Big retail sites' share of estimated Google traffic shrank this autumn

Shoppers put a little more of their spending online

The received wisdom starts from something true. Office for National Statistics figures show that internet sales made up 28.3% of the value of retail sales in Great Britain in September 2026, up from 27.3% in September 2025. That series counts online sales by every kind of retailer, from web-only shops to high street chains with a website. The ONS also revises its monthly figures as later returns arrive, so the fair comparison is like month against like month.

A one-point move sounds modest. It is still the direction retailers want, and it is why the assumption feels safe: more money passing through online checkouts ought to mean more shoppers arriving through search.

Retail's slice of estimated Google traffic got thinner

The Google estimates tell a different story. News sites rose from 22% to 25.2%.

These figures are estimates. The model never sees a real visit and misses brand-new or very niche searches. The panel is a chosen set of well-known sites in each sector, so it does not include every site in that sector. Each figure is a three-month average set against the same three months a year earlier.

The obvious objection is that everyone's Google traffic moves around, so a smaller slice might mean nothing. Shares deal with that. The model is rescaled from time to time, which is why raw visit totals are left aside here. A share compares each sector with the rest of the same panel under the same model. On that like-for-like basis, retail finished the autumn with a smaller piece of the pie and news with a bigger one.

One news site takes four in five of its sector's visits

News is also the most top-heavy sector in the panel. That is the highest single-site share in any of the panel's 14 sectors. Rightmove took well over half of property among nine sites, and Tripadvisor more than half of travel among 15. They rank second and third of the 14 sector leaders.

These sector shares come from the same traffic model and carry the same limits. They describe how concentrated each sector is. They do not show how the news sector's gain was split between its sites. By simple arithmetic, though, most of whatever slice news holds in the panel belongs to a single name.

Fashion has no runaway leader

Fashion and sportswear sits at the other end of the scale. No clothing site has anything like the grip one broadcaster has on news. For a fashion brand, that makes for a crowded field, and an open one.

More money online, a thinner slice of search

Taken together, the two sets of figures do not support the comfortable assumption. The ONS series counts money spent. They also cover slightly different windows: a single September month for spending, and a September to November average for search.

The data cannot explain the gap. It also does not show that shoppers bought less from the big retail sites. What it shows is narrower and still useful.

What this means for UK businesses

Visibility has to be earned in its own right. Your share of a fixed group of competitors is a steadier guide.

• Know how your sector is shaped. In a sector where one site holds most of the visits, you are competing with that giant for every search. In an evenly split sector such as fashion, there is more room for a smaller brand to take ground.
• Use estimates for direction. Model figures show which way things are moving. For your own site, check them against the real clicks in Google Search Console.

The shopper on the bus will keep tapping links all season. The question for a retailer is which links those are, and the autumn estimates show it is less safe than ever to assume they lead to the biggest shop.

How we built it

Search traffic shares (estimates). Estimated monthly UK Google visits come from a commercial SEO traffic model. The model infers visits from ranking positions, modelled search volumes and typical click-through rates. It is not measured traffic. The panel covers well-known UK sites in 14 sectors, chosen by the analyst rather than drawn as a complete or random list. Shares are each sector's (or site's) part of the panel's combined estimate. Figures are three-month averages for September to November 2026, compared with September to November 2025.

The panel's overall estimated total is not reported because the model is periodically rescaled. Shares are used instead, because they compare sectors under the same model at the same time. Single months can jump when the model updates; the three-month averages smooth out some of that, but not all.

Scope limits. The window ends in November, so this piece does not cover the full October to December quarter. Shares are calculated across the whole multi-sector panel. Keyword search-volume data for this piece was collected but came back empty, so search demand is not measured here.

Online share of retail (measured). This uses ONS retail sales series J4MC: internet sales by all retailers as a share of total retail sales value in Great Britain. Like months are compared across years. Monthly values are subject to revision.

No statistical tests were run, and no correlation between the ONS series and the traffic estimates is claimed. The two sources measure different things over different windows.

Data table