E-commercee-commerce visibility

When e-commerce visibility moves: from conventional SEO to AI answers

E-commerce visibility and AI: what changes when assistants answer in place of Google, and how to make your product pages citable.

Mehdi B.

Co-founder · SEO & digital strategy

When e-commerce visibility moves: from conventional SEO to AI answers
In brief

The visibility of an online shop is no longer decided only by Google rankings: AI-generated answers are becoming an entry point, sometimes without a click. Traditional SEO remains necessary, but we recommend adding product pages marked up with Schema.org, FAQs that answer real questions and detailed customer reviews, so that you become a source AI can cite.

For years, the magic formula for selling online seemed set in stone: choose the right keywords, polish your meta tags, and earn quality links to climb onto Google’s first page. But the rules of the game have recently been changing before our eyes. Users are no longer looking only for a list of blue links. They are looking for an immediate, synthesised, precise answer. E-commerce visibility is therefore played out on two fields: ranking in Google, which remains necessary, and presence in AI answers, which requires marked-up product pages, clear answers to buyers’ questions and detailed reviews.

Generative artificial intelligence (such as ChatGPT, Google’s AI Overviews or its AI Mode, launched in Switzerland in October 2025) is redefining how your customers discover your products. It is no longer only a question of ranking, but of presence in the conversation. If your e-commerce strategy rests solely on the old methods of SEO, you risk missing a growing share of purchase journeys. Adobe measured that, over the 2025 holiday season, traffic reaching US retail sites from generative AI tools had risen by 693.4% year on year, while specifying that the user base remains modest. The figure concerns the United States, but it indicates the direction in which buying habits are moving.

So how do you adapt your online shop to this new reality? How do you make your brand the one AI recommends? Let us dive together into this major transformation to understand how to come out ahead.

Understanding traditional e-commerce SEO

Before tackling the AI revolution, it is worth recalling what the current visibility of most merchant sites rests on. “Classic” SEO remains a fundamental pillar, even as it evolves.

Keyword research

Historically, everything starts with search intent expressed in words. You identify that your customers type “waterproof hiking boots” or “best Swiss coffee beans”. Your goal is to place those terms strategically in your product pages and categories to signal to search engines that your page is relevant.

On-page optimisation

Once the keywords are found, the technical work begins. That involves:

  • Compelling title tags and meta descriptions.
  • A clear heading structure (H1, H2, H3).
  • Clean URLs.
  • Optimised images (alt tags, file weight).

This is the foundation for crawlers to understand and rank your content.

Netlinking and content marketing

To gain authority, your site needs external recommendations. Backlinks (inbound links) act as votes of confidence. In parallel, content marketing through a blog makes it possible to capture traffic on informational queries (“how to care for leather shoes”) and then guide the user towards purchase.

These methods still work, of course. But they are no longer enough to capture the attention of users who turn to conversational assistants or AI-augmented search engines.

The rise of AI-driven answers

Artificial intelligence is disrupting the buying journey. Instead of typing a query, scrolling, and clicking on three or four different links to compare, the user now asks a complex question and expects a synthesis.

The end of keyword search, the beginning of semantic search

AI does not read isolated keywords: it understands context and overall intent. If a user asks: “I am looking for a connected watch for trail running, compatible with iPhone, max budget CHF 500.-”, the AI will scan the web to build a tailored answer.

It will not simply list pages containing those words. It will:

  1. Understanding the criteria (trail = precise GPS, battery life; iPhone = iOS compatibility).
  2. Filter by price (under CHF 500.-).
  3. Summarising the reviews and technical specifications found across several sites.

“Zero-Click” visibility

The risk for e-commerce operators is real: the user gets their answer directly in the AI interface or at the top of the search results (via Google’s AI Overviews). If they have all the information, why would they click through to your site?

Your objective is no longer only to attract the click, but to be the source cited in the generated answer. That is sometimes called GEO (Generative Engine Optimization). If the AI presents your product as “a good choice for beginners”, the visitor who clicks arrives with a decision already formed, even if overall traffic to your site falls.

A group of people stands around a humanoid robot, which is writing on a transparent board covered with sticky notes, capturing the emotions shared within the group.

Optimise for AI-generated answers

How do you convince these new gatekeepers of the digital temple to talk about you? Unlike classic SEO where you optimise for a robot, here you optimise for a “human understanding” simulated by the machine.

1. Structure your data (Schema Markup)

Structured data tells the machine what the page displays. Google states that no special markup is required for its AI features, but Product markup remains the clearest way to give a page’s price, currency, availability and reviews, provided it matches the visible text. For a Swiss shop, the currency is given using the ISO 4217 code CHF :

{
  "@context": "https://schema.org/",
  "@type": "Product",
  "name": "Veste de randonnée imperméable",
  "offers": {
    "@type": "Offer",
    "price": 289.00,
    "priceCurrency": "CHF",
    "availability": "https://schema.org/InStock"
  }
}

Google requires at minimum the product name and, within the offer, the price; it recommends the currency and availability, and makes currency mandatory for merchant listings. Then add what the buyer compares: dimensions, material, customer reviews.

2. Favour conversational content and FAQs

Users ask natural questions. Your content has to answer them directly. Turn your classic product descriptions into useful guides. Instead of a plain list of features, add sections such as “Why choose this product?”, “Who is it for?” or “How do you use it?”. A well-stocked FAQ section on every product page is a gold mine. If you explicitly answer the question “Can this blender crush ice?”, the AI will be able to reuse that information word for word in its answer.

3. Cultivate authority and genuine reviews

AI models are trained to favour reliable sources. Detailed customer reviews play a crucial role here. An AI will be more inclined to recommend a product if it “reads” hundreds of comments confirming its sturdiness or its effectiveness.

Encourage your customers to leave descriptive reviews, not just stars. Comments mentioning specific use cases (“Perfect for my bike rides in the rain”) provide valuable semantic context.

4. Be the technical reference

For technical or B2B products, publish white papers, detailed comparisons and case studies. If your content is used as a training or reference source by language models, your brand can gain organic visibility in the generated answers.

Two typical scenarios

To make this concrete, here are two situations imagined for the example.

A mountain shop in Sion

Take a sports shop in Valais that also sells online. Its product pages reuse the manufacturer’s texts, whereas its customers ask comparative questions: “Which jacket for a high-mountain race in summer?”.

We would start with detailed buying guides that cite its products but above all explain the technical criteria (waterproofing in Schmerber, RET breathability), then with complete Product markup on each page. What we then track: the queries that lead to the guides in Search Console, and how assistants answer questions about the product category, citing these pages or not.

A B2B software publisher in Lausanne

Imagine a Lausanne-based publisher of management software for SMEs that hides its pricing behind a contact form. When a business owner asks an AI to compare the solutions on the market, it has nothing precise to say about it.

The approach is to publish its pricing and features, with an honest comparison table against the alternatives. The AI then has clear data to draw on, and the prospect arrives with realistic expectations. Being cited is never guaranteed; we give ourselves the means to be readable.

E-commerce visibility: what to take away for your strategy

Moving from classic SEO to optimisation for AI does not mean throwing everything away. It is an evolution, an extra layer of intelligence to bring to your content.

Here are the concrete actions to prioritise for your company:

  1. Audit your existing content: is it written for robots, or does it genuinely answer people's questions?
  2. Strengthen your structured data: make the reading work easier for AI.
  3. Build on expertise: become the essential source of information in your niche, the one AI cannot ignore.
  4. Monitor your online reputation: reviews and external mentions feed the perception AI has of your brand.

The commercial objective does not change. The path simply requires more finesse, authenticity and technical skill, and it is better to prepare before AI answers start to carry weight in your sales.

Our approach to these topics is described on the AI and automation page. To look at how your shop appears in these answers today, let’s talk at a first conversation.

Sources

digital marketingdigital toolsSEO
Mehdi B.
About the authorMehdi B.

Co-founder of Smart Impact. Hooked on the web from the start, he launched his first project in 2006: an online music magazine still running today. With close to 20 years of SEO experience, a Swiss federal diploma in marketing and a solid geek streak, he turns clients' ideas, sometimes hazy ones, into concrete digital projects with his team.

LinkedIn
Further reading