Came across this article called How We Built & Optimized Google Shopping For 192% More Revenue by Ed Leake, founder of PPC agency Midas Media.
In this article Ed goes through the Search Terms Report in Google Shopping and shows you how to optimize your shopping campaign accordingly.
He takes you through different types of search intent: generic, brand, and specific. Then takes you through each associated mindset, with the specific searchers being the ones most likely to buy now.
He then sets larger bids & lower priority for the more specific search queries, which goes against conventional wisdom of bidding less for long tail searches because they have less competition. The result is a layered campaign structure that gets more and more aggressive as the searches get more and more specific.
This is a really interesting idea and makes a lot of sense. Length of search query is an approximation for specificity, so that would have to be your starting point, but I wonder how much more sophisticated you could get in evaluating the specificity of a search term and therefore how much to bid for it.
If you could identify the attributes that are being searched for, I think you'd have a much better sense of actual search term specificity. For example, a search query may be short but specific, e.g. "Dogtra 3502NCP." This is short, but its attributes are Brand + Part Number which are highly specific. I would rank that as more specific than a search query like "Electronic Training Dog Collar" (which is a specific Product Type + Animal Type + Key Feature). And that search query is more specific than "Pet Training System for Dog" which is a less specific Product Type + Animal Type).
Our technology allows us to automatically parse search queries in this way & come up with these attribute patterns. If the patterns can be scored based on the degree of specificity, then we have a great start toward gauging the searcher intent of each search.