On this page

Good ecommerce website design is mostly about five pages: the category list, the product page, the cart, the checkout and what happens after the order. Large usability benchmarks show that most leading stores are still mediocre on all of them. Fix hidden costs and forced accounts at checkout first, then product page information, then filters, then speed.

Why does store design matter more each year?

Online shopping keeps taking share. US retail ecommerce sales reached $340.2 billion in the second quarter of 2026. That was 17.1% of all retail sales, and 12.2% more than a year earlier, while total retail grew 6.7%.1

Yet the big stores are not good at it. Baymard Institute scores hundreds of leading US and European sites by hand against its usability guidelines. Its latest benchmarks rate most of them mediocre or worse on the pages that sell.234

Share of leading stores rated mediocre or worse

Desktop sitesMobile sitesProduct lists and filtering (2025)58%78%Product page (2026)52%62%Checkout (2025)64%63%
Source: Baymard Institute benchmarks of leading US and European stores (research vendor data). Product page figures are the share not rated decent or better.

That gap is your opening. A small store can't outspend a big one on ads. It can have a clearer product page and a shorter checkout.

Which pages does an online store need?

Every store has the same spine. A shopper lands, narrows down, decides, pays and waits. Each page has one job, and each has its own article here.

The store, page by page

Landingpagekeep thead's promiseCategorynarrow itdownProductpageanswer everyquestionCartshow thefull costCheckoutnosurprisesConfirmationreassureNextorderbring themback
Guides: landing pages, category pages, product pages, cart and checkout, after the order.

The homepage matters less than most owners think. Many shoppers arrive straight on a product or category page from search, ads or a link.

What does the research say about category pages?

Finding the right product is the first place stores lose people. In Baymard's 2025 benchmark, 58% of desktop sites and 78% of mobile sites were poor to mediocre at product lists and filtering.2

The cost is large. In Baymard's testing, sites with mediocre list tools saw 67% to 90% of users abandon while looking for a product. Sites with only slightly better tools saw 17% to 33%.5

The fixes are mostly about data. Each category needs filters that match how people choose in it, such as size and fit for shoes or wattage for heaters. The category page guide covers filters, sorting, loading more products and the SEO side.

What does the research say about product pages?

The product page decides the sale. In Baymard's 2026 benchmark, only 48% of leading desktop sites had a decent or good product page. On mobile it was 38%.3

The common misses are simple. Product photos without anything to show scale. Delivery cost hidden until checkout. Sizes in a dropdown instead of buttons. Reviews tucked away.

Reviews carry a lot of weight. A Northwestern University study found that a product with five reviews was 270% more likely to be bought than one with none.6 The product page guide covers each element, and the free product page audit scores yours.

What does the research say about checkout?

Checkout is where the order is won or lost. Baymard averages 50 studies and puts cart abandonment at 70.22%.7

Some of that can't be avoided. In Baymard's survey, 42% of US shoppers had left a cart because they were just browsing. The rest is mostly design. Among the other reasons, 40% left over extra costs and 18% because the site wanted an account.7

In its 2025 benchmark, 64% of desktop sites had a mediocre or worse checkout, and only 2% rated good.4 Baymard estimates that a large store can lift conversion by 35.26% through checkout design alone.7 The checkout guide has the fixes and a table of what each platform lets you change.

How fast does a store need to be?

Fast enough that nobody waits between steps. In a Google-commissioned study, Deloitte looked at mobile sites from 37 brands. When pages got 0.1 seconds faster, retail conversions rose 8.4% and average order value rose 9.2%.8

What 0.1 seconds faster did for retail sites on mobile

Product list to product page+3.2%Product page to add to basket+9.1%Conversion rate+8.4%Average order value+9.2%
Source: Deloitte Digital, "Milliseconds Make Millions", commissioned by Google, 2020. Changes after a 0.1 second gain on four speed metrics.

Treat those numbers as direction, not a promise. Google paid for the study and wanted to show that speed matters. Still, the lift showed up at every step of the funnel, and it matches what store owners see.

Most store speed problems come from three places: huge product images, too many apps or plugins, and heavy third-party scripts. Check a page with the free page weight checker, and read what page speed means for the basics.

What do Google and AI shopping tools need from a store?

Clean structure and clean product data. Google's own ecommerce guide asks for links from menus to category pages, from categories to sub-categories, and from there to every product.9 If a product can only be found through the search box, Googlebot may never find it.

Product data matters next. Pages with Product markup can show in Google's merchant listings, the shopping knowledge panel, Google Images and product snippets.10 You can also send the same data through a Merchant Center feed. The glossary entry on product schema explains the markup.

Filters are the most common technical trap. Each filter combination can create a new URL, and Google calls this a cause of overcrawling and slower discovery of new pages.11 The category page guide shows how to decide which combinations deserve to be found.

AI shopping assistants read the same things: clear product names, prices, stock, delivery and returns, in text and in structured data. A store that is easy for Google to read is easy for them too. I cover the AI side in how to use AI in ecommerce, and the free AI shopping readiness checker shows what they can read on one page.

Which platform should a store run on?

Pick it for how you sell, not for its templates. For most small and mid-sized stores, Shopify is the safe default. WordPress with WooCommerce suits stores that publish a lot or want to own every file. Wix and Squarespace suit a few products sold beside a service.

Bigger catalogues with complex pricing, many warehouses or B2B accounts may need Adobe Commerce, formerly Magento, or a headless setup. Expect a much larger build and running cost. I compare the options in B2B ecommerce platforms.

The full comparison of speed, fees, lock-in and security is in WordPress vs Webflow vs Wix vs Shopify. The platform picker gives you a match in seven questions.

What should an ecommerce website design company deliver?

More than page designs. A store is a system of pages that share one product data model. When you hire, ask for these in writing:

  • A page inventory. Category, product, cart, checkout, confirmation, account, search results, and any landing page templates.
  • A product data plan. Which attributes each product type has, which become filters, and which show on the product page.
  • Mobile layouts first. Most store traffic is on phones, and Baymard's mobile scores are the weakest.2
  • Structured data and feeds. Product markup on product pages, and a plan for Merchant Center if you use it.
  • A speed budget. Target page weight and Core Web Vitals, checked on real product pages with real photos.
  • Analytics events. View item, add to cart, begin checkout and purchase, set up and tested before launch.
  • A handover. How to add products, edit categories and change shipping rules without calling anyone.

The guide to choosing a website designer has the questions to ask and a free scorecard.

Template or custom store design?

A good theme is enough for most stores. Theme makers have already built the patterns shoppers know, and familiar layouts are easier to use. Custom work pays when a key page has measured friction or when your products need something no theme does, like a configurator.

The middle path works well here. Start from a fast theme, add your own design system, and redesign only the pages where data shows a problem. Template or custom design has a decision tree.

Where should you start?

Start where the money leaks, in this order:

  1. Checkout costs and accounts. Show delivery cost early, and make guest checkout the most obvious option.
  2. Product page information. Scale photos, delivery date, returns, and reviews near the price.
  3. Filters on your top three categories. Filters that match how people choose those products.
  4. Speed on product and category pages. Compress images and remove apps nobody uses.
  5. The order confirmation and the first emails. Clear status, a tracking link and one useful next step.

Each step is a few days of work, not a rebuild. Measure the funnel before and after each one.

Questions people ask

What makes a good ecommerce website design?

A good store makes products easy to find, answers every question on the product page and takes payment without surprises. Large benchmarks show most leading stores still fall short on lists, product pages and checkout, so getting these right sets you apart.

How many pages does an online store need?

At minimum a homepage, category pages, product pages, a cart, a checkout, an order confirmation, and policy pages for delivery, returns and privacy. Add landing pages for campaigns and a contact page.

Is Shopify good for ecommerce website design?

For most small and mid-sized stores, yes. It handles hosting, security and checkout, and about 79% of Shopify sites pass Core Web Vitals. Check our platform comparison if you publish a lot or need to own every file.

How long does it take to design an online store?

A theme-based store with up to a few hundred products often takes six to twelve weeks, including product data and testing. A custom design or a complex catalogue takes longer. Product photos and data usually set the pace, not the design.

What should I fix first on my store?

Checkout. Show the full cost early and make guest checkout obvious. Among shoppers who weren't just browsing, 40% abandoned over extra costs and 18% over forced accounts in Baymard's survey.

Do I need a custom ecommerce design?

Usually not at first. A fast, well-chosen theme with your own brand on top covers most stores. Go custom for a page where your data shows friction, or for products that need a configurator or a builder.

Sources

  1. "Quarterly Retail E-Commerce Sales, 2nd Quarter 2026," U.S. Census Bureau, 18 August 2026. census.gov
  2. Mark Crowley, "Product List UX 2025: 8 Common Pitfalls & Best Practices," Baymard Institute, updated 9 September 2025. Research vendor data. baymard.com
  3. Edward Scott, "Product Page UX 2026: 10 Pitfalls and Best Practices," Baymard Institute, updated 18 March 2026. Research vendor data. baymard.com
  4. "Checkout UX: 10 Common Pitfalls & Best Practices," Baymard Institute, 2025 benchmark. Research vendor data. baymard.com
  5. "Filtering Options," Product Lists & Filtering UX benchmark, Baymard Institute. Research vendor data. baymard.com
  6. "How Online Reviews Influence Sales," Spiegel Research Center, Northwestern University, 2017, with PowerReviews. spiegel.medill.northwestern.edu
  7. "Cart Abandonment Rate Statistics," Baymard Institute, last updated 22 September 2025. Research vendor data. baymard.com
  8. "Milliseconds Make Millions," Deloitte Digital for Google, 2020, summarised on web.dev. web.dev
  9. "Help Google understand your ecommerce website structure," Google Search Central. developers.google.com
  10. "Merchant listing (Product, Offer) structured data," Google Search Central. developers.google.com
  11. "Managing crawling of faceted navigation URLs," Google Search Central, December 2024. developers.google.com

Keep reading

All insights

Product page design that converts

Images, price, delivery, sizes, reviews and SEO.

Read the guide

Checkout page design

Fixes for the avoidable reasons people leave.

Read the guide

Product page audit

Twenty-four checks and a prioritised fix list.

Audit a page

Planning or fixing a store?

I design stores around the pages that sell: category, product and checkout. Twenty minutes is usually enough to find the biggest leak.

Book a 20-minute call