How to A/B test a landing page without a developer
You do not need a developer to A/B test a landing page. With a testing tool that has a visual editor, you paste one script tag once, then change text, images and buttons by pointing at them. This guide walks through it, from choosing what to test to reading the result.
What you need
- A site where you can add a script to the head of every public page, once. Most site builders and content management systems have a setting for custom code in the site-wide head; check your builder's help pages for where it is.
- You only touch that setting once. After that, every test is set up in the testing tool.
- A goal you can measure on the page or the next one: a button click, a visit to a thank-you page, or a custom event.
Step 1: choose one thing to test
Start from what you think stops visitors converting. Good first tests on a landing page:
- the headline: a different promise, not different words for the same one
- the main call to action: its wording, or what it offers
- the hero image or the first screen
- the form: fewer fields, or a different first question
- social proof: customer logos or quotes near the call to action
Write the idea down as "If we change X, more visitors will do Y, because Z." It keeps you honest when you read the result.
Step 2: install the script once
Copy the testing tool's script tag into the head your public pages share, through your builder's custom code setting, then publish the site. It has to be on the page you test and on the page where the goal happens, such as a thank-you page. Use your live domain: preview and staging addresses on site builders are not the domain your project is set to, so visits there are not counted. In Twinpage, the install page checks that the script is working on your domain.
Step 3: create the variant without code
Open the page in the visual editor, click the element you want to change and edit it: text, colours, images and links, or hide or move an element. If you have two different page designs, use a redirect test that sends half the visitors to the other page instead. On Growth, you can also add your own JavaScript.
Step 4: pick one goal and start
Choose the goal from Step 1's "more visitors will do Y": a click on the call to action, a visit to the thank-you page, or a custom event. Start the test. Visitors are split evenly between the original and the variant, and each visitor keeps seeing the same version.
Step 5: read the result
Wait until each version has enough visits; see How long to run an A/B test. Twinpage shows visits and conversions for each version with a plain-language verdict: not enough data yet, no clear difference, leading, or likely better. A winner needs at least 1,000 visits and 25 conversions per version, and Twinpage warns you if the traffic split looks uneven, which often means the install is broken on one version.
When you have a winner, make the change on the page itself and stop the test. When there is no clear difference, keep the version you prefer and test something bolder.
Common mistakes
- Testing several changes in separate tests on the same page at the same time, so the tests affect each other.
- Changing the page while a test runs.
- Stopping the test the first day one version looks ahead.
- Testing a tiny change on a page with little traffic. If that is your situation, read A/B testing with low traffic.
Frequently asked questions
Will the test slow down my page? A well-built testing script is small. To stop visitors seeing the original flash first, the install code keeps the page hidden until the change is applied, for a short moment at most; after that limit the page shows whatever happens. Ask any tool you consider how long that hide can last.
Can I test a page built with a site builder? Yes, if the builder lets you add a script to the page's head on your own domain.
How many versions can I test at once? In Twinpage, up to 5 versions in all depending on your plan. On a small site, start with two.
Testing with Twinpage
Twinpage is A/B testing for marketing sites: one script tag, a visual editor, one main goal per test and a verdict you can trust, with prices you can read before you start.