How to Write a Smoke-Test Landing Page in 60 Minutes

A founder has spent three weeks talking to potential customers. The interviews point in one direction — the problem is real, the audience is reachable, the current workaround is bad enough that people would switch. The temptation at that point is to build. The discipline is to do one more test that costs an afternoon and produces a number, not an opinion.

The test is a smoke-test landing page: a single page that names the product, names the audience, names the alternative, names the gain, and asks the visitor for one of three actions. The page is shipped in sixty minutes. Traffic is paid or borrowed. The conversion rate over the next seven to fourteen days is the signal. The signal is read against a benchmark the founder has named in advance, not invented after the fact.

This article is the playbook for writing that page. It covers the structure, the headline, the body, the seven things to leave off, the build order that gets you to a live URL in an hour, and how to read the conversion data at small sample sizes. The conceptual home — what a smoke test is and how it differs from a fake-door test — is in Fake Door Test: How to Validate Demand Without Writing Code. The broader landing-page discipline (the channel mix, the conversion-rate thresholds, the multi-page sequence) is in Landing Page Validation.

Quick answer

A smoke-test landing page is a single-page website that describes a product that does not yet exist, asks the visitor for a real action (an email signup, a reservation deposit, a pre-order, or a waitlist join), and produces a conversion rate that the founder reads as evidence. The page is built and shipped in sixty minutes. The traffic is small and borrowed. The signal is the conversion rate at one specific action, compared to a benchmark the founder named before running the test.

The page has exactly five structural elements, in this order: a headline that names the buyer and the gain, a subhead that names the alternative, three to five signal lines (what the product does, in plain language), a single call to action, and a footer with the founder's name and an unsubscribe link. The page leaves off seven things: testimonials, team bios, "as seen in" press logos, feature lists, pricing for buy-now boxes, video, and chat widgets. Each of those seven items either requires evidence that does not exist yet or pulls the visitor toward an action other than the one being measured.

The build order for the sixty-minute version: write the headline and subhead first (twenty minutes), write the signal lines second (fifteen minutes), wire the form to a real email-collecting service (ten minutes), buy a domain and deploy to Vercel / Netlify / GitHub Pages (ten minutes), and buy $50 to $100 of traffic to test the page (fifteen minutes). The total elapsed time is sixty minutes if the founder already has a domain and an account on at least one deployment platform; it is ninety minutes if both need to be set up.

Key takeaways

• The smoke test is a real page on a real domain, not a Figma mockup. A mockup produces no signal. A real page on a real URL, with real traffic and a real form posting to a real service, produces a conversion number. The conversion number is the artifact.
• Five structural elements, in order: headline, subhead, signal lines, call to action, footer. Everything else is noise that pulls the visitor off the one action being measured.
• The headline names the buyer and the gain in plain language. Not "the smartest way to manage your workflow" but "A renewal reminder system for independent insurance agencies that recovers three days a month."
• The subhead names the alternative. "Unlike Salesforce, which assumes inside-sales workflows, our CRM ships with the renewal-reminder workflow on day one." Naming the alternative is what makes the page specific enough to defend in a sales call.
• The seven things to leave off: testimonials (no customers yet), team bios (not the test), press logos (none exist), feature lists (the signal lines are the features), pricing for buy-now (the test is the conversion, not the purchase), video (too slow to ship in an hour), chat widgets (they pull the visitor off the form).
• The build order matters. Headline and subhead first, signal lines second, form wiring third, deploy fourth, traffic fifth. Reordering produces a page that does not have the discipline of a real test.
• Read the conversion rate against a benchmark named in advance. A 4% waitlist signup rate at $5 of paid traffic per click is not the same as a 4% rate at $0.50 per click, and neither is the same as a 4% rate on traffic from a Reddit post. The benchmark is what makes the number mean something.

Why this matters

The most common mistake at this stage is to skip the smoke test and go straight to building. The founder has done the customer interviews, heard the same complaint five times, and assumes the evidence is enough. The interviews are evidence that the problem is real. They are not evidence that the named solution will convert. The smoke test converts the interview signal into a different signal — a conversion rate — that the founder can compare against a benchmark.

The cost of skipping the smoke test is not just a wrong build. It is the three to nine months the founder would have spent building the wrong thing, the morale damage of the failed launch, and the opportunity cost of the idea that should have come next. The cost of running the smoke test is one afternoon of work and $50 to $100 of paid traffic. The asymmetry is large enough that the smoke test is the rational default for any founder who has not yet validated at the conversion-rate level.

The smoke test is also the cheapest way to test positioning. A founder who runs the same smoke test with two different headlines — same subhead, same signal lines, same call to action — gets a comparison of two positioning wedges at small N. The headline that converts better is the wedge the founder should defend in sales calls. The smoke test is the discipline that produces that comparison in a week, not in a quarter of failed launches.

Definitions

• Smoke test — A pre-launch test of a product hypothesis by shipping a real page that asks for a real action, before the product itself exists. The page is the test; the conversion rate is the signal.
• Fake-door test — A variant of the smoke test in which the call to action leads to a "coming soon" or "thanks for your interest" page rather than to a real form. Useful for products where collecting an email is impossible or premature. The smoke test is the more disciplined version; the fake-door test is the fallback when the smoke test cannot run. The conceptual comparison is in Fake Door Test: How to Validate Demand Without Writing Code.
• Conversion rate — The fraction of unique visitors who complete the action being measured. For a waitlist signup, it is the number of email addresses collected divided by the number of unique visitors. The conversion rate is the artifact, not the count.
• Benchmark — A conversion rate the founder names in advance as the threshold at which the test is positive, neutral, or negative. The benchmark is named before the page is shipped; naming it after the data arrives biases the read.
• Signal lines — The three to five short sentences on the page that describe what the product does, in the buyer's outcome language. Signal lines are not features; they are outcomes, each ending in a measurable gain.
• Paid traffic — Traffic the founder buys from an ad platform (Google Ads, Reddit Ads, Meta Ads, LinkedIn Ads, X Ads) at a fixed cost per click. The cost of the traffic is part of the test; a $5 cost-per-click audience is not the same as a $0.50 audience at the same conversion rate.

Framework: the five-element structure

The smoke-test landing page has exactly five structural elements, in this order. Each element exists to do one job; the order exists so the visitor's attention flows toward the call to action.

Element 1 — the headline

The headline names the buyer and the gain in plain language. The buyer is a specific role or segment. The gain is a specific, measurable outcome in the buyer's life. The headline does not name the product, does not name the alternative, does not name the technology, and does not name the founder.

Three patterns that work, with a worked example for each.

Pattern 1 — buyer + gain + timeframe. "A renewal reminder system for independent insurance agencies that recovers three days a month."

Pattern 2 — buyer + pain + cure. "For operations directors at US manufacturers who produce monthly OEE reports by hand, a reporting tool that produces the report in five minutes."

Pattern 3 — buyer + alternative + new gain. "Unlike Salesforce, a CRM for independent insurance agencies that ships the renewal-reminder workflow on day one."

The discipline across all three patterns is the same: the headline is specific enough that a buyer can read it and say "yes, that is me" or "no, that is not me." A headline that says "the smartest way to manage your workflow" produces both reactions for everyone; it produces no signal.

Element 2 — the subhead

The subhead names the alternative. The alternative is the incumbent, the spreadsheet, the manual process, or the status quo that the buyer is currently using. The subhead also names the specific reason the alternative is not enough — the failure mode the smoke test is testing.

A worked example.

Headline: A renewal reminder system for independent insurance agencies that recovers three days a month.

Subhead: Independent insurance agencies with 10 to 50 agents currently spend three days a month on renewal reminder calls. Salesforce assumes inside-sales workflows; spreadsheets assume someone will remember. Our CRM ships the renewal-reminder workflow out of the box, integrated with the agency's existing IVR.

The subhead is the longest block on the page. It carries the burden of specificity the headline cannot carry. A page with a specific headline and a generic subhead produces the same conversion rate as a page with a generic headline.

Element 3 — the signal lines

The signal lines are three to five short sentences that describe what the product does. Each signal line ends in a measurable gain the buyer will experience, not in a feature the founder built.

A worked example for the same product.

1. Renewal reminder calls ship on day one — no configuration, no CSV import, no setup call.
2. The reminder integrates with the agency's existing IVR, so the calls go out through the same number the agency has used for ten years.
3. The dashboard shows which agents have called which clients, so the agency owner can audit compliance without running a separate report.
4. The system sends a daily digest to the agency owner, so renewals that need attention surface without anyone having to remember.
5. The first 100 agencies get a 50% discount for the first year, in exchange for feedback.

Each line ends in a buyer outcome. None of them describes the technology. The lines are the page's claim that the product solves the named problem in the named way. The conversion rate is the buyer's verdict on the claim.

Element 4 — the call to action

The call to action is the form. One form, one field, one button. The field is usually email; it can be email plus role, or email plus company size, when the segmentation matters. The button text names the action, not the product. "Reserve a spot," "Join the waitlist," "Pre-order the annual plan," and "Get the early-access link" are the four patterns that work. "Sign up," "Submit," and "Learn more" do not work.

The form posts to a real service that collects the email and either sends a confirmation or shows a thank-you page. The thank-you page is also the place where the founder can ask a second question — "what problem were you hoping this would solve?" — that produces qualitative evidence at the same time as the quantitative conversion rate.

Element 5 — the footer

The footer carries the founder's name, an unsubscribe link, a privacy link, and a contact email. Nothing else. The footer is not the place for social-media icons, a newsletter signup, a link to the founder's Twitter, or a link to the founder's blog. Each of those items pulls the visitor away from the action being measured.

The seven things to leave off

A smoke-test landing page is shorter than a marketing landing page. The discipline of leaving things off is what makes the page measure one thing cleanly. The seven items below are the most common things founders add to a smoke-test page that make the test less useful.

Leave off 1 — testimonials. There are no customers yet. A testimonial from a friend, an advisor, or a beta-tester is not a testimonial in the buyer's eyes; it is a piece of evidence that the founder does not have enough customers to fill the slot. Leave the slot empty.

Leave off 2 — team bios. The visitor is not buying the team. The visitor is buying the claim that the product will solve the named problem. A team bio pulls attention away from the claim.

Leave off 3 — "as seen in" press logos. There is no press coverage yet. The slot is empty by definition. A page that fills the slot with logos of outlets that have not covered the product is a different signal — the visitor reads it as dishonest, and the conversion rate drops.

Leave off 4 — feature lists. The signal lines are the features. A separate feature list duplicates them in shorter, jargon-heavy language. The duplication makes the page longer without making the claim clearer.

Leave off 5 — pricing for buy-now boxes. If the call to action is a waitlist signup, there is no price on the page. If the call to action is a pre-order, the price is on the form, not in a separate pricing section. A pricing section for a product that does not exist yet is premature; it asks the visitor a question the visitor is not ready to answer.

Leave off 6 — video. A sixty-second explainer video takes longer than sixty minutes to ship. A placeholder video produces a click on the play button, which is a different action than the form submission. The visitor's attention that the play button captures is attention the form does not capture. Leave video for the post-launch page.

Leave off 7 — chat widgets. Chat widgets pull the visitor into a conversation before the visitor has decided whether to convert. The conversation produces no measurable artifact; the form submission does. A chat widget on a smoke-test page is the equivalent of asking the visitor "would you like to talk to someone about this?" before the visitor has answered the question the page is asking.

How to build the page in sixty minutes

The build order is the discipline that gets the page shipped in an hour. Reordering produces a page that does not have the discipline of a real test.

Minute 0 to 10 — open the doc, name the five elements

Open a Google Doc or a Notion page. Write five lines: "Headline," "Subhead," "Signal line 1," "Signal line 2," "Signal line 3," "Call to action," "Footer." The five elements, in order. The discipline of the structure starts with naming the structure.

Minute 10 to 30 — write the headline and subhead

Write three candidate headlines. Pick the one that names the buyer and the gain most specifically. Write three candidate subheads. Pick the one that names the alternative most specifically. Twenty minutes is the budget; if the headline is not settled in twenty minutes, the problem is upstream — the founder has not done enough customer interviews to know what the headline should be.

Minute 30 to 45 — write the three signal lines

Write five candidate signal lines. Pick the three that end in measurable buyer outcomes. The signal lines are the hardest part of the page to write; the budget of fifteen minutes is intentional. The signal lines get better with revision, but the first three to five lines are good enough for a smoke test. The next revision can come after the first wave of conversion data.

Minute 45 to 55 — wire the form

Pick a form service. ConvertKit, Beehiiv, Loops, Tally, and Typeform are the common choices; any of them wires an email-capture form to a landing page in under ten minutes. Create the form, copy the embed code, and add it to the page. Test the form by submitting it once. The thank-you page is the page the founder will look at every morning for the next seven to fourteen days.

Minute 55 to 60 — deploy

Deploy the page to Vercel, Netlify, GitHub Pages, or Cloudflare Pages. A static HTML page with the form embed is the fastest to deploy. Buy a domain if the founder does not have one that fits the product name; the cost is $10 to $15 for a year. Ship the page.

If the founder already has a domain and a deployment account, the build order above is sixty minutes. If both need to be set up, the elapsed time is ninety minutes; the founder should plan for that.

After sixty minutes — buy the traffic

Set up a paid campaign with $50 to $100 of total budget. The campaign targets the named buyer in the named geography. The campaign runs for seven to fourteen days. The cost-per-click varies by channel — Google Ads at $2 to $10 per click, Reddit Ads at $0.30 to $1.50 per click, Meta Ads at $0.50 to $3 per click, LinkedIn Ads at $5 to $15 per click. The channel the founder picks should match where the buyer actually is. The conversion rate at small N (typically 100 to 1,000 unique visitors) is the signal.

How to read the conversion rate at small N

The conversion rate the smoke test produces has to be read against three things: the benchmark the founder named in advance, the channel that produced the traffic, and the cost-per-click the founder paid. The three together produce a verdict; any one alone is incomplete.

Step 1 — name the benchmark before the page ships

The benchmark is the conversion rate that the test is positive at. Three rules for naming it.

Rule 1 — base the benchmark on a comparable product. A B2B SaaS smoke test on a LinkedIn audience is not the same as a consumer app smoke test on a Reddit audience. The benchmark should come from a comparable product in a comparable channel. Founders who have not run smoke tests before can use the published benchmarks in the landing-page validation guide as a starting point.

Rule 2 — name three bands, not one threshold. A positive band (the test passes), a neutral band (the test is inconclusive, run more traffic), and a negative band (the test fails). The bands should not overlap; the neutral band exists to keep the founder from making a decision on data that does not deserve one.

Rule 3 — write the benchmark on the same doc as the page. The benchmark lives next to the page, not in the founder's head. The doc is the artifact the founder reads when the data arrives.

Step 2 — separate the conversion rate by channel

The conversion rate on Reddit traffic is not the same as the conversion rate on Google traffic. The conversion rate on cold LinkedIn outreach traffic is not the same as the conversion rate on retargeting. The smoke test produces a single overall conversion rate, but the verdict depends on which channel produced the rate.

Two rules.

Rule 1 — run at least one channel at a time. A smoke test that runs Reddit, Google, and Meta at the same time produces a blended conversion rate that no single channel can claim. The blended rate is informative but not actionable. Run one channel at a time when the budget is small.

Rule 2 — segment the data by source. UTM parameters on the URL let the founder read the conversion rate per channel. The form service the founder picked in minute 45 to 55 should record the UTM source on each submission; if it does not, the founder should add a hidden field that does.

Step 3 — read the verdict against the benchmark and the cost

The verdict is the combination of three numbers: the conversion rate, the benchmark band the rate falls into, and the cost-per-click the founder paid to produce the rate. A 6% conversion rate on Reddit traffic at $0.40 per click is a different verdict than a 6% conversion rate on Google traffic at $8 per click. The first is a positive signal at a cost the founder can afford at launch; the second is a positive signal at a cost the founder cannot afford.

The verdict has three outcomes.

Positive verdict. The conversion rate is in the positive band and the cost-per-click is within the budget the founder planned for customer acquisition. The founder builds the smallest version of the product that delivers the claim in the headline.

Neutral verdict. The conversion rate is in the neutral band, or the cost-per-click is above the planned budget. The founder runs more traffic to either confirm the rate or fail it cleanly, then re-runs the verdict.

Negative verdict. The conversion rate is in the negative band, or the cost-per-click is so high that the positive conversion rate cannot support the planned business. The founder tests a different positioning wedge (a new headline and a new subhead, same signal lines) before declaring the hypothesis dead.

Worked example

A founder has built a CRM for independent insurance agencies with 10 to 50 agents in the US. The interviews pointed in one direction — the renewal reminder workflow is the named pain. The founder spends ninety minutes building the smoke-test landing page and seven days running $75 of Reddit traffic at the named audience.

The page.

Headline: A renewal reminder system for independent insurance agencies that recovers three days a month.

Subhead: Independent insurance agencies with 10 to 50 agents currently spend three days a month on renewal reminder calls. Salesforce assumes inside-sales workflows; spreadsheets assume someone will remember. Our CRM ships the renewal-reminder workflow out of the box, integrated with the agency's existing IVR.

Signal line 1: Renewal reminder calls ship on day one — no configuration, no CSV import, no setup call.

Signal line 2: The reminder integrates with the agency's existing IVR, so the calls go out through the same number the agency has used for ten years.

Signal line 3: The dashboard shows which agents have called which clients, so the agency owner can audit compliance without running a separate report.

Call to action: Email — Join the early-access list.

Footer: Founder name, contact email, privacy link, unsubscribe link.

The traffic.

Seven days on Reddit. Two subreddits relevant to the audience. Total spend: $75. Total clicks: 187. Total submissions: 14. Conversion rate: 7.5%. Cost-per-submission: $5.36.

The verdict.

The benchmark was 5% positive, 2-5% neutral, under 2% negative. The 7.5% rate is in the positive band. The $5.36 cost-per-submission is below the planned $20-per-customer-acquisition budget. The verdict is positive. The founder builds the smallest version of the CRM that delivers the renewal-reminder workflow out of the box and integrates with the named IVR — nothing else.

The same page with a different headline — "A CRM for independent insurance agencies" — produced a 1.8% conversion rate in the same channel. The comparison tells the founder that the headline carries most of the page's persuasion. The wedge is the renewal-reminder workflow, not the CRM category. The positioning pillar at Startup Positioning walks through the discipline of naming that wedge in a sentence the buyer can repeat in a sales call.

Common mistakes

Six failure modes that produce a smoke test the founder cannot read.

Mistake 1 — using the same page for the smoke test and the launch. The smoke test is built to be measured. The launch page is built to convert. The two pages have different jobs and different structures. Mixing them produces a page that does both jobs badly.

Mistake 2 — running paid traffic to a page the founder cannot change. A smoke-test page that is hard to edit is a smoke-test page the founder cannot iterate. The founder should be able to swap the headline in five minutes and ship the change. The build should use static hosting or a no-code platform with versioning.

Mistake 3 — naming the benchmark after the data arrives. The benchmark is named before the data arrives. Naming it after biases the read. The discipline of naming the benchmark in advance is what makes the smoke test a test and not a fishing expedition.

Mistake 4 — confusing traffic volume with conversion rate. A smoke test that gets 5,000 visitors and 25 submissions has the same conversion rate as a smoke test that gets 500 visitors and 2.5 submissions. The volume is not the artifact. The rate is.

Mistake 5 — declaring the test negative on a single channel. A smoke test that produces a 1% conversion rate on Reddit traffic may produce a 6% conversion rate on LinkedIn traffic to the same named audience. Channels differ; the founder should run the test in at least two channels before declaring the hypothesis dead.

Mistake 6 — converting the waitlist into a sales pipeline before validating the product. A 7.5% conversion rate produces 14 email addresses. The 14 addresses are not customers; they are signal that the hypothesis converts at that rate. Sending the 14 addresses a sales pitch before the product exists is the polite-yes problem in reverse — the founder is asking the 14 to predict their behavior on a Tuesday morning three months from now. The discipline is to wait until the product ships.

FAQ

How long does it actually take to ship a smoke-test landing page?

The build itself is sixty minutes if the founder already has a domain and a deployment account. The full process, including the headline iteration, is closer to ninety minutes for a first-time founder. The page does not need to be polished; it needs to be live and measurable. A page that takes three days to ship is not a smoke test; it is a launch page in disguise.

What conversion rate should a smoke test produce?

There is no universal benchmark. The benchmark depends on the channel, the audience, the product category, and the call to action. A B2B SaaS smoke test on LinkedIn traffic at a waitlist signup typically runs 3-8%. A consumer app smoke test on Reddit traffic at a waitlist signup typically runs 2-6%. The benchmark the founder names should come from a comparable product in a comparable channel. The published conversion-rate bands for landing pages in the landing-page validation guide are a starting point, not a verdict.

Can a smoke test use organic traffic instead of paid?

Yes, with two cautions. Organic traffic is harder to attribute to a specific audience; the conversion rate on organic traffic is a blend of motivated and unmotivated visitors. And organic traffic is harder to scale; a smoke test that gets 50 visitors from a tweet produces a less reliable conversion rate than a smoke test that gets 500 visitors from a paid campaign. Paid traffic is the disciplined default; organic traffic is the fallback when the budget is zero.

How is a smoke test different from a fake-door test?

A smoke test asks the visitor for a real action — an email signup, a reservation deposit, a pre-order, a waitlist join. A fake-door test asks the visitor for a button click that leads to a "coming soon" or "thanks for your interest" page. The smoke test is the more disciplined version because it produces a real lead the founder can talk to. The fake-door test is the fallback when the smoke test cannot run — for example, when collecting an email is illegal in the founder's jurisdiction or premature for a sensitive product. The conceptual comparison is in Fake Door Test: How to Validate Demand Without Writing Code.

What if the conversion rate is high but the cost-per-submission is also high?

A high conversion rate at a high cost-per-submission is a different verdict than a high conversion rate at a low cost-per-submission. The first is a signal that the audience is interested but expensive to reach; the second is a signal that the audience is interested and affordable. The verdict depends on the planned customer-acquisition budget. A founder whose planned CAC is $20 cannot afford a $50 cost-per-submission; the same founder can afford a $5 cost-per-submission.

What if no traffic converts at all?

A 0% conversion rate at any sample size is a signal. The signal is one of three things: the wrong audience, the wrong headline, or the wrong channel. The first iteration is a new headline; the second is a new audience; the third is a new channel. A founder who cannot produce any conversion on any iteration is testing a hypothesis that does not survive contact with a real visitor. The discipline is to test three iterations before declaring the hypothesis dead.

Can the smoke test run on a friend's audience?

No. Friends who convert because they want to support the founder are noise, not signal. The conversion rate on a friend's audience is always higher than the conversion rate on a stranger's audience. The smoke test is meaningless if the traffic is not strangers.

How long should the smoke test run?

Seven to fourteen days is the typical window. Shorter than seven days produces too few conversions to be reliable; longer than fourteen days allows the audience to saturate and the conversion rate to drift. A founder who runs the test for twenty-one days is reading a different signal at the end than at the beginning. The discipline is to name the duration in advance and stop the test at the planned date.

Summary

A smoke-test landing page is a real page on a real domain, built in sixty minutes, that asks the visitor for one action and produces a conversion rate the founder reads against a benchmark named in advance. The page has five structural elements — headline, subhead, signal lines, call to action, footer — and seven things left off — testimonials, team bios, press logos, feature lists, pricing, video, chat widgets. The build order is headline first, signal lines second, form third, deploy fourth, traffic fifth. The conversion rate is read against three things: the benchmark, the channel that produced the traffic, and the cost-per-click the founder paid.

The smoke test is the discipline that converts customer-interview signal into conversion-rate signal. The interview signal is evidence that the problem is real. The smoke-test signal is evidence that the named solution will convert. The asymmetry between the cost of running the test (one afternoon) and the cost of skipping it (three to nine months of building the wrong thing) makes the smoke test the rational default for any founder who has not yet validated at the conversion-rate level.

The next move depends on the verdict. A positive verdict means the founder builds the smallest version of the product that delivers the claim in the headline. A neutral verdict means the founder runs more traffic. A negative verdict means the founder tests a different positioning wedge — a new headline, a new subhead, same signal lines — before declaring the hypothesis dead.

Related reading

• Fake Door Test: How to Validate Demand Without Writing Code — the conceptual sibling: the variant of the smoke test where the call to action leads to a "coming soon" page rather than to a real form.
• Landing Page Validation — the broader discipline: the channel mix, the conversion-rate bands, the multi-page sequence, and the named benchmark sources.
• How to Validate a Startup Idea Before Building — the canonical four-stage validation framework that the smoke test sits inside.
• How to Validate Demand Before Coding — the practical bridge from "I have an idea" to "here is the cheapest experiment that produces real evidence."
• Willingness to Pay Validation — the framework for the next step once the smoke test has produced a conversion rate: turning the waitlist signal into a willingness-to-pay signal.
• How to Test Whether People Will Pay Before You Build — the tactical playbook for the six experiments that follow a positive smoke-test verdict.
• Startup Positioning — the discipline of writing the positioning wedge the headline carries; the wedge is what makes the headline specific enough to convert.
• Startup Validation Checklist — the operational checklist that turns the smoke test into a week-by-week plan.
• Startup Validation hub — the full reading path and topical taxonomy.

Sources

The smoke-test framework, the five-element structure, and the read-the-conversion-rate discipline draw on the following primary and secondary sources.

• Rob Fitzpatrick, The Mom Test (2013), free at momtestbook.com — the canonical rule against asking "would you buy this?" in interviews; the smoke test is the answer to that question when an interview cannot.
• Eric Ries, The Lean Startup (2011), Chapter 9 on Validated Learning — the build-measure-learn loop and the discipline of testing the riskiest assumption before building. The smoke test is the smallest validated-learning experiment that produces a number, not an opinion.
• Steve Blank, The Four Steps to the Epiphany (2005) — Customer Development as the published methodology; the customer-discovery interviews that produce the named buyer and the named gain in the headline are Blank's.
• April Dunford, Obviously Awesome (2019) — the modern positioning methodology; the headline and the subhead carry the wedge Dunford describes.
• Ash Maurya, Running Lean (2012) — the discipline of validating the problem before the solution; the smoke test is the smallest version of the problem-solution-fit test.
• Teresa Torres, Continuous Discovery Habits (2021) — the modern Customer Discovery methodology; the weekly iteration cadence between smoke tests is Torres'.
• Paul Graham, "Do Things That Don't Scale" (2013) — the essay that named the founder-psychology observation that the unscalable work (hand-shipped landing pages, hand-collected waitlist emails) is the path; relevant where the smoke test is itself the unscalable work.
• The Reddit Ads documentation (advertising.reddittor.com/docs) and the Google Ads documentation (support.google.com/google-ads) — the actual platform surfaces the founder uses to buy traffic and to read the conversion-rate data per channel.
• The ConvertKit, Beehiiv, Loops, and Tally documentation — the form services that wire the call to action to a real email-collecting backend in under ten minutes.
• The Stripe checkout documentation and Payment Links documentation — relevant when the smoke test upgrades from a waitlist signup to a reservation deposit or pre-order; the conversion-rate discipline is the same.

Next action

Build the page tonight. Open a Google Doc. Write the five element labels — Headline, Subhead, Signal line 1, Signal line 2, Signal line 3, Call to action, Footer. Write three candidate headlines. Pick the one that names the buyer and the gain most specifically. Write three candidate subheads. Pick the one that names the alternative most specifically. Write the three signal lines. Wire the form to a real service. Deploy to Vercel, Netlify, GitHub Pages, or Cloudflare Pages. Buy $50 to $100 of traffic in a channel the named buyer actually uses. Run the test for seven to fourteen days. Read the conversion rate against the benchmark you named in advance.

If the verdict is positive and you want a structured second opinion on which of the headline's claims carries the most risk before you commit the build, Startup MRI's validation analysis takes about five minutes and surfaces the parts of the idea most likely to break under real-world pressure. The smoke test is the conversion-rate signal; the analysis is the structural signal. The two together are the smallest disciplined set of inputs a founder can have before writing the first line of product code.