CRO Test Variants for Landing Pages and Forms

Flint makes it straightforward to set up CRO (conversion rate optimization) test variants. You can create variant pages, instrument them with your preferred A/B testing platform, and run experiments on everything from hero copy to form placement and button text.

For a general overview of A/B testing in Flint, see the A/B Testing guide.

The first step is to create a duplicate of the page you want to test.

1. 1.Open the AI chat and ask Flint to duplicate your page:

Create a duplicate of the /lp/my-landing-page page at /lp/my-landing-page-v2

2. 2.Once the variant page is created, navigate to it using the Pages menu.
3. 3.Ask Flint to apply the changes you want to test:

On this page, change the hero headline to "Start building in minutes" and move the email capture form above the fold.

Flint integrates with Statsig, LaunchDarkly, Eppo, and other feature flag platforms. To instrument your variant:

1. 1.Add the SDK script to your site via Settings > Site Settings > Custom Code > Head Scripts.
2. 2.Ask Flint to add the appropriate CSS classes or data attributes to the elements you want to test:

Add the class "hero-headline" to the main headline in the hero section, and "cta-form" to the email capture form container.

3. 3.In your testing platform, set up an experiment that targets those classes and routes traffic between your control and variant URLs.

To test different form configurations:

• •Form placement - Ask Flint to move the form to a different position on the variant page (e.g., above the fold, inline with content, or in a sticky sidebar).
• •Form copy - Update the form headline, placeholder text, or button label on the variant.
• •Form style - Ask Flint to change the visual design of the form section (background, layout, button color) on the variant page.

If you are using a HubSpot or embedded form, you can also test different form IDs to compare conversion rates between form configurations.

Once your variant page is ready and instrumented:

1. 1.Publish both the control and variant pages in Flint.
2. 2.Set up your traffic split in your testing platform (e.g., 50/50 between the control and variant URLs).
3. 3.Configure your conversion goal (form submission, button click, scroll depth) in your platform's event tracking.
4. 4.Start the experiment and monitor results in your testing platform dashboard.

For Google Ads campaigns, you can also use the Google Ad Experiments module to split traffic between your control and variant landing pages.

• •Test one change at a time - Isolating a single variable (headline, form position, CTA text) gives you clearer signal on what drove the difference.
• •Give tests time to reach significance - Avoid stopping tests early. Most platforms will indicate when results are statistically significant.
• •Use query parameters for personalization - Combine CRO testing with Flint's query parameter support to create dynamic variants without duplicating pages.
• •Iterate on winners - Once a variant wins, make it your new control and test the next hypothesis.