Stop Blaming the Budget: What’s Actually Killing Your Google Shopping Performance
Let’s be honest — Google Shopping looks simple on the surface. You upload your products, set a budget, let PMAX do its thing and wait for the sales to roll in. Except they don’t. The ROAS is underwhelming, the clicks aren’t converting and you’ve got a vague sense that money is disappearing somewhere but you’re not quite sure where.
If that sounds familiar, you’re not alone. It’s one of the most common situations I see when auditing Shopping accounts, and nine times out of ten the problem isn’t the campaign. It started long before that — in the feed, on the product page, or in tracking that’s been quietly misfiring for months.
Here’s what’s actually going wrong and, more importantly, what to do about it.
1. Your Feed Titles Are Doing the Bare Minimum
This is where most Shopping problems begin, and it’s also where most people invest the least amount of effort. Your product feed title isn’t just a label — it’s how Google decides which searches your products are eligible to appear for. Get it wrong and you’re invisible to the people who are actively looking for what you sell.
The most common mistake? Pulling feed titles straight from the website and leaving them as is. That works fine for navigation. It doesn’t work for search. A title like “Classic Tee — Blue” tells Google almost nothing. But “Men’s Classic Fit Crew Neck T-Shirt Blue Cotton — S to 3XL” gives it everything it needs to make a decent match.
“Your feed title is your biggest keyword opportunity in Google Shopping. If it isn’t optimised, you’re essentially bidding blind.”
Think about how your customers actually search. They don’t search for your internal product names or SKU codes. They search for what the product is, what it’s made of, what size it comes in and who it’s for. Your feed titles need to reflect that.
A few things worth doing right now:
- Lead with the most important information — brand, product type, key attributes in that order
- Include size, colour, material and gender where they’re relevant search modifiers
- Keep feed titles under 150 characters but make every word earn its place
- Test title variations using a supplemental feed rather than editing your source data directly
- Stop burying the most important details at the end — Google weights the first few words more heavily
It sounds like a small thing. It isn’t. Better feed titles mean better match rates, better placement and better traffic quality before you’ve touched a single bid.
2. The Rest of Your Product Data Probably Needs Work Too
Feed titles get most of the attention but the surrounding product data is just as capable of quietly tanking your performance. Google Shopping runs on structured data and the more accurate and complete yours is, the better the whole thing works. The more gaps and errors there are, the more Google either limits your reach or pulls your products from auction altogether.
Disapproved products are the obvious issue — you can see those in Merchant Center. But it’s the subtler stuff that tends to do the most damage over time. Vague product categories. Missing GTINs. Product descriptions that read like they were written by someone who’d never seen the product. Images that are technically acceptable but not actually convincing.
“A feed with 30% disapprovals isn’t a campaign structure problem. It’s a product data problem. And you can’t bid your way out of a product data problem.”
The things that tend to get overlooked:
- GTINs need to be accurate — Google cross-references them against manufacturer data and mismatches cause disapprovals
- Price and availability must match your website in real time, not just at the point of upload
- Product categories should be as specific as possible, not just the top-level option
- Product descriptions are a missed opportunity for most retailers — use them, make them detailed and write them for a person not a crawler
- Custom labels are massively underused — segment by margin, bestseller status, seasonality or sale items so you can control where budget goes
Set a reminder to run a feed audit every month. It’s not the most exciting task in digital marketing but every fixed error is a product back in the auction.
3. PMAX Is Not a Strategy On Its Own
Performance Max campaigns get a lot of praise and a lot of criticism, often from the same people. The truth is it can work brilliantly or burn through budget with very little to show for it, and the difference usually comes down to how it’s set up.
The most common mistake is running one PMAX campaign across the entire product catalogue with no real campaign structure behind it. No segmentation by product type. No asset groups organised by theme. No separation between high-margin products and clearance stock. Just everything lumped together and handed over to Google’s automation to figure out.
“Handing Google a single PMAX campaign with no structure is like giving someone your entire stockroom and asking them to run your shop. They’ll sell something, but probably not what you’d have chosen.”
What better campaign structure actually looks like:
- Separate PMAX campaigns by product category, margin tier or intent where your budget allows
- Use asset groups properly — organise by audience signal and product theme, not just one group for everything
- Add audience signals that mean something: customer lists, website visitors, high-intent custom segments
- Use campaign-level negative keywords to stop irrelevant traffic eating your budget
- Consider running a standard Shopping campaign alongside PMAX for your top performers — it gives you more bidding control where it matters most
- Use listing groups to separate bestsellers, high-margin lines and clearance so budget allocation reflects your actual priorities
The goal isn’t to fight Google’s automation. It’s to give it enough to work with so it makes the right calls rather than expensive ones.
4. The Landing Page Is Where Shopping Clicks Go to Die
You can have a perfectly optimised feed and a well-structured campaign and still haemorrhage money if the landing page doesn’t convert. Google Shopping traffic is high intent — these people have already seen your product, your price and your image before they click. They arrive closer to buying than almost any other traffic source. If the landing page doesn’t hold up its end of the deal, they leave.
Most product pages were built for browsing. They weren’t built for converting someone who’s already made up most of their mind. They load slowly, bury the important information, skimp on trust signals and make the path to purchase more complicated than it needs to be.
“Google Shopping traffic arrives with intent. The landing page either honours that intent or wastes it. There’s not a lot of middle ground.”
The landing page fixes that actually make a difference:
- Match the landing page to the ad — if your Shopping ad shows a blue size 10 trainer, the page should load on that variant
- Get the price, availability and add-to-cart button above the fold — don’t make someone scroll to buy
- Add trust signals where they’re visible: reviews, star ratings, delivery times, returns policy and payment options
- Write product descriptions for people, not spec sheets — explain why it matters, not just what it is
- Take page speed seriously — a one second delay can reduce conversions by up to 7%
- Mobile experience is non-negotiable — a significant portion of Google Shopping traffic arrives on a phone
Run your key landing pages through PageSpeed Insights and throw Microsoft Clarity on the site if it isn’t already. Heatmaps and session recordings will show you exactly where people drop off and that’s worth more than guessing.
5. If Your Tracking Is Off, Everything Else Is Guesswork
This one doesn’t get talked about enough. Smart Bidding — including everything inside PMAX — is entirely powered by conversion data. Feed it accurate conversion data and it learns fast and optimises well. Feed it bad data and it optimises confidently in completely the wrong direction. Feed it no data and it spends your budget on its own education.
Broken conversion tracking is more common than most people realise. Conversion actions firing on the wrong pages. Duplicate tags inflating reported results. GA4 and Google Ads telling completely different stories. Micro-conversions like add-to-cart being weighted the same as actual purchases. All of it distorts the picture and quietly trains your campaigns to fail.
“Smart Bidding is only as smart as the conversion data you give it. Bad conversion tracking doesn’t just affect your reports — it actively points your campaigns in the wrong direction.”
What to check before anything else:
- Audit your conversion actions in Google Ads and make sure you’re optimising toward purchases only, not every micro-event on the site
- Check for duplicate conversion tracking — a GA4 imported conversion plus a Google Ads tag on the same page will over-report
- Set up enhanced conversions if you haven’t already — it plugs the gaps caused by ad blockers, cookie consent and browser restrictions
- Make sure your conversion value reflects actual revenue, not a static placeholder figure
- Use Tag Assistant regularly, not just at the point of setup
Sort your conversion tracking before you think about scaling spend. Every pound going into a campaign with broken data is a pound being optimised for the wrong outcome.
6. Wasted Spend Is Hiding in Plain Sight
Even well-run Google Shopping accounts accumulate wasted spend over time. Search terms that never convert. Products that eat budget without contributing revenue. Audiences that click regularly and buy never. It adds up faster than most people expect and it tends to go unnoticed until someone actually sits down and looks for it.
The search terms report is always the first place to start. Look for queries that are pulling clicks but no purchases, terms that are too broad to be worth showing for, and competitor brand terms you’re appearing against with no real reason to be there.
“Most Google Shopping accounts have around 20% of products driving 80% of revenue. The other 80% are often just consuming budget that could be working a lot harder somewhere else.”
Finding and cutting wasted spend:
- Review your search terms report weekly and add negative keywords consistently — this is ongoing maintenance, not a one-time job
- Use custom labels to segment products by ROAS or conversion rate and adjust bids by performance tier
- Pause products with significant spend and zero conversions over a 30 to 60 day window — they’re not about to turn around on their own
- Check impression share data — it’ll tell you whether underperformance is a budget issue or a relevance issue
- For PMAX, use the asset group performance breakdown to see which combinations are working and which aren’t
- Watch your Target ROAS threshold — set it too high and you’ll starve the campaign of traffic; too low and you’re buying unprofitable clicks
Wasted spend reduction isn’t a project with an end date. Build it into your weekly routine and you’ll consistently find budget to reinvest without needing to increase what you’re spending overall.
So Where Do You Start?
The honest answer is: wherever the biggest problem is. For most accounts that’s the feed, because everything else flows from it. Get the product data right and the campaign has something worth working with. Build proper campaign structure and the automation makes better decisions. Fix the landing page and the clicks start converting. Get the conversion tracking right and Smart Bidding learns fast. Cut the wasted spend and the budget works harder across all of it.
None of this is particularly complicated. It just requires actually doing it — methodically, consistently and without waiting for the campaign to magically sort itself out.
If you’ve been staring at a flat ROAS and wondering what you’re missing, there’s a reasonable chance the answer is somewhere in this list.
Recent Posts


