Google Shopping doesn't use keyword targeting in the traditional PPC sense — you don't bid on specific keywords. Instead, Google matches your products to queries based on what's in your feed: primarily the title and description, secondarily the category and attributes. This means keyword research in Shopping is really feed content research: which words and phrases should be in your product titles and descriptions to match the queries your potential customers are actually searching?

Start with your Search Terms report

The Search Terms report in Google Ads (under Insights & Reports for Performance Max, or directly in the Search Terms tab for Standard Shopping) shows you the actual queries that triggered Shopping impressions. This is the most direct source of keyword intelligence for your feed because it tells you exactly which queries Google considers your products relevant for — and by implication, which queries you're missing.

Filter the Search Terms report for three distinct lists:

The queries you're NOT getting — finding impression share gaps

Search Terms only shows queries that triggered impressions. It can't show you queries that should trigger your products but don't, because your feed content doesn't match them. For this, you need external keyword research combined with feed gap analysis.

The process: use Google's free Keyword Planner (available in Google Ads) to research search volumes for terms related to your product categories. Look for high-volume queries that include attributes your products have but which aren't in your feed titles. Common examples:

Feed rule approach: systematic attribute injection

Once you've identified a keyword gap — "merino wool" is high-volume and your products contain merino wool but the titles don't say so — the fix is a Trajekt rule rather than manually editing every product title.

# Inject material terms from product attributes IF material CONTAINS "Merino" AND title NOT CONTAINS "Merino" THEN title = title + " — Merino Wool" IF material CONTAINS "Cashmere" AND title NOT CONTAINS "Cashmere" THEN title = title + " — Cashmere" # Inject fit terms from tags IF tag CONTAINS "slim-fit" AND title NOT CONTAINS "Slim" THEN title = title + " — Slim Fit"

This approach systematically injects keyword-valuable attributes across your catalogue. New products with those attributes automatically get the keyword injection on their first feed run. Existing products update on the next daily refresh.

Monitoring keyword adoption

After implementing keyword injection rules, track impression share changes for the affected product groups over 30 days. If the material term injection is working, products with the injected term should start appearing for material-specific queries that weren't triggering them before. The Search Terms report will show new query patterns that weren't present before the feed change.

Also monitor quality score equivalents in Shopping — Shopping doesn't show explicit quality scores, but changes in average CPC at stable bid levels indicate whether Google's relevance assessment of your products has improved. Lower CPCs at similar impression share suggests higher relevance scoring, which means your feed content is better matching the queries it's entering.

🌎
Trajekt Editorial
Product Feed & Ecommerce Specialists

Trajekt is a UK-built product feed management platform. Our editorial team covers feed optimisation, Google Shopping strategy, ChatGPT commerce, and ecommerce channel management for UK merchants and agencies.

Share: X / Twitter LinkedIn