Ecommerce conversion rate optimization: a hands-on course in 8 lessons

Ecommerce conversion rate optimization (CRO) is the practice of increasing the share of visitors who place an order, without buying more traffic. You measure where shoppers drop out, remove the friction at that step, and check whether the change worked. This course teaches the method in eight lessons, each with a short answer, a task for the week and an exercise, written for stores selling to shoppers in the United States and the United Kingdom, where expectations around guest checkout, express wallets, shipping transparency and returns are high.

Lesson 1: What is conversion rate and how do I calculate it?

Conversion rate is the number of orders divided by the number of sessions in the same period, shown as a percentage. The formula is simple; the mistake is to use only one number.

A single storewide conversion rate misleads because it averages very different visitors: mobile and desktop shoppers, returning email subscribers and first-time visitors from a social ad. One product with a broken size chart can drag the whole average down. The storewide number says something changed; the segmented numbers say where.

What to do this week

1. Write down the storewide conversion rate for the last four full weeks.
2. Split it by device, traffic channel and landing page, and look separately at your ten most visited product pages.
3. Decide what you count (sessions or unique visitors, cancelled orders in or out) and keep that definition fixed; ad blockers and consent choices make every conversion rate an estimate, so compare only numbers counted the same way.

Exercise

Find the segment with the biggest gap between traffic and orders, often mobile sessions from paid social. Write it at the top of a page titled "CRO notes"; the rest of this course works on that page.

Lesson 2: Where do visitors drop? Mapping the funnel

The ecommerce funnel is the sequence of steps from arriving on the site to paying: landing page, category or search, product page, cart, checkout and payment. The weakest step is the one where the largest share of people who reached it do not continue. Fix that step first, because the improvement helps everyone who gets there.

What to do this week

Build the funnel for the segment you chose in Lesson 1. In Shopiyz, the analytics dashboard shows the conversion funnel from visitor to customer and splits it by device. Find the step that is clearly worse than its neighbors.

Exercise

List three possible reasons shoppers leave at the weakest step, then go through that step yourself on a phone, on mobile data, as a new visitor. Your own walk-through is free and usually enough to start.

Lesson 3: Product page fixes that move the needle

A product page converts when it answers "Is this the right product for me, and is it safe to buy here?" before the shopper scrolls away. The fixes that matter most remove an unanswered question: unclear photos, hidden total cost, missing delivery information, no size or compatibility guidance, and a mobile layout that buries the add-to-cart button. The parts themselves are covered in the seven parts of a product page that sells; this lesson applies that list to CRO.

What to do this week

Audit the ten product pages with the most traffic and the fewest orders, and fix them in order of traffic:

Exercise

For each of the ten pages, complete the sentence "A shopper leaves this page because…". If you cannot finish it, the problem is probably elsewhere.

Lesson 4: Search and navigation

Shoppers who use site search or filters have told you what they want, so a failure here loses your most motivated visitors. Fix two things: zero-result searches, where a query returns nothing, and filters that hide products or use words customers do not. The detail is in how to fix zero-result searches.

What to do this week

1. Export last month's search log and look at the top queries that returned nothing, and those that returned results but no clicks.
2. Add synonyms and spelling variants, including US and UK vocabulary ("sneakers" and "trainers", "sweater" and "jumper"), and make the zero-result page suggest related categories.
3. Check category filters on a phone; group long option lists and show the most used options first.

Shopiyz AI Search understands natural phrasing and misspellings and logs zero-result queries for you.

Exercise

Search your store for your three bestsellers using a customer's words, then search for something you do not sell. The first test shows whether search works; the second shows whether the zero-result page keeps the shopper.

Lesson 5: Cart and checkout

Cart and checkout convert when they are predictable: the total cost is known before the shopper types anything, there is a guest option, the payment methods are ones the shopper already uses, and every error says what to change. In the US and UK, shoppers expect Apple Pay and PayPal, a visible shipping cost and no forced account creation.

What to do this week

1. Place a guest test order on a phone and note every moment you were unsure what would happen next.
2. Show the order total, with shipping and tax, in the cart before any form; the shipping guide covers turning shipping rules into a clear promise.
3. Enable express wallets, and replace generic error messages with specific ones next to the field.
4. Set up abandoned cart recovery for shoppers who reach checkout and leave anyway.

Exercise

Ask someone who has never used your store to buy something on their own phone while you watch without helping. Every hesitation goes on your backlog.

Lesson 6: Trust: policies, contact, reviews and security signals

Trust converts when a first-time visitor can answer, within seconds, "Is this a real business, and what happens if something goes wrong?" The answers are a clear return policy, a way to reach a human, genuine reviews, recognizable payment logos and a secure connection. Trust is the absence of unanswered doubts.

What to do this week

• Return policy: plain language, linked beside the add-to-cart button and in checkout. Reducing ecommerce returns shows how a clear policy lowers returns and raises conversion.
• Contact: an email address, a contact form, a phone or chat with real hours, and a business address.
• Reviews: collect them automatically after delivery and show them, including critical ones with your reply.
• Payment and security: accepted payment methods in the footer and checkout, and HTTPS on the whole site.

Exercise

Open your store on a phone in a private window and give yourself ten seconds to find the return policy, a way to contact a human, and the delivery cost for a typical order. Anything that took longer is your first trust fix.

Lesson 7: How do I run an A/B test correctly?

An A/B test shows two versions of a page or message to two randomly split groups of visitors at the same time and compares one success metric chosen before the test starts. A test is valid only with one hypothesis, one change, enough traffic and time to separate a real difference from random noise, and a stopping rule set in advance.

The honest part of this lesson: most small stores do not have enough traffic for a valid A/B test on page layouts. With a few hundred sessions a week, the difference between two versions is almost always within random variation, so the test either runs for months or gives a confident-looking answer that is wrong. Fix obvious problems directly instead and measure the funnel before and after.

If you do test, the rules are short: write the hypothesis down, change one thing, decide the metric and sample size before launch, run whole weeks including weekends, and decide on the planned end date. Peeking and stopping the first day one version pulls ahead is the most common mistake; let the test finish.

Just fix it when the problem is clear and the fix is standard: hidden shipping cost, no guest checkout, a broken size chart. Test it when two reasonable options exist: popup wording, cart reminder timing, free shipping threshold versus flat rate, email subject lines. Messages reach many people quickly, so even a mid-sized store gets a usable answer.

What to do this week

If you have the traffic, write one hypothesis as "If we change X, metric Y will improve, because Z", set the metric and end date, and launch. If you do not, take the top item from your "just fix it" list, note today's funnel numbers, make the change, and compare two weeks later.

Shopiyz lets you A/B test the message in a smart popup and in automated email, SMS and WhatsApp flows. Page-level layout testing is not part of the platform; third-party tools exist for that.

Exercise

Rewrite any idea from your CRO notes as a hypothesis with a single metric. If you cannot name the metric, it is not ready to test; if the metric will obviously improve, skip the test and ship it.

Lesson 8: Prioritization, a weekly routine and common mistakes

Prioritize CRO work by scoring each idea on impact (how much it could move the weakest step) and ease (how quickly you can ship it), and work from the top of the combined score. This stops a small team from spending a month on a homepage redesign while mobile checkout loses shoppers every day.

The impact × ease score. Give every idea one to five for impact and one to five for ease, then multiply. Impact is highest for ideas that touch the weakest step on the most visited pages; ease is highest for changes made in settings or copy. Re-score monthly, because the weakest step moves.

A weekly routine. Monday: read last week's funnel per device and channel. Tuesday: read the search log and return reasons, score new ideas. Midweek: ship the top-scored item. Friday: log what changed and when.

Common mistakes. Optimizing the homepage first, when most shoppers land on product pages. Changing five things at once, so a win cannot be repeated. Adding urgency instead of removing doubt. Forgetting the traffic source: if an ad promises a discount the landing page does not mention, no page change fixes the drop.

A hypothetical example: a small UK homeware store

This is a hypothetical example with no real figures. Imagine a UK store selling ceramics and kitchen textiles, run by two people, with most traffic landing on product pages from organic search and social ads, mostly on mobile.

Measure. Splitting conversion rate by device shows mobile converting far less often than desktop, and the mobile funnel's biggest drop is between cart and checkout.

Diagnose. A phone walk-through finds that shipping cost appears only after the address form, checkout demands an account, and the only payment option is a card form.

Fix, scored. Guest checkout and express wallets score highest on both impact and ease, because they are settings. Estimated shipping in the cart is next, then a return policy link beside the add-to-cart button. The planned homepage redesign scores low and is parked.

Verify and repeat. The owners note the funnel numbers, ship the changes over two weeks and compare the mobile funnel a month later. No A/B test is run, because the traffic could not reach a usable sample and the problems were not in doubt; they do test two versions of a cart recovery popup message, since the platform splits that traffic. The weakest step then becomes category to product, and next month's work is search synonyms.

Conversion rate optimization with Shopiyz

Shopiyz puts the measurement and the fixes in one place. The real-time analytics dashboard shows a conversion funnel from visitor to customer and device analysis, and site search analytics show which queries return nothing and whether a search leads to an order.

Checkout supports more than thirty payment methods, including Apple Pay, PayPal, Stripe and Klarna for UK shoppers. Smart popups offer exit-intent and cart recovery messages with A/B testing, workflows run automated email, SMS and WhatsApp flows with A/B testing, and reviews are requested automatically after every order.

Explore the feature list, compare plans, or contact the team for help reading your funnel.