Ecommerce CRO: the tests that actually move revenue

Introduction

Most conversion rate optimisation programmes fail for an unglamorous reason: the tests are too small to matter. Changing a button colour on a store doing a few thousand sessions a month will never reach significance, and even if it did, the effect would be noise. Real CRO works on the parts of the journey where money is actually lost. Here is where we look first.

1. Fix the mobile product page before anything else

For most consumer brands the majority of sessions are mobile, and the product page is where the buying decision happens. The recurring problems are consistent: the price and variant selector sit below the fold, image galleries do not communicate scale or texture, shipping and returns information is buried in an accordion nobody opens, and the add-to-cart button competes with a sticky promo bar. Getting price, variant, delivery promise, and a clear primary action into the first screen is often the single largest available win.

2. Answer objections in the order customers have them

Every category has a short list of questions that block purchase. Apparel: will it fit, and can I return it easily. Skincare: is it right for my skin, and what is actually in it. Furniture: will it fit my space, when does it arrive, and who carries it in. If your page answers these in the wrong order, or makes people hunt, you lose people who were ready to buy. Read your support inbox and reviews. The objections are written down for you already.

3. Treat delivery and returns as conversion features

Unexpected cost at checkout is one of the most consistently cited reasons for abandonment across the industry. The fix is not always free shipping. It is honesty earlier: show the delivery estimate and cost on the product page, show progress toward a free-shipping threshold in the cart, and state the returns window in plain language. Certainty converts.

4. Make search and filtering work

Visitors who use site search convert at meaningfully higher rates than those who do not, because they arrive with intent. If your search returns nothing for common misspellings, ignores synonyms your customers actually use, or your category filters do not match how people shop the category, you are turning your highest-intent traffic away at the door.

5. Reduce cart and checkout friction, not steps for their own sake

Forced account creation, surprise fields, aggressive upsells between the cart and payment, and a checkout that does not accept the wallets your customers use all leak revenue. Guest checkout, express wallet buttons above the fold, and address autocomplete are unglamorous and reliably positive.

6. Use social proof where the doubt is

Reviews at the bottom of a long page do less work than a rating near the price, a photo review beside the size guide, or a short line about how many people bought this month. Match the proof to the specific hesitation at that point in the page.

7. Measure like an adult

Decide the primary metric before the test. Revenue per session beats conversion rate, because it catches the change that lifts conversion while shrinking basket size. Run tests for full weekly cycles, do not peek and stop early, and be honest when a result is flat. A flat result that you accept is more valuable than a false win you build a strategy on.

Where to start if you have low traffic

If you cannot reach statistical significance, stop A/B testing and switch to evidence gathering: session recordings, on-site polls asking what almost stopped you buying, customer interviews, and a heuristic audit against the points above. Ship the obvious fixes, and save formal testing for when volume justifies it.

Takeaway

CRO is not a library of tricks. It is the discipline of finding where your specific customers hesitate and removing the reason. Start with mobile product pages, objections, and delivery clarity, and most stores find more upside there than in a year of button tests.