Landing Page Test SOP

A landing page test SOP tells you whether strangers want your solution badly enough to hand over their email, and it takes one to two weeks from first draft to decision. You build a single page that promises one clear outcome, send targeted visitors to it, and count what they do. The page is cheap to build and expensive to fake, because cold visitors owe you nothing. If they sign up and then answer a follow-up question or book a call, you have demand. If they bounce, you saved months of building.

What is the landing page test and when should you use it?

A landing page test is a one-page website for a product that does not exist yet. It explains the problem, the promised result and how it works, with one button: join the waitlist, request early access or book a demo. You measure the share of visitors who click and complete the form, then filter for intent with a second step.

Use this SOP:

• After customer interviews confirmed the problem and you want to know if your proposed solution attracts people who never met you.
• When you need a number to compare ideas. Two pages with the same traffic source give you two conversion rates you can put side by side.
• Before you spend on design or code, so the first build serves people who already raised their hand.

This SOP tests the whole offer on its own page. If you want to test a single button inside a product or newsletter, use the fake door testing SOP. If you want people to pay, move to the pre-order campaign SOP.

What do you need before you start?

• Quotes from 10+ customer conversations. Your headline should use their words. The customer interview SOP gives you those quotes.
• One target segment and one channel where that segment gathers: a subreddit, a LinkedIn audience, a Facebook group, a Slack community.
• A no-code page builder such as Carrd, Webflow or Unbounce, plus a form tool such as Tally.
• Analytics with a signup event: Plausible or Google Analytics, and Microsoft Clarity for free session recordings.
• A budget of $100 to $300 for ads, or permission to post in two or three communities.
• A written pass threshold, set before any traffic arrives.

How do you run it, step by step?

1. Write the hypothesis and the pass line

Put it in one sentence: "At least 10% of [segment] visitors from [channel] will join the waitlist, and 30% of those will answer the follow-up question." Writing the numbers first stops you from moving the goalposts after you see the data.

2. Write the page from customer words

Use a headline that names the outcome, a subheadline that says how, three benefits, one picture of the result and a short FAQ. Borrow phrases straight from interview notes. Keep the page to a single promise and a single button. Add an honest line such as "We are building this now. Join to get early access."

3. Build it in one day and set up tracking

Publish on a cheap domain or a subdomain. Add analytics, fire an event when the form is submitted and tag every link with UTM parameters so you know which source sent each visitor. Test the form on mobile, because most social traffic arrives on a phone.

4. Add a second, harder step

After the email form, show a thank-you screen with one more ask: answer "What do you use for this today?", pick a plan tier, or book a 15-minute call. Signups measure curiosity. The second step measures intent.

5. Send targeted traffic

Run one small ad set aimed at your segment, or post in communities where the rules allow it. Keep each source separate so you can compare them. Friends, family and your own followers do not count, because they click to support you.

6. Run until the sample is big enough

Wait for at least 300 unique visitors per source, or two weeks, whichever comes later. Do not edit the page during the test. If you want to compare headlines, run two pages at the same time with the same traffic split.

7. Contact every signup within 48 hours

Send a short personal email asking what made them sign up and what they use today. Reply rate is a quiet signal of how much the problem matters. Book calls with the people who answer.

Template you can copy

LANDING PAGE TEST: PAGE SKELETON

Headline (max 10 words): [Outcome] for [segment] without [main frustration]
Subheadline (1 sentence): [Product] does [how it works] so you can [result].
Button: Get early access

Section 1, The problem: 2 sentences in the customer's own words.
Section 2, How it works: 3 steps with one icon each.
Section 3, What you get: 3 benefits, each with a number or a time saving.
Section 4, FAQ: price range, launch date, who it is for.
Footer: "We are building this now. No spam, unsubscribe any time."
        Link to privacy notice.

Thank-you screen (second step):
"Thanks! One question so we build the right thing:
 What do you use for [problem] today?" [short text field]
 Optional: [Book a 15-minute call] button

TEST LOG (one row per source)
Source | UTM | Spend | Visitors | Signups | Signup % | Cost per signup |
Second step done | Second step % | Replies to follow-up email

Worked example: a meal planner for parents of kids with allergies

This example is illustrative. A founder interviewed 14 parents whose children have several food allergies. Her page promised "A week of allergy-safe dinners in 10 minutes." She spent $240 on Meta ads aimed at members of allergy parenting groups and posted once in two forums.

Cost per signup from ads was $3.24. Nine people booked a call. Answers to the second step showed that 22 of 48 respondents already paid for a generic meal kit and threw away half of it. The test passed, and her next step was a pre-order campaign at $12 a month.

What should you measure, and when do you stop?

Stop when each source reaches 300 visitors or two weeks have passed. Treat the result as directional: with a few hundred visitors, the difference between 4% and 6% is noise. Big gaps such as 2% versus 12% are a real signal. If you land in the middle, change one thing (headline or audience) and run a second round.

Which mistakes should you avoid?

• Testing with warm traffic. Your LinkedIn followers sign up to be nice. Pay for or post to strangers who match the segment.
• Mixing audiences. Freelancers and agencies on the same page give you one blended number that describes neither group.
• Writing a clever headline. Visitors decide in seconds. Say the outcome in plain words.
• Skipping the privacy notice. If you collect emails from people in the EU, the GDPR requires a clear notice about how you use them.
• Inventing proof. Fake testimonials, logos or user counts break trust and, in many countries, consumer protection law.
• Editing during the test. Every change resets your sample.

How can AI speed this up?

Paste your interview quotes into an AI assistant and ask for five headlines that reuse the customers' exact words, then pick the plainest one. AI site builders inside Webflow, Framer or Carrd can produce a first layout in minutes. After the test, export the second-step answers and ask the assistant to group them by current tool and frustration. The human check: read every answer yourself, and make sure every claim on the page is something you can deliver.

The free landing page module of the F/MS startup school walks through page structure, and the marketing budget calculator helps you size the ad spend before you start.

Where does this SOP fit in your validation plan?

The landing page test sits early in the Solution Validation stage, right after you confirm the problem.

1. Before: customer interviews and the problem statement validation SOP give you the words for the page.
2. In parallel: compare messages with solution value proposition testing when you are unsure which promise to lead with.
3. After: ask for money with a pre-order campaign, and check whether your cost per signup can turn into a healthy cost per customer with customer acquisition cost validation.

See all Solution Validation SOPs.

FAQ

What is a good conversion rate for a landing page test?

For a waitlist page with cold, targeted traffic, 10% or more is a strong signal and under 3% is weak. Warm traffic from your own network converts higher and tells you little.

How many visitors do I need for a landing page test?

At least 300 unique visitors per traffic source. Below that, a handful of signups swings the rate too much to trust.

Do I have to say the product does not exist yet?

Yes. Say it is in development and that signups get early access. Honest pages still convert, and the people who join know what they signed up for.

Can I run a landing page test for a B2B product?

Yes. Use LinkedIn ads or direct outreach as the traffic source and make the button "Book a 20-minute demo". Expect fewer visitors, so judge the result on booked calls and their quality.