Ecommerce conversion rate optimization is the practice of raising how much revenue or profit each store visitor produces, by finding where shoppers drop off between the collection page and the order confirmation and testing fixes there. It works best when you measure revenue per visitor instead of conversion rate alone, base test ideas on research, and run controlled A/B tests. The tests in our database show that the same change can win on one store and lose on another, so testing beats copying.
This guide covers what makes ecommerce CRO different, how to work through the funnel, which research inputs to use, what our database of 285 real A/B test results says works, and a 90-day plan you can follow.
How ecommerce CRO differs: RPV, AOV and profit per visitor
On a lead-gen site, a conversion is a form fill and every conversion is worth roughly the same. In a store, two orders can be worth $20 and $400. That is why ecommerce teams track three numbers together.
- Conversion rate (CVR): orders divided by sessions or visitors.
- Average order value (AOV): revenue divided by orders.
- Revenue per visitor (RPV): revenue divided by visitors, which equals CVR times AOV.
RPV is the safer primary metric because it catches changes that raise conversion rate by cutting order size. Profit per visitor goes one step further and subtracts discounts, shipping costs and cost of goods.
Shipping tests show why this matters. Dave Diederen raised a free shipping threshold to $165 and saw +6.27% profit per visitor. A higher threshold can look bad on conversion rate while it earns the store more money.
The opposite can happen too. When Dave Diederen introduced subscriptions with the subscription option pre-checked, the result was -9.17% RPV upfront and a +1,295% share of subscription orders. Whether that is a win depends on how long subscribers stay, which a two-week RPV readout cannot show.
To see where your store stands, compare it with our ecommerce conversion rate benchmarks.
The funnel: collection, product, cart, checkout
Most store sessions follow the same path: a shopper lands on a collection or product page, adds to cart, opens the cart and goes through checkout. Each step loses people, and each step has its own typical problems.
Collection pages
The job of a collection page is to get a shopper onto the right product page or into the cart fast. Common fixes are showing products above the fold, adding best-seller labels and letting shoppers add to cart from the grid.
Oliver Kenyon added Quick Add buttons to collection page product cards and reported +$135,307/month. Our full test database has dozens more collection page results.
Product pages
The product page is where the buying decision happens, so it is also the most tested page type in our database: 128 of our 285 results. Tests usually focus on the title, the first image, the buy box, reviews and guarantees.
Dave Diederen changed a product title to a benefit-led "Achieve [outcome] with [product]" format and saw +15.61% CVR. Browse more in our product page tests.
Cart
The cart is where shoppers check the total and decide whether to trust the store with their card. Shipping, returns and payment information near the checkout button often matter more than design.
Oliver Kenyon added a dispatch, returns and payment block under the checkout button in a cart drawer, plus a size dropdown, and saw +23.8% revenue per visitor. Removing the cart is not a reliable shortcut: when Dave Diederen skipped the cart and sent shoppers straight to checkout, the result was -10.23% CVR (a loss). See all cart tests.
Checkout
Checkout is where cost surprises hit. Baymard Institute puts the average documented cart abandonment rate at 70.22%, based on 50 studies. In the same research, 40% of US shoppers who abandoned for a reason other than "just browsing" said extra costs such as shipping, tax and fees were too high.
On many platforms you can only change parts of the checkout, so tests here often target trust messages, payment options, urgency and upsells.
Research inputs: analytics, heatmaps and surveys
Good test ideas come from evidence about your own shoppers, not from lists of tactics. Three sources cover most of what a store needs.
Analytics
Analytics shows where people leave. In Google Analytics 4, funnel exploration lets you build up to 10 steps, choose open or closed funnels and break the results down by device. Build one for collection, product, add to cart, begin checkout and purchase, then split it by mobile and desktop.
The step with the biggest drop on your highest-traffic device is usually the best place to start.
Heatmaps and session recordings
Heatmaps show where attention goes on a page, and recordings show where shoppers hesitate. Microsoft Clarity offers both for free and says it has no traffic limits.
Dave Diederen used heatmap data to change above-the-fold USPs to the points buyers were engaging with and saw +6.23% CVR.
Surveys
Surveys tell you why. Post-purchase surveys ask buyers what almost stopped them, and on-site polls ask browsers what is missing. Use the answers as copy, word for word if you can.
Oliver Kenyon added three survey-based reassurance lines above a collection page grid and reported $44,890/month.
Want results like this on your store?
Oliver Kenyon at Conversionwise. Oliver co-founded Conversionwise and has spent 13 years helping ecommerce brands turn clicks into customers.
What the database says works
We have 285 ecommerce A/B test results from CRO practitioners. By page type, 128 are product page tests, 28 are collection page tests, 27 homepage, 23 site-wide, 19 checkout, 19 landing page, 18 cart and 14 navigation. Here are the patterns that show up most.
Trust and cost clarity near the buy button
Shipping dates, return policies, guarantees and payment icons placed right next to the add-to-cart or checkout button win again and again. This lines up with Baymard's finding that extra costs and delivery time are leading reasons people abandon.
Clear benefit copy
Benefit-led product titles and headlines are among the most consistent winners in the database. They work because a shopper coming from an ad needs to know in a few seconds what the product does for them.
Faster paths to the product
Quick add buttons, tighter collection page headers and better navigation menus help shoppers reach products sooner. Most of the navigation results in our database come from Oliver Kenyon's mega menu and mobile menu tests.
Urgency depends on context
Countdown timers are the clearest warning against copying. Dave Diederen has a checkout timer test that won and another that lost, and a timer for a European market that lost badly. Kasper Dolk removed a countdown timer from a checkout and saw CVR +25.61% and RPV +26.18%, an estimated +€150k/month.
The same idea can help or hurt depending on the audience, the market and how honest the urgency is. That is the case for testing on your own store.
Want results like this on your store?
Kasper Dolk at Maison Commerce. Kasper runs Maison Commerce, which builds CRO programs, landing pages and themes for fast-growing ecommerce brands such as ICON and HiStrips.
Losses are part of the data
Our database includes losing tests on purpose. A loss tells you what not to ship and stops you from copying a tactic that only worked on someone else's store. For a broader approach to growth, read our guide on how to increase conversion rate.
A 90-day ecommerce CRO plan
This plan fits a store with enough traffic to run at least one A/B test a month. Stores with less traffic can follow the same steps but should make bigger changes and run tests longer.
Days 1-30: measure and research
- Check that purchase, add-to-cart and checkout events track correctly.
- Pick a primary metric. Use RPV, or profit per visitor if discounts and shipping vary a lot.
- Build a funnel report by device and find the biggest drop.
- Install heatmaps and session recordings on your top collection and product pages.
- Launch a post-purchase survey asking what almost stopped people from buying.
Days 31-60: first tests
- List 15-20 ideas, each tied to a piece of research.
- Score them on expected impact, confidence and effort, and pick the top two or three.
- Start with the page that has the most traffic and the biggest drop, often the product page on mobile.
- Decide the sample size and run length before launch, and do not stop a test early because it looks good.
Days 61-90: learn and scale
- Ship the winners and write down why each test won or lost.
- Run follow-up tests on the same idea in other places, such as moving a winning product page message into the cart.
- Test the cart and checkout once the product page work is underway.
- Review the backlog every month and add new survey and recording insights.
If you would rather hand this work to a specialist, our CRO experts directory lists the practitioners behind the tests in this guide.
Want results like this on your store?
Dave Diederen at Syntra Labs. Dave builds and tests Shopify stores at Syntra Labs, and shares most results in public.
Questions
What is ecommerce conversion rate optimization?
Ecommerce conversion rate optimization is the process of raising the revenue or profit each store visitor produces. You research where shoppers drop off between the collection page and checkout, then A/B test changes to fix those problems.
Which metric should an ecommerce store optimize for?
Revenue per visitor is usually the best primary metric because it combines conversion rate and average order value. If discounts or shipping costs vary a lot between variants, profit per visitor is even better.
Where should ecommerce CRO start?
Start where your funnel loses the most shoppers on your highest-traffic device. For many stores that is the mobile product page or the cart. A GA4 funnel exploration split by device will show you the biggest drop.
Do CRO tactics like countdown timers work for every store?
No. Our database has countdown timer tests that won and others that lost, including a timer for a European market at -35.61% RPV and a checkout where removing the timer raised RPV by 26.18%. Test tactics on your own store before rolling them out.
How long does ecommerce CRO take to show results?
A focused program can run its first tests within 30 to 60 days. Plan for about 90 days to set up tracking, do research, run a few tests and ship the first winners.