How to increase ecommerce conversion rate: 10 tips from our A/B tests

Most lists of conversion tips are a pile of ideas with no order. This one is the order I work in. The first 4 tips decide what to test. The next 4 are changes that have moved conversion in our A/B tests. The last 2 keep you from fooling yourself with the results.

1. Start with the step that loses the most shoppers

I look at the conversion rate of each funnel step and attack the step with the highest drop-off. If no step stands out, I start with the landing pages that get the most traffic. That is where a small lift adds up fastest. Our guide on How to do a CRO audit walks through how to find that step.

2. Use analytics for what, and usability testing for why

Analytics shows what visitors did. Usability testing shows why they did it. You need both. A page with a high exit rate tells you where to look, but only watching real people tells you what confused them.

3. Rank your ideas before you build anything

I score every idea on PIE: Potential, Importance and Effort, each from 1 to 10. The highest scores go first. It keeps the team from testing whatever was mentioned last in a meeting. How to build a CRO strategy covers the full process.

4. Write a hypothesis you can prove wrong

Every test gets 1 sentence: “By [change X], we will [affect Y], resulting in [Z].” If you cannot fill in all 3 parts, the idea is not ready to test.

5. A yellow add to cart button with the price inside lifted conversion 13.9%

On a set of desktop product pages, the add to cart button did not stand out. We made it yellow and put the price inside it. Conversion rose 13.9% and revenue per visitor rose 16.9% over 24 days.

The takeaway is not “use yellow.” The button had to stand out from everything around it, and the price inside it answered a question shoppers had at the moment they were deciding.

6. Clearer homepage category buttons lifted conversion 27.3%

A desktop homepage opened with an animated banner and small category buttons. We stopped the animation and made the buttons larger and clearer. Conversion rose 27.3%, and we implemented it.

Shoppers who land on a homepage are trying to get somewhere. Make the paths to your main categories the easiest thing on the page to see and tap.

7. A cart timer lifted mobile checkout conversion 8.0%

Our mobile checkout gave shoppers no reason to finish now. We added a countdown timer that showed how long their cart was held. Conversion rose 8.0% in 18 days, and we implemented it.

Urgency only helps when it is honest. A timer that holds a cart is a real benefit. A fake deadline is not, and shoppers can tell.

8. Show the full cost before checkout

Extra costs are the top reason shoppers leave checkout. Show shipping and fees on the product page or in the cart, so the total at checkout is no surprise.

9. Run every test for at least 2 full weeks

I want 2 weeks of data before I call a test. That covers both weekends and smooths out paydays and email sends. During peak weeks like Black Friday, I watch results several times a day but never make a call on less than a full day of data. Peak results get retested once traffic returns to normal. Our experimentation and testing service runs every test this way.

10. Read the signal, then plan the next test

A result is more than win or lose. I look at how much the tested element moved the number. A big move, in either direction, means shoppers care about that element. The next test usually tries 2 different directions on it. A small move means the element does not matter much, so I move on.

The best tip is the right order: find the leak, learn why, then test.

Frequently asked questions

What is a good ecommerce conversion rate?

It depends on your category, price point and traffic mix. Compare your rate to your own past results, not a broad industry average.

How can I increase my ecommerce conversion rate fast?

Find the funnel step with the biggest drop-off and test fixes there first. Small gains on a high-traffic step add up fastest.

What should I A/B test first on an ecommerce site?

Start with the step that loses the most shoppers. If no step stands out, start with your top landing pages.

How long should an ecommerce A/B test run?

At least 2 full weeks, so the data covers both weekends. Results from peak weeks should be retested later.

Do countdown timers increase conversion?

They can when they are honest. Our mobile cart timer lifted conversion 8.0%, but a fake deadline can cost you trust.