CRO checklist for online shops

If people cannot tell what you sell and who you sell it to, they leave before anything else you built gets a chance.

• A headline that names the product category in plain words
• A subline that answers what is in it for the visitor
• One image that supports the message instead of competing with it
• The main call to action reachable without scrolling

The reasons to buy here rather than at the next shop have to be visible before the first scroll, not on an about page.

• Short, concrete lines such as free shipping from a set amount, next day delivery, returns within a named period
• An icon per point so they can be scanned rather than read
• Placed next to the headline or the button, where the eye already is
• No vague claims like best quality, which every shop writes

Interest turns into revenue at the button. If a visitor has to hunt for it, most of them simply will not.

• High contrast against everything around it, not a third shade of your brand colour
• Text that names the action: add to cart, see the sizes, start the order
• Reassurance right underneath: free returns, delivery time, no account needed
• Repeated at every point where someone might be ready, not only once

Nobody believes what a shop says about itself. They believe what other customers said about it.

• Reviews with a name and a date, not anonymous five star blocks
• A star average near the top, the individual reviews further down
• Payment and delivery logos where they answer a doubt, not as decoration
• Any certification, award or membership you genuinely hold

A shop that looks abandoned feels risky to order from, even when everything else is in order.

• Recent orders, stock counters or a restock date where they are real
• A social feed only if you actually post regularly, otherwise it proves the opposite
• Fresh content: new arrivals, a seasonal block, a dated note
• Never invent scarcity. Customers check, and a fake counter costs more trust than it buys

Design is not decoration here. It decides in which order things get read and where the eye ends up.

• Clear sections with room between them instead of one continuous wall
• One dominant element per screen, not four competing for attention
• The same layout logic on every page, so the second visit is faster than the first
• A path that runs from the headline to the button without a detour

The category page is where most shops lose the sale, because it is treated as a list instead of a decision aid.

• A short intro that says what belongs in this category and what does not
• The most wanted products first, not whatever the database returns
• Sub categories visible as links, so a wrong turn costs one click
• Enough products per view that scrolling feels productive

Every question a tile leaves open costs a click, and a share of visitors will not spend it.

• A good image, with a second view on hover where it helps
• Name plus one line of description, so similar products can be told apart
• Price, and the struck through price when there is a discount
• Availability, and a button that either adds to the cart or opens the detail

Filters are the single biggest lever on a catalog with more than a screen of products, and they are usually generic.

• Filter by the attributes people actually decide on: price, brand, size, colour, availability
• Sorting by popularity, newest and price in both directions
• On mobile a slide in panel that is reachable with a thumb, not a desktop sidebar squeezed in
• Show how many products a filter leaves, and never let it lead to an empty page

Plenty of visitors arrive on a category page from search and never see the home page at all.

• A slim bar at the top with shipping, returns and delivery time
• Sticky on mobile if it does not eat the screen
• The same wording as on the home page, so it reads as a promise and not as a banner

A badge works because it is rare. Put one on every tile and it stops meaning anything.

• New, bestseller, last few in stock, a named discount
• The same colour and the same corner everywhere
• Only one badge per product
• Only where it is true, including the stock count

A ragged grid reads as carelessness, and carelessness on the catalog raises doubt about the packing station.

• The same tile size and image ratio for every product
• Backgrounds either all cut out or all photographed, not mixed
• Enough spacing that the eye can separate the tiles
• A grid that reflows properly on desktop, tablet and phone

Hidden shipping costs at the checkout are the most common reason a full cart is abandoned.

• The price large and unmissable, with the struck through price when there is a discount
• Shipping cost and delivery date visible on the product page, not first in the checkout
• Real urgency only: a genuine stock count, a real order cut off time
• Quantity discounts or tiered pricing where you offer them

On the product page every element competes with the one that earns money. That competition has to be settled by design.

• High contrast, large enough for a thumb, and visible without scrolling on mobile
• Microcopy underneath: ships today, free returns, delivery date
• A second button further down for long pages
• Nothing next to it that looks equally clickable

Doubt peaks at the button. Trust elements three screens further down arrive too late.

• Buyer protection, secure payment, the return period, directly under the button
• The star rating repeated near the price
• The payment methods you accept, as logos
• A phone number or chat for anything that needs advice

Online the pictures do the job the shelf does in a shop. Thin photography is the most expensive saving in e-commerce.

• Several angles, not one press shot
• Zoom or a full screen gallery for detail and material
• At least one image in use: worn, in a room, in a hand, for scale
• A short video or a rotating view for anything expensive or explanation heavy

A size chart is cheaper than a return, and a sold out variant that only reveals itself in the cart loses the whole order.

• Size and colour selectable by tapping, with the unavailable ones visibly out
• Availability per variant, stated on the variant itself
• A size chart, or fit guidance such as runs small, where clothing is involved
• The image switching with the colour, so the choice is confirmed

A shop wide rating helps a little. A rating on this product answers the question that is actually open.

• Written reviews with the star distribution above them
• Sorting and filtering, so a doubter can go to the one star reviews
• Customer photos where you can get them
• Answers from you under the critical ones, which reads better than only having good ones

Supplier titles are written for warehouses. Customers search and think in different words.

• Brand, product, and the variant that matters, in that order
• The words customers use, not the internal article name
• No article numbers or supplier codes in the visible title
• Short enough to survive being cut off in search results

Most descriptions list what the product is. Almost none say what it solves, which is what the visitor came for.

• Open with the problem it solves, in two sentences
• Bullet points for the benefits and the use
• Technical data at the end, in a table or an accordion
• Your own text, not the manufacturer's, which every competitor also has

Cross selling raises the basket when it completes the purchase and lowers conversion when it reopens the decision.

• Accessories that belong to this product, placed under the button
• Other colours and variants of the same product, close to the selection
• Alternatives only further down, once the decision has had its chance
• Never a carousel that moves on its own

Speed is a conversion factor and a ranking factor at once. On mobile connections it is usually the bigger of the two.

• One H1 per page, a sensible H2 and H3 structure underneath
• Its own title and meta description per category, not a generated pattern
• Images in WebP, sized correctly and lazily loaded
• Check the loading time on a phone connection, not on office wifi

Most shops get most of their traffic from phones and are still designed on a desktop screen.

• Tap targets large enough for a thumb, with space between them
• No horizontal scrolling anywhere, tables included
• Forms with the right keyboard per field and autofill switched on
• Walk through a full order on your own phone, once a quarter

The footer is where people look when they are checking whether a shop is real before they hand over a card number.

• Address and a phone number, not only a form
• Imprint, privacy policy, terms and the return policy
• Newsletter signup with a reason to sign up
• Social links opening in a new tab, plus quick navigation to the main categories

Colour is the fastest instruction on a page. It only works while it points at one thing.

• Pick one colour for add to cart and order, and use it nowhere else
• Every other button, catalog links included, gets a different colour
• Keep it consistent across the whole shop, including the checkout
• Check the contrast so the button stays readable for everyone

Mark's own shops sell better on prices ending in nine. It is cheap to try and it is a textbook case for a test rather than a belief.

• Round down to the next nine instead of up, so the price also drops
• Apply it per category, not to single products, otherwise the list looks random
• Watch the margin: a franc off on a thin product is a lot of margin
• Run it as an A/B test if you have the traffic to measure it

All three kinds work, in different situations, and the wrong one costs more attention than it returns.

• A chatbot catches standard questions around the clock and qualifies at night
• WhatsApp creates closeness and lets you follow up later, which suits after sales
• A human chat is worth it where advice decides the purchase
• The strongest setup combines them: bot first, handover when it matters, WhatsApp afterwards