Which retail websites attracted the most UK traffic in August 2026? The earlier Semrush snapshot used for this article placed Amazon.co.uk first at an estimated 212.67 million visits, followed by eBay.co.uk at 95.42 million and Argos.co.uk at 38.04 million. Etsy, Tesco, DIY.com, Asda, Currys, AliExpress and Next completed the stated top ten. These are inherited third-party estimated website-visit figures, not independently re-extracted in this update and not official revenue statistics. A large audience does not automatically indicate more orders, greater profitability, better customer outcomes or a stronger UK domestic supply chain. The distinction between traffic and commercial results must accompany every interpretation of the ranking.
The value of the list is not merely knowing which names appear at the top. Their different operating models reveal why identical headline traffic figures may demand different technology and operational decisions. A marketplace handles sellers and disputes, a grocer manages perishables and delivery slots, and an electronics retailer supports detailed comparisons and warranties. This guide shows how to read the original ranking without conflating websites, businesses, visits and purchases, then turns those distinctions into practical considerations for UK ecommerce teams.
The ten sites in the dated traffic snapshot
The previously reported estimated visits were Amazon.co.uk 212.67 million, eBay.co.uk 95.42 million, Argos.co.uk 38.04 million, Etsy.com 25.61 million, Tesco.com 23.55 million, DIY.com 23.39 million, Asda.com 22.91 million, Currys.co.uk 21.85 million, AliExpress.com 20.30 million and Next.co.uk 20.22 million. These figures are specific to the cited Semrush August 2026 UK retail website ranking. They should not be presented as live traffic counts in a later month. A fresh ranking would require checking the original provider's exact current scope, country filter, device coverage and observation period.
Why the top two are not ordinary single-brand stores
Amazon and eBay are marketplaces with substantial third-party seller activity as well as other commercial features. The traffic figure attaches to a website, not to one merchant selling an identifiable inventory. A buyer may visit the same marketplace to compare products, contact a seller, track an order or manage a listing. A retailer comparing its own single-brand site with a marketplace needs to account for this aggregation rather than assume the same browsing patterns, revenue mix or operational responsibilities.
Understanding the gap between first and third
The inherited estimates show Amazon exceeding eBay by more than twice and eBay exceeding Argos by a substantial margin. This is a comparison of estimated monthly visits, not market concentration measured by sales value. Customers can visit a site multiple times without purchasing, and one purchase can generate several sessions. A third-party traffic methodology may estimate from panels and observations rather than counting each retailer's audited server logs. Avoid treating a rank position as a certified financial statement.
Mobile shares can matter more than rank for design
The prior article cited approximately 78% mobile traffic for Argos, 80.47% for Asda, 74.85% for Currys, 72.85% for DIY.com, 73.18% for Next and 65.9% for Tesco. These are dated provider estimates, not independently measured device analytics supplied by those retailers. They nevertheless illustrate a useful design question: can a shopper search, filter, read product information and complete checkout on a narrow screen? Teams need to test real devices, network conditions, accessibility and payment states, not equate a responsive layout with a reliable mobile transaction.
Desktop use remains commercially relevant
The original snapshot suggested approximately 50.7% desktop share for Amazon and 63.85% for eBay. Those proportions should not be turned into claims about the ages or intentions of visitors. Some tasks, including multi-product comparison and seller administration, may work differently on larger screens, but establishing why requires additional research. Retailers should test important customer journeys across devices and check whether baskets, identity, order history and preferences remain coherent after a device switch.
Retail business models change the meaning of traffic
Grocers need delivery-slot and substitution workflows. Electronics retailers benefit from reliable specifications and warranty information. Home improvement businesses may need inventory visibility across delivery and store collection. Fashion retail brings sizes, returns and photography requirements. Marketplaces add seller, listing and dispute complexity. A shared visit metric conceals these operational differences. For a smaller retailer, choosing relevant peers by product and fulfilment model is often more informative than copying the interface of the highest-ranked site.
Do not translate visits into orders without evidence
Conversion rate, basket size, delivery fees, margins and customer retention are different quantities. Multiplying estimated visits by a guessed conversion rate would create an invented sales number. Nor can gross merchandise value be inferred directly from traffic for marketplaces with varied merchant arrangements. If financial benchmarking matters, collect separately published company data with aligned reporting periods and definitions, then explain where comparison remains impossible. A striking traffic table should not create false precision.
Warehouse systems sit behind every appealing website
Traffic growth is useful only when product availability and fulfilment remain dependable. A promotion may drive shoppers to a product that is already reserved elsewhere. Returns can create apparent stock before inspection, and split shipments can leave an order status misleading. These operational problems do not appear in a monthly visitor ranking. A retailer should trace important order events across the storefront, warehouse, carrier and payment provider, with clear ownership of mismatches and customer communication.
Search and navigation shape the actual customer task
A large catalogue needs understandable categories, filtering, accurate variants and search results that match customer intent. A small speciality shop may succeed with a shorter path and better explanatory content instead of complex marketplace-style filtering. Test product discovery from real customer questions, including spelling variations and missing specifications. Counting visits cannot tell a team whether users find suitable products quickly or abandon searches because the data are inconsistent.
Check mobile performance as an end-to-end journey
Optimising an image alone may not repair a slow checkout caused by payment scripts, inventory lookups or excessive third-party tags. Review representative navigation, search, basket changes, login, address entry, payment and failure recovery. Account for accessible controls, unstable connectivity and smaller screens. Record real observed bottlenecks before implementing a large redesign. Provider traffic estimates can identify where mobile deserves attention, but cannot diagnose a particular shop's code.
Prepare for peaks without copying enterprise complexity
Big retailers operate at enormous scale, but a smaller team need not reproduce all their internal systems. Prioritise reliable stock reservations, observable payment processing, safe order retries and transparent customer updates. Define who investigates an event that occurs twice or never reaches the warehouse. Test a promotion scenario before it happens and establish a practical rollback or escalation plan. A stable operational foundation often provides more value than a fashionable new interface feature.
Use the ranking as a benchmark with limits
The original Semrush ranking is appropriate for describing estimated UK retail-site visitation in its stated month. It is not appropriate for declaring the most profitable company or the strongest customer experience. Recheck whether the provider has updated its category and whether international domains are included in a consistent way. Record observation dates with every refreshed table. If a site moves position, investigate whether the change reflects genuine demand, the estimation method or seasonal shopping rather than automatically attributing it to product quality.
What smaller ecommerce teams can actually adopt
Choose one operational bottleneck visible to customers: inaccurate availability, confusing delivery messages, slow mobile checkout or unresolved returns. Measure the present outcome, change the relevant workflow and assess whether quality improves. Use the leaders as examples of different business models, not a universal template. Review product-information accuracy, customer support and mobile journey alongside raw acquisition traffic. The aim is a trustworthy buying experience for the retailer's own audience.
The headline conclusion and its evidence boundary
The dated third-party snapshot showed a diverse UK top ten dominated in raw estimated visits by Amazon and eBay. It also suggested that mobile and desktop usage mixes vary substantially across retailers. Those facts can guide questions about technology and fulfilment but cannot establish sales, profit, market share or customer satisfaction. The strongest use of the ranking is to decide which business processes deserve investigation, then obtain direct operational evidence before making an investment.
Source and method: Semrush UK ecommerce and retail website traffic rankings (provider page; August 2026 figures inherited from the existing article, not independently re-extracted in this update); GOV.UK online selling requirements for the distinction between customer commitments and promotional traffic.
Traffic ranking versus actual retail commercial performance
| Question | Estimated website visits | Retail sales and operational results |
|---|---|---|
| Measurement | Estimated visits to a site in a defined period. | Orders, value, margins, fulfilment and customer outcomes. |
| August example | Amazon 212.67m and eBay 95.42m inherited from dated Semrush snapshot. | No comparable audited sales values established by these figures. |
| Multiple events | One person can generate many visits. | One order can involve several parcels and payments. |
| Primary risk | Interpreting estimated visits as purchases. | Comparing dissimilar business models or periods. |
| Appropriate decision | Audience and device-experience benchmarking. | Direct operational and financial evidence for investment. |
Source: Semrush retail web-traffic provider. The table explains measurement distinctions; no sales metrics are inferred.
Frequently asked questions
Was Amazon the busiest UK ecommerce site in the cited snapshot?
Yes, according to the existing article's Semrush August 2026 estimated-visit dataset; the original provider figure was not re-extracted during this repair.
Does a higher rank mean greater UK sales revenue?
No. Visits, transactions, gross merchandise value and profit are different metrics.
Are app sessions included automatically in website visits?
Not necessarily. Check the provider's specific website and device measurement rules.
Why should smaller shops examine mobile use?
Real shoppers need usable navigation, search, checkout and recovery on narrow screens and weaker networks.
Can a marketplace be compared directly with a single retailer?
Only with caveats about multiple sellers, product mix and distinct operating models.