The Uncomfortable Premise
Most Google Ads troubleshooting starts with bids. ROAS too low? Adjust targets. Spend too high? Lower bids. Performance volatile? Try a different bidding strategy.
This is like adjusting your speed when you're driving in the wrong direction. You'll get there faster, but you're still going the wrong way.
Your product feed determines which searches you appear for. Your bids determine how aggressively you compete for those searches. Fix direction before speed.
This isn't about "feed hygiene" or filling in optional fields. This is about understanding that bidding optimisation happens downstream of a decision that's already been made: which auctions you're entering in the first place.
Matching Happens Before Bids
When someone searches on Google, Shopping campaigns go through a sequence. First, Google determines which products are relevant to the query. Then, among relevant products, it runs an auction.
Step 1: Query Matching
Google reads your feed data (titles, descriptions, attributes) to determine if your product is relevant to the search. This happens before any auction.
Step 2: Auction
Among products deemed relevant, Google runs an auction based on bids and quality signals. Your bid strategy only affects this stage.
Step 3: Ad Delivery
Winners show ads. Clicks happen. Conversions (or lack thereof) feed back into quality signals and bidding algorithms.
If your feed data causes Google to match your products to wrong searches, you're entering the wrong auctions. Bidding higher just means paying more for traffic that was never going to convert.
Bidding lower might reduce waste, but it doesn't fix the underlying problem. You're still appearing for wrong searches. You're just appearing less often.
Bad Data Creates Wrong Traffic at Any ROAS Target
We see this pattern constantly: a brand sets a 4.0 ROAS target, gets traffic that doesn't convert, so they raise the target to 5.0. Traffic drops. They lower it to 3.5. Volume returns, but so does waste. They try tROAS, then Maximise Conversion Value. Nothing stabilises performance.
The Pattern
If your feed shows a "Black T-Shirt" when you're selling an £85 merino crew neck, you'll appear for bargain-hunting searches. Those users want cheap. Your product isn't cheap. No bid strategy fixes this mismatch.
The problem isn't the bid. The problem is the traffic you're buying. Bad feed data attracts wrong-intent searches:
- •Vague titles expand matching to generic, low-intent queries
- •Missing attributes force Google to guess product characteristics
- •Wrong categories place products in auctions with mismatched buyer intent
- •Poor images suppress CTR and poison quality signals
You can set any ROAS target you want. If the traffic coming in has the wrong intent, you're just adjusting how much you pay for waste.
The Accelerator vs Steering Wheel Problem
Here's the analogy we use with clients:
Bidding is the accelerator. Feed data is the steering wheel.
You can press the accelerator as hard as you want, but if the wheel is pointed at a cliff, you're just getting there faster.
Most agencies spend their time on the accelerator. It's visible, measurable, and looks like work. Adjusting bids generates activity reports. Feed work is invisible until it's done right.
Accelerator (Bidding)
- • ROAS targets
- • Budget allocation
- • Bid adjustments
- • Campaign structure
- • Audience layering
Controls how aggressively you compete in auctions you're already entering.
Steering Wheel (Feed)
- • Product titles
- • Descriptions
- • Attributes
- • Categories
- • Images
Controls which auctions you enter in the first place.
The brands that win at Shopping aren't the ones with the cleverest bidding strategies. They're the ones whose feed data ensures they only enter auctions where their products are genuinely relevant.
When matching is right, bidding becomes simple. When matching is wrong, no bidding strategy saves you.
What "Good" Feed Data Actually Means Commercially
"Good" feed data isn't about filling in every optional field. It's about ensuring Google understands what you're selling and who should see it.
Titles That Qualify Traffic
Your title should repel wrong-intent searches as much as it attracts right-intent searches. "Premium Merino Wool Crew Neck Sweater" attracts different clicks than "Black T-Shirt."
Test: Would someone searching this title be disappointed by your price point?
Attributes That Narrow Matching
Material, size, colour, pattern, and custom attributes help Google understand exactly what you're selling. Missing attributes force Google to guess, and it often guesses wrong.
Test: Could Google distinguish your product from a cheap alternative based on feed data alone?
Images That Set Expectations
Product images are your first filter. They should communicate quality level, use case, and price bracket before anyone clicks. A premium product needs premium imagery.
Test: Does your image look like it belongs next to £20 alternatives or £200 competitors?
Categories That Place You Correctly
Wrong Google product categories mean competing in auctions with mismatched intent. A "Home & Garden" product categorised as "Toys & Games" enters completely wrong auctions.
Test: If someone filtered Google Shopping by your category, would they expect to find your product?
The commercial goal of feed optimisation isn't "more impressions." It's fewer wrong impressions. You want Google to show your products only to people whose intent matches what you're selling.
Who This Matters For (And Who Should Ignore It)
Feed optimisation matters most when there's a gap between what Google thinks you're selling and what you're actually selling.
This Matters For You If:
- You sell premium products at higher price points
- CTR is low despite competitive pricing
- Search terms report shows irrelevant queries
- Bidding changes don't stabilise performance
- You have 500+ SKUs with varied margins
You Can Ignore This If:
- •You sell commodity products at market rate
- •Search terms already match your products well
- •CTR and conversion rates are already healthy
- •You compete primarily on price
- •Feed is auto-generated and already matches products
The bigger the gap between your product and commodity alternatives, the more feed optimisation matters. If you're selling £200 products and appearing for searches that expect £20 results, no bid strategy saves you.
Feed optimisation is profit control, not housekeeping. It determines which customers see your products before you spend a penny on clicks. Get this right, and bidding becomes a refinement. Get this wrong, and bidding is just adjusting how fast you waste money.
Frequently Asked Questions
Why do bidding changes fail to fix Shopping performance?
Bidding controls how aggressively you compete in auctions, but it doesn't control which auctions you enter. Your product feed data determines query matching. If your feed shows products for wrong searches, bidding higher just means paying more for bad traffic.
What's the difference between feed optimisation and bid optimisation?
Feed optimisation controls direction: which searches your products appear for. Bid optimisation controls speed: how aggressively you compete. Most brands focus on speed while ignoring direction. You can't outrun a wrong turn.
How does bad product data affect Google Ads performance?
Bad product data causes three problems: wrong query matching (appearing for irrelevant searches), poor quality signals (low CTR and conversion), and wasted spend (paying for traffic that was never going to convert). Fixing bids doesn't fix any of these.
Related Reading
Want to See What's Actually Wrong with Your Feed?
We review your feed data, search terms, and query matching to show you exactly where the steering wheel is pointed wrong. No obligation. Just clarity.
Request a Feed AuditGet our insights in your inbox
Plain-English thinking about Google Ads. No spam, unsubscribe anytime.