Test demand with an honest landing page

Decide which uncertainty the page will test

A landing page can test whether a reachable audience understands an offer and takes a specific next step. It cannot, by itself, prove that a broad market exists, that the product works, or that sign-ups will become paying customers. Before writing copy, state the hypothesis and the decision you will make from the result.

Example: “Owners of independent language schools who currently coordinate make-up lessons by email will book a 20-minute workflow review if a proposed tool shows how it could reduce that coordination.” That is more testable than “people want easier scheduling.” If willingness to pay is the uncertainty, include a credible price or ask for a real paid commitment; an email signup tests a different, weaker behavior.

Make the page a faithful representation

Include the customer, job, outcome, essential constraints, price or pricing approach if known, and what happens after the action. Use a short headline, concrete explanation, and one primary call to action. Show a prototype or screenshots only when they represent what exists or what you can reasonably build. Say clearly when the product is a concept or pilot. Never present a mockup as an available product or imply customer results you do not have.

The GOV.UK prototype guidance treats prototypes as a way to test realistic interactions and learn before committing to a full service. A landing page is a much narrower test: it measures response to an explanation and action, not the usability or effectiveness of the future product.

Define the audience and count the right denominator

Choose one channel where the intended audience is plausibly present. Keep a record of source, date, audience, geography, and campaign. Count qualified visits, not every page load: remove your own traffic, bots, accidental clicks, and visits outside the intended segment where you can do so responsibly. A conversion rate without a defined audience and denominator is easy to misread.

Decide the outcome before launch. Examples, from weaker to stronger evidence, include: a relevant visitor reads the offer; requests a demo and describes their current workflow; schedules and attends a call; starts a pilot; provides payment details or pays a disclosed deposit. Stronger actions cost the participant more and require a more complete, honest offer. Do not use urgency or scarcity unless it is real and verifiable.

Set a threshold that matches your next decision

Do not borrow a universal “good conversion rate.” Estimate how many qualified people you can reach, the minimum number of conversations needed to learn, the cost and time of the next step, and what level of response would justify that cost. Write down:

Audience and channel:
Offer and primary action:
Qualified-visit definition:
Observation window:
Continue if:
Change the offer or segment if:
Stop if:
Known confounders:

Small, targeted traffic is usually too imprecise for fine-grained A/B conclusions. Use the page to recruit and learn from plausible users, not to claim statistical certainty from a handful of sign-ups. If you compare variants, change one meaningful element at a time, keep the audience and channel comparable, and note outside events that could affect response.

Ask neutral questions after the action

Follow up quickly and ask what the visitor does today, what prompted the action, what they expected, and what would prevent adoption. Avoid leading questions such as “Wouldn’t this save you time?” Pew Research Center documents how wording, response options, and question order can shape survey answers; apply the same care to signup questions and follow-up forms. Ask for only information needed to contact or understand the participant, explain how it will be used, and set a deletion plan.

Keep the experiment honest and safe

Do not collect payment for a product you cannot provide unless the terms, timing, refund path, and uncertainty are clear. Do not use false testimonials, fake customer counts, fabricated countdowns, hidden recurring charges, or signup flows that obscure what is being tested. The FTC’s U.S. report on dark patterns describes deceptive interface tactics such as false urgency and difficult cancellation; rules vary by jurisdiction, so this is a warning signal, not a complete global legal analysis.

At the end, report the denominator, audience, channel, dates, actions, and what you learned. Keep the page live only if the offer is still accurate. Use an experiment brief to record the decision and choose the next test.

If the product is not ready, use a clearly disclosed waitlist or intent test. If prospects are ready to commit to a defined outcome, plan a transparent paid pilot with delivery and cancellation terms in writing.