When the Route to the Customer Changes

Google’s European shopping changes and the commercial value of a direct visit

A product appearance in search is a useful thing to count. A visit to somebody else’s website is a rather different commercial asset.

That distinction deserves attention as Google changes its European product results. Retailers need to understand who receives the next click, what happens before the shopper reaches their store and whether the resulting business is worth the cost.

The evidence so far

On 18 September, Search Engine Roundtable reported that organic “Popular Products” carousels had largely disappeared in several European Economic Area markets, including Germany. It cited third-party observations and reported confirmation from Google’s Ads Liaison. This is evidence about particular search formats, not proof that every free product presence across Google has ended. Nor does it establish a loss of retail revenue. Search Engine Roundtable, 18 September

Google’s documentation describes an aggregator unit for EEA product queries. Approved comparison services supply results, and clicks within the unit take shoppers to the aggregator’s website. A separate supplier unit provides space for direct providers, appearing only alongside the aggregator unit. Google’s aggregator documentation, Google’s supplier documentation

The evidence supports a change in how access is organised. Its effect on an individual retailer remains an open commercial question.

The commercial reading

»A product can remain visible while the retailer becomes one step further from the customer.«

Consider a shopper who encounters a product through a comparison service. Before reaching the retailer, they may compare competing sellers, change their selection or leave altogether. The retailer needs to know how much worthwhile demand survives that journey.

There is an opportunity here as well as a risk. A specialist with little direct reach might benefit from a comparison service that introduces new buyers. An established retailer might find that customers who previously arrived directly now come through a partner. Those outcomes call for different decisions, even if both are recorded as growth in referral traffic.

No evidence cited here establishes that the changes have raised acquisition costs. Fees, conversion rates and the mix of genuinely additional customers would have to be examined before making that claim. It would be equally premature to assume that an existing comparison shopping service arrangement for paid Shopping ads guarantees presence in the partner’s organic product results.

This is where a discussion about search belongs in a commercial meeting. The questions concern partner coverage, the terms of access and the contribution earned from the resulting orders. A good ranking report cannot settle them.

A proportionate response

Start with a small set of commercially important searches. Observe the results repeatedly, recording the country, device, date, providers shown and destinations of the links. Keep Germany and Austria separate from Switzerland, which is outside the EEA. A single screenshot is a useful observation, not a market study.

Then compare the routes into the business. Separate direct organic visits, comparison referrals and paid traffic where the data permits. Look beyond visit counts to orders and contribution after attributable acquisition costs. Changes in promotions, stock availability or seasonal demand may explain movements that initially look like a search problem.

Finally, ask existing partners to demonstrate coverage. Which products appear, in which results, with what data requirements and on what terms? If an additional partner looks promising, test a limited range before committing a larger catalogue or budget. The useful outcome is evidence of commercially worthwhile demand.

There is no case for disabling product feeds or increasing advertising spend simply because a familiar carousel has disappeared. There is a case for checking whether the route on which the business relies still works as expected.

These developments concern conventional product search. They are not evidence that AI recommendations have changed. Their relevance to AI visibility lies in the commercial question they expose. Being included in a result and acquiring the customer are different events, with other businesses potentially making decisions in between.

Before approving a response, a retailer should be able to trace that route far enough to judge what it is buying, from whom, and with what return.

Sources checked on 21 September 2026. The commercial implications are the author’s interpretation. This Radar issue does not draw on proprietary retailer data or a causal study of revenue effects.