AI recommendations push users to named brands
Users who received a specific brand recommendation from ChatGPT were far more likely to visit that brand’s website within a week, according to a new Similarweb analysis reported by Danny Goodwin at Search Engine Land. The study, which tracked opted-in U.S. desktop behaviour from July to December 2025, finds that AI-driven suggestions materially shift downstream traffic and complicate standard digital attribution. See the Search Engine Land report for the full write up. (https://searchengineland.com/chatgpt-recommendations-brand-website-visits-study-480989) and the original Similarweb report. (https://www.similarweb.com/corp/reports/the-downstream-impact-of-ai-visibilty/)
What the numbers show
Similarweb followed users who asked ChatGPT industry-relevant questions, received a named-brand recommendation, and then visited either that brand or a competitor within seven days. The study excluded anyone who had visited the brand in the previous four weeks or who explicitly named the brand in their prompt, isolating recommendations as the likely trigger.
Across finance, travel and beauty the effect is clear and consistent:
– Finance: After an American Express recommendation 7.2 percent of users visited American Express versus 3.1 percent who went to Capital One. After a Capital One recommendation 14.2 percent visited Capital One while 3.8 percent visited American Express.
– Travel: After a Skyscanner recommendation 9.5 percent visited Skyscanner compared with 7.6 percent to Kayak. After a Kayak recommendation 12 percent visited Kayak compared with 3.4 percent to Skyscanner.
– Beauty: After a Sephora recommendation 7.9 percent visited Sephora versus 3.3 percent to Ulta. After an Ulta recommendation 7.6 percent visited Ulta compared with 4.6 percent to Sephora.
Why attribution undercounts AI’s role
A striking finding is that most AI-influenced visits do not show up as AI referral traffic in website analytics. Similarweb found that 55.9 percent of AI-influenced visits arrived via search, compared with 40.4 percent for non-AI-influenced visits, while direct traffic accounted for just 19.9 percent of AI-influenced visits versus 38.8 percent for standard visits. In short, the conversation with an AI appears to shape intent, but the later click is often routed through search and therefore misattributed.
Engagement is higher among AI-recommended visitors
Not only did AI-recommended visitors reach brand sites more often, they engaged more deeply once there. AI-influenced visitors viewed on average 12 pages and spent 11.8 minutes on site, compared with 6.5 pages and 5.6 minutes for non-AI-influenced visitors. That suggests users arrive with a narrower set of options and a stronger purchase intent, a point Similarweb highlights in its analysis.
What this means for Irish businesses and marketers
For companies operating in Ireland and beyond the takeaway is straightforward – AI visibility now matters to the bottom line and to brand health. A few practical implications:
– Reassess attribution models – Standard channel reporting understates AI influence. Marketers should augment analytics with panels, surveys, and first-party signals to capture AI-driven intent.
– Protect brand equity across touchpoints – Where AI recommendations can name brands, those brands benefit disproportionately. Firms should prioritise clarity in brand assets and SEO signals so that AI-primed searches land on owned pages.
– Consider platform engagement – Businesses should evaluate how to be discoverable in AI ecosystems, whether by working with platform partners, providing structured data, or ensuring easily parsable content.
– Focus on conversion-ready content – Given the higher engagement and time on site from AI-recommended visitors, invest in content and journeys that convert an informed user quickly.
– Demand transparency from AI providers – Firms should push for clearer reporting on how recommendations are selected so that the commercial effect can be measured and contested if necessary.
A market-first response, not a regulatory one
The most effective immediate response for most firms is commercial and technical rather than political. Companies that move fastest to understand and optimise for AI referral patterns will capture more of the demand being generated outside traditional paid and organic channels. That does not preclude asking platforms for transparency, but Irish businesses should not wait for regulators to catch up before protecting their brands.
Caveats and next steps
Similarweb’s findings are limited to opted-in U.S. desktop panel data and the three sectors studied – finance, travel and beauty – so the exact magnitudes may differ for mobile or other industries. Still the directional signal is strong: AI recommendations move people and change the shape of downstream engagement. For a deeper read consult the original coverage by Danny Goodwin at Search Engine Land and the full Similarweb report. (https://searchengineland.com/chatgpt-recommendations-brand-website-visits-study-480989) (https://www.similarweb.com/corp/reports/the-downstream-impact-of-ai-visibilty/)
If Irish marketers take one lesson from this it should be to stop treating conversational AI as an abstract technology trend and start treating it as a distribution channel that can be measured, optimised and, where necessary, contested.
