Strong on Google’s AI, nearly absent on ChatGPT: the analysis that showed a manufacturer where to act.
How OTW Motion measured what AI assistants tell buyers about an established manufacturer’s product line, and turned it into a specific list of where to act.
One industrial product line, August 25, 2026. Client anonymised.
See what we could do for your business →The challenge
An established brand cannot see what AI says about it.
This is an established manufacturer, and its team reasonably assumed AI assistants would describe its market the way the market does. They had no way to check. Sales figures and search rankings do not show which brands an assistant recommends when a buyer asks, and the answer differs from one assistant to the next. We do analysis work for this manufacturer.
What we did
We measured the answers the way a buyer meets them, then traced where each one came from.
- 01Buyer-question design. Four questions covering the path to purchase for one industrial product line: choosing a brand, choosing a supplier, buying in volume, and weighing the product against the older alternative.
- 02Repeated sampling across assistants. Each question put to Google’s AI Overview and to ChatGPT eight times on the same day, because one answer from an assistant is an anecdote and sixty-four are a reading.
- 03Brand and competitor detection. Every answer read for whether it named the brand, and which competitors it named instead.
- 04Source analysis. Every page each assistant cited, counted and attributed, to show whose content the answers were built on.
- 05A briefing the team can act on. Findings by assistant, by question and by source, delivered as a shareable report.
The findings
Strong on Google’s AI. Nearly absent on ChatGPT.
Answers that named the brand. Same four questions, same day.
Whose website ChatGPT cited. One competitor’s site appeared more than twice as often as the brand’s own.
Where in the purchase the brand goes missing. Answers naming the brand, by question, both assistants.
August 25, 2026. Four buyer questions, each asked eight times of each assistant, 64 answers.
Why it matters
The team now knows where to act, and where not to spend.
On Google’s AI the brand is already recommended in about two answers out of three, so that position needs defending, not rebuilding. The opening is ChatGPT, where buyers asking the same questions almost never hear of it, and the earliest stage of the purchase, where a buyer is still choosing what kind of product to buy. The analysis also names the competitor pages the brand’s own content has to match. That turns a general worry about AI into a short, specific list of work.
Notes on the numbers
- This is a reading of one product line on one day, not a trend. No improvement is claimed; our work here is analysis.
- Citation counts show which pages each assistant pointed to. They do not prove why an assistant chose its answer.
References
The client has agreed to be contacted for a reference check on request. Contact OTW Motion and we will arrange it.
Do you know what AI tells your buyers about you?
Most established brands have never seen it measured. We show you where you are recommended, where you are missing, and whose pages are winning the answer instead.