Artificial intelligence isn’t just changing the way people work, crunch data or maintain trucks. It’s also changing the way customers make purchasing decisions.
Here’s an example. Searching in Google for an EGR cooler replacement digs up an AI Overview explaining how the part is replaced and a range of total cost. After a bit of a scroll, Google then displays a range of parts, followed by the more traditional list of links. It’s changing the way people use and peruse the internet.
AI Overviews and other generative AI searches are changing the landscape for e-commerce sites.Beth Colvin
Greg Tull, marketing director at Classy Llama, a Missouri-based e-commerce services agency, says when Google flipped the switch on AI Overviews last year saw impressions go up but clicks go down. Industry experience backs that up.
[RELATED: Procede shares its AI journey, and the potential it offers users]
“Gartner’s prediction of a 50% drop in traditional organic traffic … validated what we began seeing in our own data,” says Gary Willis, vice president of marketing and communications at Rush Enterprises. “Because our SEO fundamentals were already strong, including a robust content strategy and a sound technical framework, we started noticing ‘dual visibility’ in our metrics: Impressions from both AI Overviews and traditional organic search results. That was the clear signal that discovery behavior was changing.”
Building an AI search vocabulary
As with any innovation, there’s some new terminology to master. One of the most important is the zero-click search. That is where the searcher finds what they’re looking for without having to click away from the search engine results page (SERP).
“Click-through rates from AI Overviews (AIOs) are extremely low, typically between 1-3%” says Gavin Duncan, vice president of digital marketing at Epicor. What that means is searchers aren’t necessarily finding companies’ carefully crafted e-commerce sites. “Keyword listings and paid search have taken a hit, and our click-share data reflects this trend.”
[RELATED: Thinking about opening a new location? Why not online?]
Because of this trend, there’s another new term. Instead of search engine optimization (SEO), companies are focused on GEO, generative engine optimization. This means making your site more attractive to AI agents such as Google’s Gemini, ChatGPT from OpenAI and Microsoft Copilot.
“One of the best things we’ve done was partner with industry experts in SEO and AI search to refine our strategy,” Willis says. “Together, we’re continuously improving our content and identifying optimization opportunities for our website so our brand remains highly visible and competitive in AI-driven search, especially within the heavy- and medium-duty truck market.”
Tull says there also are small but growing visits to sites directly from AI search engines. These visits often come with higher engagement than traditional search traffic, but again, making the site appealing to the AI agent is key.
Navigating a new search landscape
It means without action, e-commerce sites may continue reaching current customers who navigate directly but might miss new contacts from keyword searches. The goal with optimizing sites for AI agents is to build authority in a meaningful way, not just with lines of keywords.
“AI-driven search is transforming how people find and consume information by presenting topics in a more contextual, conversational way,” Willis says. “That means we need to meet customers where they already are — in AI search engines. Integrating GEO into our strategy ensures we continue delivering the accurate, relevant, high-quality content our customers expect from a trusted, forward-thinking authority.”
[RELATED: Three questions to answer before diving into AI]
Tull says AI agents prefer structured, specific and verifiable content. Products should have clean, complete product data, including unique product IDs everywhere it appears, improving matching, citations and commerce surfaces. Part identifiers should be “unavoidable,” Tull says.
“AI search is redefining how authority and expertise are demonstrated online,” Duncan says. “Websites remain critical, but distribution models are shifting. Instead of relying on keyword density and backlinks alone, success now depends on creating authoritative, structured content that LLMs [large-language models, another term for AI agents] can easily interpret and cite.”
When possible, information should be standardized to make it easy for AI agents. Tull says using data from the Aftermarket Catalog Exchange Standard (ACES) or the Product Information Exchange Standard (PIES) helps. Make the most of that structured data, putting in offers and reviews, a frequently asked questions page, how-to information, and other timestamped data to make websites more attractive to AI agents.
“Our goal is to make sites more ‘agent-friendly’ by emphasizing depth, relevance and semantic richness over sheer keyword density,” Duncan says. “Authority and citations matter more than keywords.”
Build that authority with an answers library that has scannable headings, numbered steps, spec tables and summaries. Pair on-page structured data with feeds that allow Google to corroborate prices, availability and specs, increasing accuracy. Monitor activity such as intent, watching for patterns and tracking referral sources and citations.
“The goal of traditional SEO was to earn a place in search listings and drive clicks/visits to your website,” Willis says. “AI search turns all of that upside down. It’s about being recommended, not just listed because your website was optimized for the right keywords. While Google isn’t entirely transparent about when an AI Overview will appear for a given search, question-based searches with informational intent tend to trigger them most. Searches with transactional or commercial intent seem to return fewer AI Overviews or none at all (not yet, at least).”
Where’s the good news?
Right here. Even though zero-click searches and AI-led searches may generate fewer clicks, the ones they do delivery are golden.
“Semrush reports that AI-search visitors convert at 4.4 times the rate of traditional organic visitors, which warrants a re-examination of the metrics and KPIs used to evaluate performance and ROI,” Willis says. Tull says Adobe saw AI-referred shoppers stay longer, view more pages and bounce less.
And here’s some more. Even though the vocabulary is different, the goal remains the same.
“We are all still competing to have our brands appear in searches more often than our competition,” Duncan says. “Adapting to change has been a constant in the SEO space since forever, and that is what we are doing now.”
Willis says a strong technical foundation helps, as does working with trusted partners. Testing and learning also are essential.
“Every business is different in how it defines success,” he says. “We’re seeing positive results in our work, but would highly encourage our peers in the industry to do their own research so they can determine how generative AI affects their specific audiences and what works for their website, technical architecture and strategic goals.”
