Landing Page A/B Testing: What to Test First for More Leads

If you are paying for every click, a landing page that converts even slightly better lowers the cost of every lead that follows. Landing page A/B testing is how you find those improvements with evidence instead of opinions, but only if you test the right things in the right order.

Most advertisers start with button colours and stop after three days. This guide explains what to test first, how much traffic you need, how to measure it properly in GA4, and where Indian businesses with modest budgets tend to go wrong.

What is landing page A/B testing?

Landing page A/B testing is showing two versions of a page to similar visitors at the same time and comparing how many take the action you care about. Version A is your current page, version B changes one element, and traffic is split between them at random.

The word "random" matters. If you run page A this week and page B next week, you are comparing different days, different ad budgets and different buyers, not different pages. A real test runs both versions in parallel.

It also helps to be clear about what a test cannot do. It cannot fix a weak offer by itself, and it cannot rescue a page that is slow or poorly matched to the ad. Fix those first, as covered in our guide to PPC landing page best practices, and then start testing.

What should you A/B test first on a landing page?

Test the elements with the largest likely effect first: the offer, the headline and first screen, the call to action and form, then proof and layout. Visitors decide in the first few seconds whether the page is for them, so changes high on the page and close to the promise usually matter more than visual polish.

This ranking is a working assumption based on how landing pages are usually built, not a guarantee. Your own audience may react differently, which is exactly why you test.

Write a one-line hypothesis for each test, in the form "If we change X, then Y will rise because Z." For example: "If we replace the 6-field form with name, phone and one question, more mobile visitors will submit because typing on a phone is slow." A hypothesis tells you what you learned even when the new version loses.

How much traffic do you need for a valid test?

You need enough visitors per variant to detect the smallest improvement you care about, and you should calculate that number before you launch. A commonly used rule of thumb for 95% confidence and 80% power is:

n per variant ≈ 16 × p × (1 − p) ÷ d²

Here, p is your current conversion rate and d is the absolute lift you want to detect. These worked examples use made-up inputs to show the scale, not benchmarks:

Halve the lift you are trying to detect and you need roughly four times the traffic. That is why small tweaks are hard to prove unless you have a lot of visitors.

Many Indian service businesses send only a few hundred paid clicks a month to one page. On that traffic, a button-colour test could take months to read. Test bold changes instead, such as a new offer, a much shorter form or a different page structure, where the difference between versions is large enough to show up quickly. For a rough sense of what clicks cost, see our breakdown of Google Ads cost in India.

How do you run an A/B test step by step?

Follow the same sequence every time, so results are comparable and nothing is decided by gut feel.

1. Pick one primary metric. Use a real business outcome such as a qualified lead, booking or purchase, not page views.
2. Check tracking first. Submit the form, tap call and WhatsApp buttons, and confirm each event arrives in GA4 and Google Ads.
3. Write the hypothesis and decide what single element changes.
4. Calculate sample size and the minimum run time before launch.
5. Split traffic evenly between the versions using a testing tool or two URLs.
6. Send the variant name to GA4 as an event parameter so you can report by version.
7. Run whole weeks and avoid launching during a sale, festival or major campaign change.
8. Decide the winner only after reaching your planned sample, then check lead quality with your sales team.
9. Record the result, win or lose, and plan the next test.

Step 8 is where many tests go wrong. A version that gets more form fills can still bring worse leads, so compare cost per qualified lead, not just submissions.

Which tools can you use for landing page A/B testing?

After Google Optimize ended on 30 September 2023, you need a separate testing tool or separate page URLs to split traffic. Google's own help page confirms that Optimize and Optimize 360 are no longer available and that experiments still active on that date ended.

GA4 is best treated as the scoreboard, not the referee. It records events and can report conversion rates by variant if you pass the variant name in, but it does not randomly assign visitors or calculate significance for you.

Google describes custom experiments in Google Ads as typically used to test Smart Bidding, keyword match types, landing pages, audiences and ad groups. That is useful for comparing two destinations at campaign level, but it will not test a headline inside a single page.

Whichever tool you pick, keep the page fast. A testing script that slows the page can hurt the very conversion rate you are trying to improve. Keep it light and test on a real mid-range Android phone.

What mistakes ruin landing page A/B tests?

The usual mistakes are stopping early, changing several things at once and testing on too little traffic. Use this table as an audit before you trust a result:

Also look for practical significance. A tiny statistically significant gain may not justify the effort of changing and maintaining a page, while a large gain from a short test deserves a follow-up check before you scale spend.

How Arham Technology builds landing pages for testing

Our landing page development service is designed so that testing does not mean rebuilding the page. The pieces we typically provide are:

1. Offer and angle brief: we pull the winning ad claim, audience objection and primary action from your campaign notes.
2. Variant-ready sections: headline, proof, form and call-to-action blocks are built so they can be swapped without a rebuild.
3. Lightweight tracking: form, call and WhatsApp events are tagged for GA4, Google Ads and Meta, with test submissions verified before spend scales.
4. Fast delivery: clean markup and compressed media, checked on mid-range Android devices.

If you want the page and the ads planned together, our PPC and Google Ads service covers campaign structure so each ad group points to a page worth testing. For the structure to start from, read our guide to a high converting landing page.

What should you do next?

Pick your highest-spend campaign, write down its current conversion rate, and choose one hypothesis about the offer or headline. Work out the sample size you need, then confirm your tracking before changing anything.

If you would like help setting up a test-ready page, share your current ads and page on our contact page or explore landing page development and we will suggest the first test worth running.