The highest-intent search in automotive is a string of digits

A shopper who types a category phrase is browsing. A shopper who types a part number has already been under the car, already read the number off the old component, and is buying within minutes. That second search is the most valuable query in the category and it is also the cheapest, because most brands never think to compete for a string that is not a word.

This brand published its cross-reference data in a PDF on its own website and nowhere else. Its Amazon listings carried the brand's internal SKU and nothing more, so an interchange search returned three competitors and a marketplace reseller of the original part — never the brand that had spent two years engineering an equivalent.

We treated the cross-reference table as a keyword list, because that is what it is.

• Interchange data extracted and normalized — OE numbers and their supersessions for all 340 SKUs, from the brand's own engineering references.
• Numbers placed where the parts catalog accepts them — the OE and interchange attribute fields plus a cross-reference table in the copy, not stuffed into generic backend keywords, where another manufacturer's identifiers do not belong.
• Number formatting variants covered — with dashes, without dashes, with and without the manufacturer prefix, since shoppers type all three.
• Exact-match campaigns on the numbers themselves, which is where the bargain sits: almost nobody bids on them and the intent is close to absolute.

Revenue attributable to part-number searches nearly tripled, from $21k to $60k a month, and total revenue grew 51%.

Those searches convert at 22% — about three times the account average — at a cost per click of $0.44 against a blended $1.36, because the brand is bidding against almost nobody. The data required no research: it had existed in a PDF on the brand's own site since 2019.