Skip to content

eCommerce CRO Checklist: 90 Checks by Funnel Stage

90 checks you can test on your own store today, each one pass or fail in seconds. They come from the rule set behind our paid 277-point audit, grouped by the order a shopper meets them.

Free 24-point instant audit.

54

Example report

example-store.com

24 points checked · 9 fail

61

Product page

42

Checkout

58

Trust

49

Mobile

  • No shipping cost before checkoutCheckout
  • Add to cart is below the fold on phonesMobile
  • No reviews near the priceTrust

Seven areas, one report

  • Product pages
  • Checkout
  • Trust signals
  • Mobile
  • Page speed
  • Navigation
  • Email capture

By Mantas Karmaza, CRO specialist at MDigital · Updated

An eCommerce CRO checklist is a list of pass/fail checks for the pages a shopper uses to buy: navigation, product pages, cart, checkout, trust, mobile, speed and email. Walk your store as a new buyer, on a phone first, mark each check, then fix the failures closest to payment first.

How to run this checklist in under an hour

Most store owners audit their store from the admin panel or from a desktop they are logged in to. Your buyers see neither. Do the walk as a stranger.

  1. Minutes 0-5: set up. Write down your current conversion rate (the conversion rate calculator does the math). Pick three URLs: the homepage, your best-selling collection and your best-selling product.
  2. Minutes 5-35: phone first. Use a real phone in a private browser window, so you see the cookie banner and popups a new visitor sees. Go from the homepage to search, to the collection, to the product, add to cart, and go through checkout up to the payment button. Do not pay.
  3. Minutes 35-50: desktop and speed. Repeat the walk on a desktop for the checks that differ. Then run your three URLs through Google's PageSpeed Insights and read the mobile tab.
  4. Minutes 50-60: email. Subscribe with an address you have never used on your store. Start a checkout with a second fresh address and leave. Check both inboxes over the next 24 hours.

Record each check in a sheet with five columns: check, pass or fail, page, device, screenshot. Take a screenshot of every fail. A screenshot ends the argument with a developer or a co-founder faster than a description does.

How to decide what to fix first

You will fail more checks than you can fix this week. That is normal. Sort the fails with these four questions, in this order.

  1. Does it stop a purchase? A broken add-to-cart, a checkout error or a forced account. Fix these today, before anything else on this page.
  2. How close is it to payment? A shopper in checkout has already decided to buy. A leak there costs you more per visitor than a leak on the homepage. Work from checkout back: checkout, cart, product, collection, homepage.
  3. How many visitors meet it? A fail in the header or on your top product hits every session. A fail on a page with 40 visits a month can wait.
  4. How long does the fix take? Text, settings and app toggles take minutes. Theme changes take a developer. Do the quick ones in the same sitting.

Do not A/B test a bug fix. Ship it. Save tests for changes where you honestly do not know which version wins, and check first whether you have the traffic: the A/B test sample size calculator tells you how many visitors a test needs. Most small stores do not have enough for small changes.

1. Sitewide and navigation

Why it matters: if a shopper cannot find the product or the cart, nothing further down the funnel gets a chance.

  1. A search box or search icon sits in the header on every page. People who search already know what they want. Open three random pages and confirm it is in the same place.
  2. Search suggests products as you type. Type the first three letters of your best seller. Suggestions with a product image and price pass. Nothing until you press Enter fails.
  3. A search with no results offers a way forward. Search for a misspelling and for something you do not sell. The page should suggest other terms, categories or popular products, not stop at "0 results".
  4. The cart icon is in the header on every page and shows the item count. Add one item and check that the number changes without a page reload.
  5. Any product is 3 clicks or fewer from the homepage. Pick a product deep in your catalog and count the clicks through the menu. Four or more means the menu is too deep.
  6. Collection and product pages show breadcrumbs. Home > Category > Product, each part clickable, on mobile too. A shopper who lands on a product from Google uses them to see the rest of the range.
  7. The 404 page has search, top categories and a link home. Type your domain followed by /xyz. Old ad links and deleted products send real visitors here every day.
  8. The cookie banner closes with one tap and does not cover the buy button. Check it in a private window on a phone. Every new visitor meets this banner before anything else.
  9. All text meets WCAG AA contrast (4.5:1 for normal text). Light grey prices, small print and text on top of photos are the usual fails. A free contrast checker browser extension gives you the ratio.
  10. You can reach the menu, add to cart and the checkout with only the Tab key, and you can see where you are. Press Tab from the top of a product page. Focus must stay visible and must never get stuck inside a popup.

2. Homepage and landing pages

Why it matters: the first screen decides whether a new visitor stays long enough to see a product.

  1. The first screen says why to buy from you, not only what you sell. "Handmade leather bags, free returns for 60 days" passes. "Welcome to our store" fails. Check on a phone, without scrolling.
  2. The hero has one clear button above the fold. One primary action, such as "Shop the summer range". Three buttons of equal weight split the click.
  3. The button text matches the headline next to it. If the hero promotes the sale, the button says "Shop the sale" and opens the sale collection. A generic "Learn more" fails.
  4. Real customer proof is visible within the first two screen heights. A star rating with the review count, a review platform widget, or testimonials from real buyers. Invented reviews are a legal risk, not a tactic.
  5. Category tiles with images let a visitor narrow down before they browse. A wide catalog shown as one long product grid makes people scroll instead of choose.
  6. A running promotion is visible on the first screen or in the announcement bar. If you have a sale or a free-shipping threshold, say so where new traffic sees it. A promotion that lives only in your emails does nothing for ad traffic.
  7. A paid-ad landing page has one goal. Remove links and buttons that do not lead to that goal. Each extra exit is a place for the click you paid for to leave.

3. Category and collection pages

Why it matters: this is where shoppers compare. Every detail missing from a product card costs them a click into the product and a click back.

  1. Every product card shows the price. No "click for price", no price only on hover.
  2. Cards show the star rating and review count when the product has reviews. Shoppers use them to pick which three products to open.
  3. Sale items show the old price struck through, or the percentage saved, on the card. The saving must be real. A "was" price the product never sold at is a pricing complaint waiting to happen.
  4. Cards show color or size swatches when the product comes in variants. A shopper who wants the green one should not open six products to find it.
  5. Filters match how shoppers choose in that category. Size, color, price and material for apparel, not only "vendor" and "type". Apply two filters on a phone, open a product, press Back, and check that the filters are still on.
  6. Sort options include price (low to high) and best selling. These two are the sorts shoppers use most. "Alphabetical" alone does not help anyone.
  7. A filter combination with no results offers to clear the filters. A blank grid with no message reads as "this store has nothing for me".

4. Product pages

Why it matters: this is where the buying decision happens, and most fails here show in one screenshot.

  1. Price and add-to-cart are both visible without scrolling, on desktop and on a 390px-wide phone. A shopper should never have to scroll back up to remember the price before tapping buy.
  2. On mobile, a sticky add-to-cart bar appears once the main button scrolls out of view. Scroll down to the reviews. The buy action should still be one tap away.
  3. The add-to-cart is the strongest-colored button on the page. Blur your eyes or step back from the screen. If the first thing you see is a chat bubble or a wishlist icon, the button fails.
  4. The product has at least 4 photos. Different angles, the product in use, and a close-up of the detail that justifies the price. Photos replace the hand that would pick the product up in a shop.
  5. Photos zoom on desktop and swipe on mobile. Try to swipe the gallery with your thumb. If the page scrolls instead, it fails.
  6. A star rating with the review count sits next to the product title. "4.7 (212 reviews)". Stars without a count hide whether 3 people or 300 rated it.
  7. A short benefit summary sits near the title, and the description is in scannable bullets with benefits before specs. "Keeps coffee hot for 12 hours" before "double-wall 18/8 stainless steel". A wall of manufacturer text fails.
  8. Choosing a size or color updates the page without a full reload. The price, stock status and URL should change in place. A reload on every click loses people on slow phones.
  9. Sold-out sizes are greyed out or crossed out, not hidden. Hidden sizes make a shopper think you never stocked their size. Pair it with a "notify me when back in stock" form.
  10. Delivery cost, an estimated delivery date and any extra charges (duties, VAT, handling) show near the add-to-cart. "Free delivery over $50. Arrives in 3-5 business days." Baymard Institute's checkout research lists extra costs that appear late (shipping, taxes, fees) as the top reason shoppers abandon, so do not leave them for checkout.
  11. A one-line returns promise next to the button links to the full policy. "Free 30-day returns" answers the last doubt at the moment of the click.
  12. Apparel and footwear have a size guide link next to the size selector. Not in the footer, not in a separate tab. Next to the sizes.
  13. Adding to cart gives clear feedback, such as a slide-out cart or a confirmation message. Tap the button. If nothing visible happens, shoppers tap again and end up with two.
  14. Any stock count or countdown is real. "Only 3 left" must match your inventory, and a timer must end when the offer ends. Reload the page: a timer that resets fails, and shoppers do notice.

The free store audit checks 24 of the most important points from this list in under 2 minutes, with no sign-up required. Use it to see where you stand before you spend the hour.

5. Cart

Why it matters: the shopper has chosen. The cart's only job is to confirm the choice and move them to checkout without a surprise.

  1. Each line shows the image, name, chosen variant, quantity and price. A shopper who sees "Shirt x1" and not "Shirt, Blue, M" goes back to check.
  2. Quantity changes with plus and minus buttons, and the total updates. A tiny number field that needs the keyboard is a fail on mobile.
  3. The checkout button is the most prominent thing in the cart, and its label says what it does. "Checkout" or "Secure checkout". Not "Continue", which could mean anything.
  4. The subtotal sits right next to the checkout button. The shopper checks the number and clicks in one look.
  5. Shipping cost, or an estimate, shows in the cart. "Shipping calculated at checkout" is the line that starts the surprise this page is trying to remove.
  6. Express payment buttons (Shop Pay, Apple Pay, Google Pay, PayPal) appear in the cart. Returning buyers skip the whole form. Check that they show on a phone, not only on desktop.
  7. If you offer free shipping over a threshold, the cart shows how far away the shopper is. "Add $12 for free shipping". Only run a threshold your margins can afford. A threshold that loses money on every order is worse than none.
  8. The return policy is one tap away from the cart. A short line under the button is enough.

6. Checkout

Why it matters: Baymard Institute puts average cart abandonment at about 70%, and most checkout leaks are friction the store added itself.

  1. Guests can buy without creating an account. Offer the account on the thank-you page, after the money has moved.
  2. The checkout has 8 or fewer visible form fields. Count them. Every optional field ("Company", "Address line 3") is a reason to stop and think.
  3. Browser autofill fills the form, and the address field suggests addresses. Tap the first field on your phone and accept the autofill. If half the fields stay empty, the field attributes are wrong.
  4. The phone field opens a number keypad and the email field opens the email keyboard. Thirty-second test on a real phone.
  5. Every field has a visible label that stays when you type. A placeholder that disappears on the first keystroke leaves the shopper guessing what the field was.
  6. Errors appear next to the field, in plain words, and the form keeps what was typed. Enter a broken email and submit. A red banner at the top or a cleared form fails.
  7. The total on the payment step matches the total the shopper saw before it. No new fees at the last step. This is the single most expensive fail on this list.
  8. The order summary (items, quantities, total) is visible at every step. On mobile it can be collapsed, but one tap must open it.
  9. At least 3 payment methods are on offer. For example card, PayPal and a wallet such as Apple Pay. Check what your main market expects, such as Klarna in parts of Europe.
  10. The discount code field is a small link, not a large empty box. A big empty box sends shoppers off to search for codes, and some do not come back.
  11. The checkout hides the main menu and footer links. The logo can stay. Everything else that leads away from payment goes.

7. Trust and policies

Why it matters: a first-time visitor knows nothing about you. These checks answer "is this store real, and what happens if it goes wrong?"

  1. Every page loads on HTTPS with no browser warning. Check the padlock on the homepage, a product page and the checkout. Mixed-content warnings count as a fail.
  2. The return policy is linked from the footer, product pages and checkout, and states the return window in plain words. "30 days, free return label" beats two screens of legal text.
  3. The footer links to contact, FAQ and order tracking. The contact page has an email address or phone number, not only a form.
  4. A physical address or company registration is visible. In the footer or on the contact page. Shoppers check this on stores they have never heard of.
  5. The About page shows the real people or story behind the store. Stock photos and "we are passionate about quality" fail.
  6. Reviews show that the reviewer bought the product. A "verified buyer" mark separates real reviews from ones anyone could type.
  7. Reviews come through a recognized review platform, not only a self-hosted widget. Shoppers trust a third party more than a store that grades itself.
  8. Privacy policy and terms of service are linked in the footer. Shoppers who are about to hand over an address and card number look for them.

8. Mobile

Why it matters: for most stores the majority of visits come from phones. Test on a real phone, not a narrowed desktop window.

  1. No page scrolls sideways. Swipe left on the homepage, a product page and the checkout. A table or image wider than the screen is the usual cause.
  2. Body text is at least 16px. If you need to zoom in to read the product description, it fails.
  3. Buttons and links are at least 48x48px. That is Google's recommendation. Size buttons and the X that closes a popup are the usual fails.
  4. Main buttons are full width and sit where the thumb reaches. Hold the phone in one hand and tap add-to-cart and checkout with your thumb.
  5. No full-screen popup covers the page on arrival. Google's guidance warns against intrusive interstitials on mobile. Let people see the page before you ask for an email.
  6. The mobile menu is grouped into clear categories. One long list of every link fails.
  7. Search is visible on mobile, not two taps deep inside the menu. Results must fit the screen, with images and prices.

9. Speed

Why it matters: every check above needs the page to load first. Run your homepage, top collection and top product through PageSpeed Insights and read the mobile results.

  1. Largest Contentful Paint is under 2.5 seconds. This is Google's "good" threshold. On a product page the main photo is usually the largest element.
  2. Interaction to Next Paint is under 200ms. Tap a variant or the menu. A delay you can feel means too much JavaScript.
  3. Cumulative Layout Shift is under 0.1. If the add-to-cart jumps down as a banner or review widget loads, a shopper taps the wrong thing.
  4. Time to first byte is under 800ms. Slower than that is the server or the hosting plan, not your images.
  5. Total page weight is under 3MB. Your browser's network panel shows it at the bottom.
  6. Images are WebP or AVIF and sized for how they display. A 4000px-wide photo shown in an 800px box has 25 times the pixels it needs.
  7. Images below the first screen are lazy-loaded, but the main product photo is not. Lazy-loading the main photo delays the very thing the shopper came to see.
  8. The page loads fewer than 10 third-party scripts. Apps you uninstalled often leave their scripts behind. See how app bloat slows a Shopify store for how to find them.

10. Email capture and post-purchase

Why it matters: most first visits do not end in a sale. Email brings people back, and the thank-you page is the cheapest moment to earn the next order.

  1. An email signup form is in the footer of every page. It catches the people who close the popup and keep reading.
  2. The signup popup appears after the visitor engages, not on page load, and does not come back after it is closed. A delay, a scroll depth or exit intent all work. On every page view, it fails.
  3. The signup offers a concrete reason. A first-order discount, free shipping or early access. "Subscribe to our newsletter" gives no reason to type an email.
  4. A welcome email arrives within 1 hour of signing up. Test with an address you have never used. Check the spam folder too.
  5. An abandoned cart email arrives within 24 hours. Enter a fresh email at checkout and leave. No email means the flow is off, broken or blocked.
  6. SPF, DKIM and DMARC are set up and aligned for your sending domain. Google and Yahoo have required this from bulk senders since 2024. Without it, the two emails above may go to spam. The email deliverability audit checks all three.
  7. Your sending domain is not on an email blacklist. Free lookup tools check the major lists, such as Spamhaus.
  8. The thank-you page confirms the order clearly and shows what was bought. "Thank you, your order #1042 is placed" with the items and total. An unclear page leads to duplicate orders and support emails.
  9. The thank-you page says when the order will arrive and how tracking will reach the buyer. "You will get a tracking link by email when it ships, usually within 2 days."
  10. The thank-you page or confirmation email gives a reason to come back. A code for the next order, for example. This is the moment the buyer trusts you most.

What this checklist does not cover

These checks are the ones a person can verify by hand in about an hour. Our full audit checks 277 points. The rest include checks that need tools, such as accessibility markup (ARIA roles, heading order), reading your Google Analytics 4 and Microsoft Clarity data when you connect them, and scoring screenshots of your real pages on desktop and mobile.

A checklist also cannot tell you which fail costs you the most. Two stores can fail the same check, and for one of them it is the main leak while for the other it barely matters. That depends on your traffic, your products and your price point. For a deeper walk-through of the process, see the guide to a 30-minute eCommerce CRO audit and how to audit an online store.

The shortcut: run the checklist automatically

If you would rather not spend the hour, the free audit checks 24 essential points in under 2 minutes and gives you a score and a summary of what fails. No sign-up required.

The full CRO audit checks all 277 points for $14.95, with step-by-step fix instructions for your platform, revenue impact estimates and a PDF report. On Shopify, see what the Shopify CRO audit checks. All options are on the pricing page.

Frequently asked questions

How many items should an eCommerce CRO checklist have?
This page has 90, grouped into 10 funnel stages, because that is what one person can test by hand in about an hour. Our full paid audit checks 277 points and the free audit checks 24. More items are not better if nobody finishes the list. Start with the stages closest to payment: checkout, cart and product pages.
How often should I run through the checklist?
Run it once in full, then again after any theme change, new app, checkout change or big promotion. Apps and theme updates break things quietly. Re-test the speed and email checks monthly, because those drift without anyone touching the store.
Can I download this checklist as a PDF or spreadsheet?
There is no separate download. Copy the checks from this page into a sheet with columns for pass or fail, page, device and a screenshot link. If you want a PDF of your own store's results, the paid audit ($14.95 per audit) includes a PDF report download.
Should I run the checklist on mobile or desktop first?
Mobile first, on a real phone in a private browser window. For most stores most visits come from phones, and many fails (sticky add-to-cart, keyboard types, tap sizes, popups) only show there. Then repeat the walk on desktop for the checks that differ.
Where do the checklist items come from?
Every item is a rule from the rule set behind our paid audit, which was built from more than 10 years of hands-on CRO work on online stores plus published usability research such as Baymard Institute's checkout studies and Google's Core Web Vitals thresholds. We rewrote each rule as a check you can verify by hand.
What should I fix first if my store fails a lot of checks?
Fix anything that stops a purchase today, such as a broken add-to-cart or a forced account. Then work back from payment: checkout, cart, product pages, collections, homepage. Within each stage, fix what most visitors see first and what takes minutes before what needs a developer.

See what your store is missing

The free audit checks 24 conversion points on your real store pages and shows which ones fail.

Free 24-point instant audit.