UX / Homepage
Ecommerce Homepage Design: What Actually Converts
// the homepage's job is to point, not to sell everything at once
By Fracto Solutions
October 8, 2026
The short answer
A good ecommerce homepage tells a first-time visitor what you sell, who it’s for, and where to go next, fast. Its job is to route people into categories, search, and key products, not to sell everything at once. Lead with clear value and navigation above the fold, build trust, and keep it fast and clean on mobile.
~57%
~1%
70%
What is the homepage's real job?
To orient a visitor and route them onward, not to close a sale on the spot. The homepage should answer what you sell, who it’s for, and where to go next, then hand the shopper to a category, a search, or a product where buying actually happens.
Teams often overload the homepage, trying to feature every product, promotion, and message at once. The result is a page that says everything and therefore nothing. Baymard rates the average top-grossing US and European site “poor” on homepage and category navigation, with 76% performing at “mediocre” or worse. On the homepage specifically, only just above half of sites manage a “decent” or better rating. Its scored set is a subset of a wider 344-site benchmark, so we do not attach that number to the 76%. Three of its findings describe exactly the overload problem: 70% of sites give no direct access to products from their inspirational imagery, so the big lifestyle hero leads nowhere; 55% run overly aggressive or distracting ads on the homepage; and 59% do not provide the full scope for links on mobile homepages specifically, so a tap is a guess.
Treat the homepage as a signpost: its success is measured by how well it sends people deeper, toward product pages and categories where conversion happens.
What should be above the fold?
Enough for a first-time visitor to instantly understand what you sell and find a way in: a clear value proposition, obvious navigation and search, and a primary call to action. The space people see first should orient, not decorate.
The area visible without scrolling carries the most weight, though not for the reason most articles claim. People do scroll, and more than they used to. Nielsen Norman Group’s eye-tracking study, 120 participants and more than 130,000 fixations, found about 57% of page-viewing time spent above the fold, with 74% in the first two screenfuls and more than 42% in the top fifth of the page alone. Two things to note before you apply it: that is an average across news, ecommerce, blogs, FAQ and encyclopedic pages rather than an ecommerce-homepage figure, and NN/g’s equivalent 2010 measure was 80% on a smaller screen, which it declines to read as proof that behaviour itself changed. What survives is the shape: attention drops sharply below the fold, and the first screen takes the majority of it.
Spend it on clarity: say what you sell and who it’s for, show the main navigation and search, and give one clear next step. Full detail is in our above-the-fold guide, but the principle is simple, orient before you impress.
What should an ecommerce homepage include?
A clear value proposition, primary navigation and search, featured categories or products, trust signals, and a path to help and policies, in roughly that priority. Each element earns its place by helping a visitor decide or move onward.
Think in priority order rather than as a checklist to cram in.
Element
Job on the homepage
Priority
Value proposition
Say what you sell and who it's for
Essential, above the fold
Navigation + search
Route people to products
Essential, above the fold
Featured categories/products
Give a clear way in
High
Trust signals
Reassure a first-time buyer
High
Social proof
Reviews, press, ratings
Medium
Help + policies
Shipping, returns, contact
Medium, usually footer
Order the homepage by what helps a visitor decide and move onward, not by what each team wants featured.
What are the most common homepage mistakes?
Trying to feature everything, a vague value proposition, rotating carousels that bury the message, weak or hidden navigation, and slow load. Each one blurs the single job the homepage has: orient the visitor and send them onward.
The biggest offenders are overload and vagueness. A homepage that promotes ten things equally promotes nothing, and a headline that could belong to any brand fails to say why this store.
Hero carousels are the classic trap, and the best public data on them is worth reading carefully. Erik Runyon tracked carousel clicks on nd.edu against homepage visitors and found that approximately 1% of visitors clicked a carousel feature at all. Of the 28,928 clicks recorded, 84% landed on the first slide, with the remaining four positions splitting roughly 4% each. Read the caveats before you quote it. It counts clicks, not views, so it does not prove people never see later slides. The comparison across Runyon’s other sites is confounded: the higher-performing ones are Notre Dame sub-sites with narrower audiences, and Runyon attributes the difference to audience rather than carousel type. The one auto-forwarding carousel in his set did draw the highest click-through at 8.8%, but that is a sample of one, and Runyon himself points readers to Nielsen’s warning against auto-forwarding. The data is from 2013 and from university rather than ecommerce sites. What it supports is narrow but useful: carousels earn very little engagement, and whatever sits on slide one does nearly all the work. We argue for a single strong hero on the simpler grounds that it is clearer and faster.
A slow homepage compounds all of this; see how speed affects conversion.
How should the homepage work on mobile?
Mobile-first, because most traffic is mobile. The value proposition, search, and a clear way in must all be reachable without pinching or hunting, and the page must load fast on a real phone connection, not just on your office wifi.
Judge the homepage on a phone first. How much of ecommerce traffic is mobile depends on how you count: Siege Media’s analysis of 13.4 billion sessions across 120 brands puts its ecommerce row at 55.8% mobile across all traffic and 65.3% for the “typical site”, where each brand counts equally. Either way it is the majority, and Baymard finds 59% do not provide the full scope for links on mobile homepages, a mobile-only measure, so shoppers tap without knowing where they are going.
On mobile, vertical space is scarce and attention is short, so the value proposition and primary path have to land immediately, and speed matters even more. This is the same mobile-first discipline covered in mobile conversion optimization.
How do I measure if my homepage is working?
By how well it moves people onward, not by how it looks. Watch homepage bounce and exit rates, click-through into categories and products, and the paths people take after landing. Low onward movement means the signpost isn’t working.
The homepage is a routing page, so measure routing. If visitors land and leave, or never click into a category, the page is failing its job regardless of how polished it looks. Session recordings and click maps show where attention goes and where it stalls. It’s the same find-the-leak approach behind our growth-strategy pillar.
How do returning visitors change what the homepage needs?
A first-time visitor needs orientation; a returning one usually wants a fast way back to products, new arrivals, or their account. A good homepage serves both without forcing either to wade through the other, which is one more reason to keep it clear and path-driven rather than crammed.
Not everyone landing on your homepage is new. Returning shoppers and existing customers often arrive with intent, to reorder, to see what’s new, to pick up where they left off, and they don’t need the full first-impression pitch. The way to serve both is not to add more; it’s to keep the primary paths (search, categories, new arrivals, account) obvious, so a confident returning visitor can move instantly while a new one still gets oriented.
This is where overload hurts twice: a homepage stuffed with promotions slows the new visitor’s understanding and the returning visitor’s shortcut at the same time. Clarity and strong paths serve every audience, which is why subtraction so often beats addition on the homepage.
How does the homepage fit the rest of the store?
The homepage is the top of a journey, not a standalone page. It should hand shoppers cleanly to category and product pages, where buying actually happens, and its promises should match what those pages deliver. Judge it as the first step in a path, not in isolation.
A homepage can look perfect and still fail if it routes people to weak category or product pages, or if its promises don’t hold up a click later. Treat the homepage as the opening of a funnel: its job is to set expectations accurately and move people onward, and it succeeds only if the pages it feeds convert. That whole-journey view is central to ecommerce growth strategy.
We treat the homepage as a routing page and measure it that way.
We cut homepage overload down to a clear value proposition and a few strong paths, make sure it orients a first-time mobile visitor in seconds, and then judge it by how many people move onward into categories and products, not by how it looks in a design review.
It’s the same measurement-first discipline behind our conversion optimization and web development work: we agree the metric before the test runs, change one thing at a time, and report the result at the confidence level the test actually reached.
// we would rather show up in your reporting than in your inbox
Frequently asked questions
What is the most important part of an ecommerce homepage?
The area above the fold: a clear value proposition that says what you sell and who it’s for, plus obvious navigation and search. If a first-time visitor can’t orient instantly, the rest of the page rarely gets a chance.
Should my homepage have a hero carousel?
Carousels earn very little engagement either way. The best public dataset, Erik Runyon’s 2013 tracking of carousel clicks against homepage visitors on nd.edu, found approximately 1% of visitors clicked a carousel at all, and 84% of the 28,928 recorded clicks went to the first slide. Note it measured clicks rather than views, is 2013 data from university sites, and that comparisons within it are confounded by audience. We recommend a single strong hero on the grounds that it is clearer and faster, rather than claiming rotation itself is fatal.
Is the homepage where sales happen?
Rarely directly. The homepage orients and routes; conversion usually happens on product and category pages. Measure the homepage by how well it sends people onward, not by on-page purchases.
How many things should the homepage promote?
Few. A homepage that features everything equally promotes nothing. Lead with one clear value proposition and a small number of strong paths rather than competing messages.
Should I design the homepage for mobile or desktop first?
Mobile first, since most traffic is mobile. Make sure the value proposition, search, and a clear way in all land on a phone, and that the page loads fast on a real mobile connection.
Sources
Homepage UX performance: Baymard Institute, homepage and category navigation usability research. The average top-grossing US and European site rates “poor”, with 76% performing at “mediocre” or worse on homepage and category navigation combined, and on the homepage specifically only just above 50% delivering a “decent” or better performance. The scored dataset is described as a subset of a wider benchmark of 344 ecommerce sites, so we do not attach 344 as the denominator for the 76%. Separately, 900+ usability issues were encountered during moderated testing of 19 sites. This page carries no publication or update date, and Baymard’s dated 2025 benchmark reports a different cut (58% of desktop and 67% of mobile sites “mediocre” to “poor”), so the two should not be mixed. https://baymard.com/research/homepage-and-category-usability
Homepage-specific compliance gaps: Baymard Institute, 2025 homepage and category navigation UX benchmark. 180+ leading US and European sites, 16,000+ UX elements manually reviewed and scored. 70% do not provide direct access to products in inspirational imagery; 55% use overly aggressive and distracting ads on the homepage; 59% do not provide the full scope for links on mobile homepages, which is a mobile-only measure. The other two are not device-split. Figures are quoted as Baymard states them. Published 6 August 2024, last updated 30 September 2025. https://baymard.com/blog/ecommerce-navigation-best-practice
Attention above and below the fold: Nielsen Norman Group, “Scrolling and Attention”, published 15 April 2018. Eye-tracking across 120 participants and more than 130,000 fixations on a 1920×1080 screen, covering news, ecommerce, blogs, FAQ and encyclopedic pages. About 57% of page-viewing time above the fold; 74% within the first two screenfuls; more than 42% within the top 20% of the page; more than 65% within the top 40%. Two limits we state rather than bury: these are averages across page types and industries, not an ecommerce or homepage cut, and NN/g’s stated goal was to characterise general ranges of behaviour rather than analyse individual sites. NN/g’s 2010 equivalent was 80%, but measured on a 1024×768 screen against 1920×1080 in 2018, and NN/g explicitly declines to separate the effect of larger screens from any change in user behaviour, so we present it as a change in measured distribution and not as proof that people scroll more. https://www.nngroup.com/articles/scrolling-and-attention/
Carousel engagement: Erik Runyon, “Carousel Interaction Stats”, published 22 January 2013, measuring nd.edu from mid-October 2012. Approximately 1% of homepage visitors clicked a carousel feature; of the 28,928 clicks recorded, 84% landed on position 1 with the remaining four positions splitting roughly 4% each. Percentages are clicks measured against homepage visitors in Google Analytics. Note that the separately reported 315,665 figure counts manual feature rotations by users, not impressions, so it is not the denominator for the click rate. Three other Notre Dame sub-sites with static carousels returned 1.7% to 2.3% click-through with position 1 taking 48% to 62%, and one auto-forwarding carousel returned 8.8% with position 1 at 40%. We do not present that as evidence that auto-forwarding works: it is a sample of one, Runyon attributes the differences between his sites to audience rather than carousel type, and he points readers to Nielsen’s warning against auto-forwarding carousels. The data is from 2013 and from university rather than ecommerce sites, and it counts clicks rather than views. https://erikrunyon.com/2013/01/carousel-interaction-stats/
Mobile share of ecommerce traffic: Siege Media, mobile vs desktop traffic share, last updated 11 August 2026. The study covers 13.4 billion sessions across 120 brands from 1 January 2023 to 30 June 2026; the ecommerce row within it reads 55.8% mobile across all traffic and 65.3% for the “typical site”, where each brand is weighted equally. The study covers only six months of 2026 and excludes tablet and smart TV traffic. https://www.siegemedia.com/research/mobile-vs-desktop-traffic-share
Stated as Fracto’s position, not research: that the homepage is a routing page rather than a selling page, the priority order of homepage elements, and judging the homepage by onward movement rather than on-page conversion. These come from client work rather than a published study, and we recommend a single strong hero over a carousel on grounds of clarity and speed rather than on the carousel data alone.