5 Google Shopping mistakes that are quietly draining your budget

Abstract

Google Shopping campaigns are deceptively easy to set up badly. Five specific, avoidable mistakes account for most of the wasted spend we encounter when auditing client accounts - and most of them have nothing to do with bidding strategy. If your Shopping campaigns have been running for more than six months without a proper audit, there is almost certainly budget going somewhere it shouldn't be.

Hypothesis

Most non-specialist Google Shopping campaigns are wasting a meaningful portion of their budget on five avoidable mistakes - none of which are related to bid strategy, and all of which are fixable with access to the account and a spreadsheet.

Required Lab Equipment

  • A Google Ads account with Shopping campaigns running
  • Access to Google Merchant Center
  • At least 30 days of campaign data
  • A willingness to actually open the search terms report
  • Honest assessment of whether your product feed was set up by someone who knew what they were doing
  • Coffee, probably (well, tea, really, seeing as we're in Britain)

Observations

  • Ignoring the search terms report. Google Shopping doesn't run on keywords; it runs on product feed attributes. But it does trigger on search terms, and those are recorded in the search terms report. Most accounts audited have been running for months with no negative keywords added. That means the campaign has been showing ads for irrelevant queries - early research traffic, competitor brand names, nonsense edge cases - for as long as it's been live. Per a Disruptive Search Advertising report, up to three-quarters of ad spend is being wasted on bad terms If you haven't looked at your search terms report in the last 30 days, money has been leaving without permission.
  • Weak product titles in the feed. Google uses the product title as its primary signal for matching ads to queries. "Blue Dress" and "Women's Cobalt Wrap Midi Dress UK Sizes 8–22" trigger very different auctions. Far too many SME product feeds were set up with the default CMS product name - written for a website navigation menu, not for a search-triggered ad. The titles look fine on the website. They are actively costing money in Shopping. Correct titles on your products can boost CTR by 15-30% and conversion by 10-20% Your product title is your Shopping ad's keyword proxy. If it's vague, you're losing relevance before the bidding starts.
  • No campaign structure - all products in one campaign. A single Shopping campaign with all products in one ad group means no control over budget allocation by category, margin, or performance tier. Your highest-margin products share a budget pool with your lowest-margin ones. Your bestsellers compete internally with dead stock that hasn't sold in six months. No structure means no leverage. You can't prioritise what you haven't segmented.
  • Performance Max doing everything. Many accounts were migrated to Performance Max (Google's black-box combined campaign type) without a full understanding of what that involved. PMax consolidates Shopping, Display, Search, YouTube, and Gmail into one campaign with minimal transparency. It frequently cannibalises branded search traffic-conversions you'd have generated anyway from people who already knew your brand-and credits them to the campaign, inflating apparent ROAS. Optmyzr studies found that over 90% of campaigns suffer from Keyword Overlap between Search and PMax and that Search outperformed PMax in CTR and conversion more often than the reverse PMax isn't wrong in every context. But it should be a deliberate, informed choice - not a default you inherited from someone else's setup.
  • Never excluding any products. If you have 200 SKUs in your feed and have never excluded a single one from Shopping, some of those products are being advertised at a loss. Low margin, high return rate, zero search demand - any one of these makes a product a budget drain. Most accounts we see have never removed anything. An ad campaign that runs on every product you stock isn't a strategy. It's a lottery with your budget as the stake.

Experimental Results

The thing about Google Shopping is that it looks like it's working even when it isn't. Clicks come in. ROAS looks acceptable. The reports show green numbers. What the reports don't show is how much better those numbers could be with two hours of housekeeping.

The five mistakes above share a common thread: they're all errors of omission, not commission. Nobody set these campaigns up wrong on purpose. They were set up quickly - with the immediate goal of getting something live - and then left to run. The machine kept going. The budget kept flowing. The problems compounded quietly behind a reassuring dashboard.

The search terms issue is the clearest example. Shopping campaigns don't use keywords, so many advertisers assume there's nothing to manage. There is. The search terms report shows exactly what queries triggered your ads, and it's often a catalogue of things you'd never consciously pay for: informational queries from people at the very beginning of a research journey with no purchase intent, competitor brand names your ads had no business appearing for, oddly specific mismatches that have somehow accumulated into a meaningful share of monthly spend. Adding negative keywords to a Shopping campaign isn't an advanced tactic. It's routine maintenance - the equivalent of cleaning the filters.

Product feed quality is the less visible version of the same problem. If Google's algorithm is trying to match your products to queries and its primary signal is a vague, generic title, you're starting every auction at a disadvantage before the bid is even considered. A feed audit focused specifically on titles, descriptions, and GTINs often produces improvements in impression share within weeks - without touching bids.

The structural and Performance Max issues take more effort to fix, but the underlying principle is the same: you cannot make intelligent decisions about budget allocation if you cannot see where the budget is going. Performance Max's opacity is by design, not by accident. Google has a financial interest in the black box remaining black. Your interest is the opposite.

The good news is that none of these fixes require specialist ad expertise. They require time, access, and the willingness to actually look at the data rather than at the headline ROAS figure.

Sources of Error

  • The ROAS trap. An acceptable return on ad spend is not the same as an efficient campaign. You could be generating the same revenue for considerably less spend. ROAS tells you what happened. A proper audit tells you why.
  • Changing bids before fixing foundations. If the search terms report is contaminated and the feed quality is poor, adjusting bids is rearranging the problem rather than solving it. Foundation before optimisation.
  • "It was set up by the platform." Google's Shopping campaign setup wizard is optimised for campaign launch speed, not campaign efficiency. Following the recommended setup is not the same as a well-structured account.
  • Assuming Performance Max is equivalent to Smart Shopping. It isn't. PMax has materially different reach, transparency, and behaviour. If you inherited PMax campaigns and haven't reviewed them since migration, you should.

The Full Picture

Start with the search terms report. Add negatives. Then look at the feed titles for your top-spending products. Then ask whether your campaign structure gives you any meaningful control over where budget flows. Then decide - actively, with evidence - whether Performance Max is working the way you assume it is.

None of this requires a specialist engagement. It requires a few hours with the account and the willingness to be honest about what you find.

If any of this has prompted the "should someone else look at our Shopping account?" question - that's exactly what a Quantised audit covers. We'll tell you what's leaking, what's worth keeping, and what to fix first.

Citations

Back to blog

Let's get you Quantised!