a total monetary cost, so the case must not be called free, low-cost, or rapid.[3]
Real sourced example - illustrative, documented product narrow/pivot
Problem thesis and context: Burbn was a broad location and social product that included check-ins, photos, and other features. The founders were deciding which part of that broad product direction to pursue. The cited record does not establish a contemporaneous Burbn revenue model or a defined buyer problem.[5] [6]
Source and source date: Kevin Systrom, interview transcript published 30 April 2019.[5] Contemporaneous TechCrunch reporting from 20 September and 8 November 2010 corroborates the product scope and the deliberate pivot.[6] [7]
What was actually tested: The closest documented test was qualitative observation of Burbn in use, combined with the founders’ review of adjacent product categories. Systrom later recalled that users only “kind of liked” check-ins, while they loved sharing what they were doing through photos.[5]
Evidence: The qualitative feature-use observation and category review informed the founders’ decision to remove Burbn’s other features and focus on social photo sharing. Contemporaneous reporting describes the team starting over and choosing to do one thing well.[5] [7]
Counter-evidence / missing evidence: Check-ins were not wholly rejected. Users reportedly “kind of liked” them, so the source supports a relative-priority decision rather than a conclusion that check-ins had no demand.[5] The sources do not provide a sample, measurement method, observation period, target segment, threshold, payment record, or revenue evidence.[5] [6] [7]
Decision actually made: The founders made a documented narrow/pivot. They stopped pursuing Burbn’s broad multi-feature configuration and focused on social photo sharing.[5] [7]
Stage 1 reading: This is a founder-reported qualitative feature-use observation that informed a product narrow. It is not a quantified public-problem validation experiment.
Not established here: This does not establish buyer budget, willingness to pay, offer fit, acquisition economics, PMF or commercial viability.
Next test: not reported.
Monetary test cost: not reported.
Source note: Later funding, launch, growth, filters, acquisition, and any planned monetisation belong to separate product, distribution, and commercial questions. They do not prove the narrow decision was a formal Stage 1 pass.[5] [6] [7]
Real sourced example - illustrative, attributed early solution-direction stop/change
Problem thesis and context: Shizune founder Pavel Gvay wrote that the team’s initial idea was a “network map” tool. The source does not define the tool’s user segment, job to be done, offer, or revenue model.[8]
Source and source date: Pavel Gvay, The mistakes I made on a journey to $90k worth of preorders, Indie Hackers, 2 November 2021.[8] An associated author comment published 3 November 2021 contains later commercial metrics, but those metrics do not establish the earlier network-map test.[9]
What was actually tested: Gvay wrote that he conducted more than 20 interviews before reaching a conclusion about the network-map idea. The post does not disclose the interview sample, questions, responses, or a defined test protocol.[8]
Evidence: The direct evidence is the founder’s dated account: after more than 20 interviews, he concluded that nobody needed the initial network-map concept.[8]
Counter-evidence / missing evidence: The page does not contain the underlying interview evidence. It does not identify who was interviewed, what they said, the two pains the founder later mentioned, or a selected next direction.[8] Later reported pilots, preorders, revenue, and economics concern an unspecified evolved or manually delivered offer. They do not validate the original network-map concept.[8] [9]
Decision actually made: This is an attributed early solution-direction stop/change. The founder describes changing away from the initial concept during a broader customer-conversation process. The source does not document a complete stop-and-narrow loop.[8]
Stage 1 reading: This is an attributed example of a founder stopping an early solution direction after customer conversations. The absence of the underlying interview record means it remains illustrative rather than independently verifiable public-problem evidence.
Not established here: This does not establish buyer budget, willingness to pay, offer fit, acquisition economics, PMF or commercial viability.
Next test: not reported.
Monetary test cost: not reported.
Source note: The source does not name the specific pain chosen next, a narrowed ideal customer profile, a causal link to later preorders or revenue, or a payment model for the initial concept.[8] [9]
| Case | Evidence type | What changed | What remained unproven |
|---|---|---|---|
| Buffer | Landing-page response, email feedback, and paid-plan selections before email capture | The founder added a pricing-page step, then chose to build a functioning product | Problem urgency, a defined buyer, payment before build, price acceptance, and commercial viability |
| Zappos | A small founder-reported online-order signal from store-sourced listings | The company continued pursuing online shoe retail | Public-problem evidence, demand scale, conversion, supply viability, margins, repeat purchase, and commercial viability |
| Burbn to Instagram | Qualitative feature-use observation and a category review | The founders narrowed from Burbn’s broad configuration to social photo sharing | Quantified user evidence, a defined buyer, payment, revenue, and commercial viability |
| Shizune | A founder-attributed conclusion after more than 20 interviews | The founder describes changing away from the initial network-map concept | The underlying interview evidence, a selected next pain or segment, payment, and commercial viability |
A genuine stop or narrow decision is documented in the source as a decision about the direction being tested. It is not a retrospective label applied because a company later struggled, changed, or succeeded.
Burbn to Instagram is the strongest documented narrow/pivot example in this set. The founders removed the other features from a broad product and focused on social photo sharing after qualitative feature-use observations and a category review.[5] [7]
Shizune is a documented early solution-direction stop/change. The founder reports deciding that the initial network-map concept was not needed after more than 20 interviews. The source does not establish a named next pain, a narrowed ideal customer profile, or a complete stop-and-narrow sequence.[8]
Buffer and Zappos are continue cases. Each source records continued pursuit, not a documented stop decision. A later company failure is not automatically evidence that an earlier validation test said stop.
These are historical public examples. GTM Right interprets them narrowly from the cited sources and keeps the source date next to every case. They are not GTM Right customer results and they do not constitute testimonials, customer evidence, or promises about outcomes.
The 2,715-record Failure Archive may provide aggregate context about startup-failure patterns. It is not the source of these case narratives and must not be used to reconstruct a founder’s hypothesis, evidence, or decision. Where the cited record does not provide an item, this page says not reported rather than estimating or filling the gap.
Use Stage 1 to assess public evidence that a problem is real, visible and urgent. It does not assess buyer budget, willingness to pay, offer fit, acquisition economics, PMF or commercial viability. Run the free Stage 1 check. Read the Commercial Validation Guide.