Published: 26.02.2026
Just yesterday, “SEO = Google.” In 2026, the logic is changing: part of search scenarios is moving to AI assistants and AI search. Gartner predicts that by 2026, traditional search volume may decline by 25% due to users shifting to AI chatbots and virtual agents. At the same time, Adobe shows that traffic to retail websites from generative AI sources has grown exponentially (for example, +4,700% year-over-year in July 2025), and during November–December 2025 there was a sharp surge in AI traffic in E-commerce retail.
AI traffic refers to visits from chats and AI search engines (ChatGPT by OpenAI, Gemini by Google, Perplexity, Microsoft Copilot, etc.), where the user receives not a list of links, but a ready-made recommendation: “where to buy,” “which model to choose,” “what is more cost-effective”. Here, the winner is not the one who “ranks higher in the SERP”, but the one the model considers the best answer and a reliable source. Analysts (and specialized agencies) expect that by 2026 around 20–25% of queries will be processed by AI systems — and this accelerates the shift of attention to new “AI comparison sites.”
Broadly speaking, there are three “comparison sites” that bring in buyers:
For example, in Google Play, Genspark AI Workspace shows 500K+ downloads and a 4,8 rating. This is a marker that “AI as an interface” is becoming mainstream even beyond the classic web.
In classic SEO, we compete for rankings. In Generative Engine Optimization (GEO), we compete for mentions and citations in a generated answer. AI typically “chooses” sources based on signals such as:
Explicit signals of expertise also help: author/date of update, links to sources, real photos, and specifications.
The key change for E-commerce: AI “prefers” not slogans, but максимально specific answers and transparent purchasing terms. And one more thing: AI traffic in E-commerce often has higher intent. According to Adobe , users who came to retail websites from generative AI assistants were 33% less likely to leave the site immediately (lower bounce rate).
AI does not “guess” your prices or availability – it reads data. A minimum checklist:
For large catalogs, it is critical to have a clean and stable product feed (Merchant Center/marketplaces). This makes the assortment “readable” for shopping responses.
AI willingly uses materials that:
Practice for E-commerce: create a cluster of pages targeting prompts such as: — “best … for …,” “which is better: A or B,” “what is the difference between …”
— “selection by budget,” “top models 2026,” “alternatives to the brand …”
On category/brand pages, add an FAQ with specific answers (no fluff) and update it regularly. On product pages, add blocks like “who it’s for” + “compare with” (internal linking to alternatives).
Models prefer brands with independent mentions and consistent facts. PR + SEO works, but with a focus on verifiability:
Capgemini reported that 71% of consumers want to see generative AI integrated into the shopping experience. Boston Consulting Group recorded a +35% increase in the use of GenAI in shopping scenarios during February–November 2025. This means that an “AI recommendation” is becoming a form of trust – and the brand must appear verifiable.
To enable AI to compile a recommendation from your content blocks:
Days 1–7: audit of schema + feed, fix crawl issues/duplicates, update policy pages.
Days 8–20: content cluster of “buying guides” + FAQ, update top product pages.
Days 21–30: mentions/case studies with numbers, AI referral analytics, test a landing page for 3–5 key prompts.
AI does not “kill” SEO, it redistributes attention. Structure, evidence, and reputation win. And even a small amount of AI traffic in E-commerce can be valuable because it comes at the decision-making stage.
If you are experiencing difficulties with a GEO strategy, optimization for AI assistants, or measuring this channel, contact Compas Agency .
Author: Artur Kvak is an SEO and AI promotion specialist, working with E-commerce and content strategies that get featured in AI recommendations.
Follow us and register on SalesDoubler platform
SalesDoubler.pro content manager