Semrush ran a manufacturing AI-search study on 458 keywords and found only two brands appear on both the most-mentioned and the most-cited top-15 lists

Semrush published an original study of AI search in the manufacturing sector this morning, run by Luke Harsel with data from Nikola Manojlović. The team tracked 458 manufacturing and industrial keywords from January through July 2026 across ChatGPT, Gemini, Google AI Mode and Google AI Overviews, and combined that with U.S. session data over ten manufacturing categories.

Being named and being linked are different outcomes

The headline finding is that a brand mention and a citation belong to different lists. Semrush defines a mention as "a brand's name showing up in the text of the AI answer" and a citation as "a link AI can point to as a source for the answer it generated." On the top-15 by each metric, only two names show up on both: Vevor and Grainger. The top-mentioned brands were 3M, John Deere and Boeing, which the study calls "names people already trust." The top-cited sources were "a mix of industry directories, brands, distributors, and reference sites." A team optimising a manufacturer's own domain for AI answers is chasing the second list, and the incumbents on the first list are largely riding brand recognition.

AI Overviews doubled their share of manufacturing queries

Across the same keyword set, AI Overviews grew from 38% of tracked search volume in January 2026 to 57% in July, a 19-point rise in seven months. The tracked categories were chemicals, civil engineering, construction and maintenance, equipment and supplies, machinery, manufacturing, plastics and polymers, renewable energy, transportation and logistics, and warehousing.

Traffic contribution is the other side of the same story. Semrush reports that "Just 0.48% of all sessions to manufacturing sites originate from AI search tools or assistants," while direct and organic search together account for close to 80% of sessions. The AI answer is showing up on more than half of the queries and delivering half a percent of the clicks.

The engineerfix.com finding

The study calls out one site that reads as a warning about routing an entire discovery strategy through Google. Semrush says engineerfix.com "dwarfs the other top sources on ChatGPT," but does not "crack the top-cited sources on any of Google's AI search platforms." Semrush Traffic Analytics puts the site at close to 300,000 monthly organic search visits. Of those, "DuckDuckGo, Bing, and Yahoo are responsible for over 68% of all traffic to the site. Meanwhile, Google accounts for less than 2%."

That is a site with near-invisibility on Google that is a top citation on ChatGPT, funded by traffic from the other three engines. Which search engine a source page ranks on turns out to influence which AI answer engine cites it, and Google-only visibility work leaves the ChatGPT surface unaddressed.

For a manufacturer running a discovery strategy, the three concrete facts to act on: mentions and citations are separate outcomes and need separate work; AI Overview presence on manufacturing queries has doubled in seven months but sends under 1% of sessions today; and a page's crawl profile on Bing, DuckDuckGo and Yahoo is a real factor in whether ChatGPT will cite it, even for a domain Google has effectively ignored.

Source

AI search & manufacturing SEO: What the data shows, by Luke Harsel with data by Nikola Manojlović at Semrush, 8 September 2026.