Negative Keyword Strategy for Standard Shopping

Why Shopping negatives differ from Search negatives

Standard Shopping does not use keywords, so negatives are the only lever you have to tell Google what a product listing should not match. That single fact changes the workflow. There is no ad-group-level negative option - the smallest scope is the campaign - and negative match types behave more loosely than they do in Search because Google is matching against the query, not against a keyword bid.

Three practical consequences: (1) precision comes from shared lists applied consistently across every Shopping campaign, not from ad-group hygiene; (2) phrase and exact negatives are the workhorse - broad negatives will strip out too much traffic; and (3) when you run the 3-tier priority structure, negatives are the mechanism that keeps queries flowing into the correct tier. Get the negative strategy wrong and the tiering collapses within a week.

This is also why teams who transplant a Search negative list straight into Shopping tend to be disappointed. Search negatives are usually built around exact commercial-intent phrases layered on top of exact-match keyword targeting; Shopping has no keyword layer to lean on, so a Search-style negative list leaves enormous gaps in the wrong-intent and competitor categories while over-blocking legitimate long-tail queries elsewhere. The fix is not more negatives - it is the right negatives, organised by category rather than bolted on ad hoc.

Google's own documentation on negative keywords confirms Shopping supports campaign-level and shared-list negatives only. Everything below assumes you are working within that constraint. If you have not yet built the tiered campaign structure this article assumes, read our companion piece on Standard Shopping priority tiers first - negatives without tiers are still useful, but negatives with tiers compound.

The weekly query-mining routine

This is the 20-30 minute weekly workflow we run on every eCommerce client. It typically surfaces 5-15 new negatives per session for a mid-sized account (£5k-£30k monthly Shopping spend).

1. Open the Search Terms report at account level, filter to the last 14 days, and segment by campaign.
2. Sort by cost descending, then filter conversions = 0. Anything spending >£20 with zero conversions across a two-week window is a candidate.
3. Classify each term into one of the five named categories below (competitor, wrong-intent, tier-bleed, out-of-stock, non-buying region). If it fits none, leave it - it is genuine top-of-funnel exploration.
4. Add to the correct shared list, not the campaign. One shared list per category makes future audits trivial.
5. Sense-check spend impact after 7 days. Impression Share Lost (budget) should not spike; if it does, a negative is too aggressive and needs narrowing to phrase or exact.

Accounts under £5k/month can drop to a fortnightly rhythm without meaningful cost leakage. Anything above £30k/month benefits from a mid-week spot-check on Wednesdays.

The routine compounds. Month one is the heaviest - you are clearing years of accumulated waste in a single sweep - and it is common to find 60-100 candidate terms in the first session on an account that has never had a dedicated negative review. By month three, weekly sessions settle into the 5-15 range because the shared lists are catching repeat offenders automatically before they can spend meaningfully.

Wasted spend before and after implementation

Across five client accounts where we introduced the full five-category negative structure, 30-day wasted spend (cost attributed to zero-conversion queries later excluded) fell sharply within the first month:

Average reduction across the five accounts was roughly 76-80% of previously wasted spend, reallocated to converting queries within the same campaigns rather than removed from budget entirely. These figures are illustrative aggregates from consented client work and will vary by category mix, seasonality, and starting negative hygiene.

Category 1: Competitor terms

Competitor brand names attract clicks with low intent to buy from you. Someone searching "adidas trainers" while your Shopping ad appears with a Nike listing converts at a fraction of the rate of a category-level query. Add competitor brand terms as phrase negatives to your generic (Tier 2) and catch-all (Tier 3) campaigns - never to your brand campaign (Tier 1), which should be the only one bidding on your own name.

Build the list from three sources: your own market knowledge, the "Auction Insights" report (competitors who impression-share you regularly), and the Search Terms report itself once negatives are seeded. Expect 20-80 competitor terms for a typical UK retailer. Review quarterly for new market entrants.

One nuance worth flagging: retailer names and marketplace names ("amazon", "argos", "very") belong in this category too, not just direct brand competitors. A query for "[your product] amazon" is someone comparing price on a marketplace, and it converts on your Shopping ad at a much lower rate than a query without a marketplace qualifier attached.

Category 2: Wrong-intent queries (research, jobs, wholesale)

This is the highest-volume category and usually the highest-waste. Split into four sub-lists so future audits are readable:

• Research intent: "how to", "what is", "vs", "review", "comparison", "guide", "tutorial", "diy".
• Employment intent: "jobs", "career", "salary", "vacancies", "hiring", "recruitment".
• B2B / trade intent: "wholesale", "bulk", "trade", "supplier", "manufacturer", "distributor" - unless you actively serve B2B, in which case route these to a dedicated campaign instead.
• Free / cheap intent: "free", "cheap", "discount code", "coupon", "voucher" (keep "sale" and "offer" - those are commercial).

Use phrase match - exact will miss variations, broad will over-strip. A typical starter list here contains 60-100 terms. This category alone typically accounts for a third to a half of all wasted spend uncovered in a first-time audit, because these queries pass every automated relevance signal Google uses - the product feed genuinely matches the words in the query - while carrying almost no purchase intent.

Category 3: Priority-tier bleed prevention

If you run the 3-tier structure from our Standard Shopping priority guide, negatives are the mechanism that keeps each query in the correct tier. Without them, the auction routes based on bid alone and the tiers collapse.

Audit this monthly. New product categories on the site should be added to Tier 3 negatives the same week they appear in Tier 2. The most common failure mode we see is a client launching a new category, adding it correctly to the Tier 2 campaign, and forgetting to exclude it from Tier 3 - the two tiers then compete against each other in the same auction, inflating CPCs for no benefit since both campaigns belong to the same advertiser.

Bleed is easiest to spot in the Search Terms report segmented by campaign: if a Tier 3 (catch-all) campaign is regularly serving impressions for a query that should clearly belong to Tier 2, the priority-tier negative list needs updating that week, not at the next scheduled monthly audit.

Category 4: Out-of-stock query catch-all

Permanently discontinued product lines still attract queries for months, sometimes years. Every click on a "product not found" landing page is wasted spend and a small SEO trust ding. Once a range is dropped, add the model name and its common variations as phrase negatives across all Shopping campaigns.

Separate this from temporarily out-of-stock lines - those will return, and negatives are hard to spot-check when reintroducing. Keep a dated shared list called "Discontinued - permanent" and never merge it into other categories.

For seasonal ranges that return every year (Christmas jumpers, summer garden furniture), avoid negatives entirely and instead pause the relevant product group in Merchant Center - a negative keyword added in January and forgotten by October will silently suppress a range that should be live.

Category 5: Non-buying regions

UK-only advertisers still see US-specific queries slip through geo-targeting when users search from the UK for products they saw on holiday, or when VPNs are involved. Common offenders: "walmart", "target", "cvs", zip-code patterns, US-specific colour spellings ("gray" vs "grey" for premium brands), and imperial-unit qualifiers where you list metric.

This category is small - usually 15-30 terms - but has a high hit rate on wasted clicks because query intent is unambiguous. It is also the category most often skipped entirely by advertisers who assume geo-targeting alone handles it; geo-targeting controls where the ad is eligible to show, not what the underlying query was about, so the two problems need separate fixes.

Managing shared negative lists at account level

One shared list per category, applied at account level, is the operational standard. The alternative - piling every negative into a single "master" list - makes future audits impossible and blocks safe rollbacks. Name lists clearly: "Shopping - Competitors", "Shopping - Wrong intent (research)", "Shopping - Tier 2 brand block", and so on.

Apply lists to campaigns explicitly during setup. Google Ads does not auto-attach shared lists to new campaigns - a new Standard Shopping campaign added six months from now will inherit zero negatives unless someone attaches them. Add a "shared list attach" step to your new-campaign checklist.

The 5-category negative framework

Here is the full framework we apply to every eCommerce account, showing which tiers each category applies to and how often it needs reviewing:

Try it: query classifier

Paste search terms from your own Search Terms report below (one per line) and the classifier will suggest which negative category, if any, each term belongs to.

Starter negative list - 200+ terms

The starter negative list we seed into every new eCommerce account contains 200+ terms across the five categories above, drawn from patterns visible across 12 UK client accounts (fashion, homewares, sports nutrition, fine jewellery, tools). A sample list is downloadable below as a plain-text file, organised by category, ready to paste into a new shared list in Google Ads.

The full 200+ term list, updated monthly and tiered against the priority framework, is available on request from the audit form below - we prefer to hand it over with a 15-minute call to explain the tiering rather than publish a static version that goes stale.

You can build the same list yourself from your own Search Terms report inside four weeks of disciplined weekly mining, using the five categories above as the classification framework and the query classifier tool above to speed up manual review.

Common negatives mistakes

1. Broad-match negatives everywhere. Broad negatives strip too much. Default to phrase; use exact for surgical exclusions.
2. One giant "master" list. Impossible to audit, impossible to roll back safely. Split by category.
3. Adding negatives inside campaigns rather than shared lists. The negative gets orphaned the day someone rebuilds the campaign.
4. Blocking your own brand from Tier 2 without a Tier 1 brand campaign live. You lose the query entirely.
5. Not attaching shared lists to new campaigns. Every new Shopping campaign inherits zero negatives by default - bake attach into the setup checklist.

A sixth, softer mistake worth naming: treating negative-list maintenance as a one-off project rather than an ongoing routine. Accounts we have audited after a single "big clean-up" six months prior typically show wasted spend creeping back to 60-70% of pre-cleanup levels, simply because new products, new competitors and new seasonal queries are never captured without the weekly routine described above.