A Pre-Launch Website Checklist for Forms, Tracking, and Mobile Usability
A website can look finished in a preview and still fail at the moment it needs to work. A contact form may send visitors into a dead end. Analytics may count every page view twice. On a phone, the main button may be hidden below a crowded menu.
These problems are easier to find before launch, when the people building the site can still make changes. This website launch checklist focuses on three areas that directly affect whether visitors can act and whether your team can tell what happened: forms, tracking, and mobile usability.
You do not need a large testing team. You do need a clear list of important actions, a few real devices or screen sizes, and someone who can verify that a test submission reaches the right place. Keep a record of what you tested, what failed, and who will fix it.
Start with the actions your website must support
Before testing individual pages, write down the handful of actions that matter to the business. For a local service company, that might be calling, requesting an estimate, and finding the service area. For a small online shop, it could be searching for an item, adding it to a cart, and completing checkout.
Make the list specific. “The website works” is too broad to test. “A visitor can submit the quote form on a phone, sees a confirmation, and the sales inbox receives the details” is testable.
- List the main actions visitors should take.
- Identify which page or button starts each action.
- Decide what successful completion looks like.
- Name the person responsible for checking the result.
If the site is a redesign, compare the new journey with the current one. A change in layout can unintentionally remove a useful route or make a key page harder to find. For broader pre- and post-launch search tasks, see our guide to SEO for a new website.
Website launch checklist for forms
Submit every important form from start to finish
Test each distinct form, not just one representative example. A contact form, newsletter signup, booking request, and quote form may use different integrations or notification rules. Use realistic test information, but do not submit a real customer’s personal details without permission.
For each form, check that required fields are genuinely required, optional fields are clearly optional, and the labels explain what information is expected. Try submitting with a required field empty and with an invalid email address. The error should identify the problem near the relevant field, explain how to correct it, and preserve any information already entered. A generic “Something went wrong” message leaves the visitor guessing.
Then complete a valid submission. Confirm all parts of the journey:
- The form accepts the information and does not get stuck loading.
- The visitor sees a clear confirmation on the page or reaches a useful thank-you page.
- The correct staff member or shared inbox receives the submission.
- The message includes the fields the visitor entered, with no missing or mislabelled details.
- Any connected system, such as a booking calendar or customer database, receives the record if that is part of the process.
Do not treat an on-screen success message as proof that the submission arrived. Check the receiving inbox or system directly. If a notification is sent to one person, test what happens when that person is away: a shared address or backup process can prevent requests from sitting unseen.
Check form details that are easy to overlook
Review the form’s privacy and consent language. If you ask visitors to agree to marketing emails, make the choice clear and separate from a request they need to submit. Do not preselect an optional marketing checkbox by default. Make sure the privacy notice linked from the form is the right one and opens successfully.
Check whether confirmation emails are expected. If they are, verify the subject line, reply-to address, and wording. A customer should be able to tell that their request was received and what happens next. Avoid promising a response time the business cannot consistently meet.
Spam protection can also create friction. Test the form as a normal visitor and confirm that any challenge, verification step, or blocked-submission message is understandable. If the form uses a third-party service, check that the service is configured for the live domain rather than only a development or preview address.
Finally, decide who will periodically test the forms after launch. An update to a plugin, email service, or form integration can break a workflow that worked on launch day.
Website tracking: verify what gets measured
Choose a small set of useful events
Tracking is most useful when it answers a real question. Start with the actions that matter: a completed lead form, a booking request, a purchase, or a click to call. Record what counts as a conversion—the specific action you consider a meaningful result—and where it should be measured.
A click on a submit button is not necessarily a completed form. Visitors may click and then encounter an error. Where possible, measure success after the form is accepted, such as on a confirmation page or through a confirmed submission event. The method depends on how the site and form are built, so have the person responsible for analytics explain what triggers the event.
Write a simple tracking plan before launch. For example:
- Event: quote request completed.
- Trigger: the form is accepted and its confirmation appears.
- Page or form: the estimate request page.
- Purpose: compare completed requests with traffic sources over time.
Keep the first version modest. Tracking every menu click can create clutter without helping you make a decision. Start with a few events that show whether key journeys work.
Test analytics without mistaking setup for proof
Before launch, confirm that the measurement code or tag is installed on the pages where it should run. Then visit the live or staging site and check whether page views and key events appear in the intended analytics property. If the site is behind a password or on a temporary domain, confirm that the setup will still work on the production domain.
Watch for duplicate measurement. A common cause is installing the same tracking code directly on the site and again through a tag manager or platform integration. Duplicate page views can make reports misleading. Check a page load and a form completion once, then compare the recorded activity with what you actually did.
Also decide how your team’s own visits will be handled. Staff and contractors may generate visits while checking pages, which can muddy early reports. Depending on your analytics setup, you may be able to exclude internal traffic or use a separate testing environment. Document the choice so people understand why internal visits are or are not included.
Campaign links need a little attention too. If you plan to use tagged links in email or advertising, agree on a consistent naming approach before different people create their own variations. Otherwise, the same campaign can appear under several names in reports.
Finally, check your consent and privacy setup. Requirements depend on where your business and visitors are located and what data you collect. Do not assume that a banner alone makes a site compliant; confirm the settings and wording with an appropriate privacy professional when needed.
For search performance, make sure the right version of the site is being monitored and that your measurement accounts are connected as intended. Our guide to finding SEO issues in Google Search Console explains what that tool can help you check after launch.
Mobile usability: test real tasks, not just a smaller screen
Use a phone to complete the main journey
A responsive preview in a design tool is useful, but it cannot show every real-world issue. Test the live site or a reliable staging version on at least one iPhone and one Android phone if you can borrow them. Also inspect a narrow and a wider browser window on a computer. Devices differ in browser behavior, screen size, and keyboard handling.
On a phone, try the same tasks you listed at the start: find a service, open the contact page, submit a request, or reach the checkout. Notice whether the menu is easy to open and close, whether important content appears before a large decorative image, and whether a visitor can get back to where they were.
Do not just tap the primary button. Scroll through the page and look for content that is clipped, overlapping, too small to read, or awkwardly spaced. Rotate the phone if that is relevant to your audience. Check that cookie notices or promotional overlays do not cover the form or make the page impossible to use.
Pay particular attention to mobile forms
Enter information using the phone’s on-screen keyboard. Make sure the right keyboard appears for each field: an email field should make it easy to enter an email address, and a phone field should not force visitors through unnecessary text entry. Test whether the browser’s autofill works where appropriate and whether the “Next” or “Done” key takes the visitor to a sensible next step.
Check that field labels remain visible while typing. Placeholder text inside a field is not a reliable substitute for a label because it can disappear as soon as someone starts entering information. Verify that the submit button is visible and easy to activate without repeatedly zooming or scrolling sideways.
Keyboard access matters for visitors who do not use a touchscreen or mouse, and it can expose usability problems for everyone. Use the Tab key to move through the page and form controls, and confirm that the current focus is visible. The W3C Web Content Accessibility Guidelines provide standards-based guidance on accessible web content, including keyboard operation and identifying form inputs.
Try to complete the form with one hand if that reflects how customers are likely to use the site. This is not a formal accessibility test, but it can reveal controls that are too close together or a process that asks for more effort than the task warrants.
A practical pre-launch test record
A short test log keeps small teams from relying on memory. A spreadsheet is enough. Include the page or task, device and browser, test result, issue, priority, owner, and retest date. Use clear outcomes such as “pass,” “fail,” or “blocked,” rather than notes like “looks okay.”
Prioritize problems by their effect on visitors. A broken lead form is a launch blocker. A slightly uneven icon may be worth fixing, but it should not take priority over a checkout error or a phone layout that hides the call button.
- Test each important form and verify the submission reaches its destination.
- Confirm that the success state, validation messages, and any follow-up email make sense.
- Verify analytics page views and a small number of key conversion events.
- Check that tracking is not duplicated and that the live domain is configured.
- Complete the main visitor journeys on real phones and with a keyboard.
- Record problems, assign owners, and retest fixes before launch.
After launch, repeat the highest-priority checks on the production site. A domain change, security setting, caching layer, or last-minute content edit can behave differently from staging. Keep an eye on the first real submissions and analytics events, while remembering that a quiet first day does not by itself prove something is broken. If the site receives visits but not enquiries, use a structured diagnosis rather than changing everything at once; our article on what to check when a website gets traffic but no leads offers a useful next step.
FAQ
How many times should I test a contact form before launch?
Test each form type at least once with a valid submission and once with a common error, such as a missing required field. Also test any different routing rules or integrations separately. Verify the result in the receiving inbox or system, not only on the website.
Should I track clicks on the submit button or successful submissions?
For measuring leads, a confirmed successful submission is usually more useful than a button click. A click can happen even when validation fails. If you also track clicks to understand form friction, keep that event separate from the conversion.
Do I need to test on both iPhone and Android?
Testing both is helpful because browsers, screen sizes, and keyboards can behave differently. If you only have access to one phone, test that device thoroughly and inspect additional screen widths in a browser. Borrowing a second device for the main forms and navigation is a practical minimum.
What should I do if tracking is not showing data before launch?
Check that the correct measurement property is connected, that the code is present on the page, and that your privacy or consent settings are not preventing the test. Confirm whether a staging domain is excluded by design. Ask the person who configured analytics to show how a test event is expected to appear.
Can a website launch before every minor issue is fixed?
Yes, if the remaining issues do not prevent visitors from completing important tasks, create accessibility barriers, or undermine data collection you rely on. Record lower-priority fixes, assign an owner, and set a review date. Do not launch with a broken lead or payment journey simply to meet a calendar date.