You're Not Overspending on Ads.
You're Underspending on Feed.
When ROAS drops, the instinct is to cut budget. But the algorithm isn't the problem. The data you're feeding it is.
The Misdiagnosis
ROAS is declining. CPA is rising. The response is almost always the same: cut budget, tighten targets, pause campaigns. It feels rational. Spend less, waste less.
But that logic assumes the problem is volume. It rarely is. The problem is almost always signal quality - the data you're giving Google to work with. You're not buying too many clicks. You're sending them to weak listings.
The Pattern
Brand cuts budget → volume drops → ROAS stabilises briefly → then declines again because the underlying feed quality hasn't changed. The cycle repeats. Each round, the account gets smaller. The feed stays the same.
Cutting budget treats the symptom. The disease is upstream.
Titles Are Targeting
In Shopping, your product title is your keyword strategy. Google matches search queries to titles. If your title says "Summer Top," you're competing against every vague query imaginable. If it says "Linen Relaxed Fit Shirt - Sage Green - Women's," you're matching specific, high-intent searches.
This isn't a marginal difference. Title optimisation routinely changes which queries you appear for, which changes click quality, which changes conversion rate, which changes ROAS. The entire performance chain starts here.
What Good Titles Do
They pre-qualify the click. A specific title attracts shoppers who already know what they want. A generic title attracts browsers. You pay the same CPC for both. One converts at 4x the rate.
Most accounts have never systematically tested title structures. They're spending five figures on traffic and sending it through a title written by a warehouse manager three years ago.
Pricing Is Eligibility
Google knows your price. It knows your competitors' prices. When your product is meaningfully more expensive than alternatives with similar attributes, the algorithm suppresses you before you even see the impression data.
This isn't a conspiracy. It's rational: Google's objective is to show products users are likely to buy. If your price signals "unlikely to convert," you don't get shown. No amount of budget fixes a pricing problem.
The Pricing Trap
You can be premium - but your feed needs to justify it. Brand authority, reviews, unique attributes, and imagery all signal value. Without those signals, a higher price just looks like a worse deal. Google knows. Shoppers know. Your impression share reflects it.
Imagery Is the Ad
In Shopping, Discovery, and YouTube placements, your product image is the creative. There's no headline to save a bad image. No description to add context. The image either earns the click or it doesn't.
Low-quality photography, inconsistent backgrounds, poor lighting, or images that don't show the product clearly all suppress CTR. Lower CTR feeds back into Google's quality signals, further reducing impression share. It's a compounding problem.
The Image Test
Put your Shopping listing next to your top three competitors. If your image is the one you'd skip, so is your customer. This isn't subjective - CTR data confirms it. Image quality is measurably a conversion rate problem.
The Compounding Effect
Feed quality compounds. A strong title matches better queries. Better queries attract higher-intent shoppers. Higher-intent shoppers convert at better rates. Better conversion rates improve Quality Score proxies. Better quality scores reduce CPCs. Lower CPCs improve ROAS.
The reverse compounds too. A weak title matches poor queries. Poor queries attract browsers. Browsers bounce. Bounces worsen quality signals. Worse signals raise CPCs. Higher CPCs destroy ROAS. And you blame the budget.
"Every pound you add to budget without fixing the feed is a pound that compounds in the wrong direction."
This is why feed optimisation isn't a "nice-to-have" project for Q3. It's the single highest-leverage investment most ecommerce brands can make in their paid media performance.
The Real Budget Question
The question isn't "are we spending too much on ads?" The question is "are we spending enough on the thing that makes ads work?"
Most brands spend £20k+ per month on Google Ads and nothing on feed quality. No title testing. No competitive price monitoring. No image refresh cycle. No attribute enrichment. The feed was set up once, during onboarding, and hasn't been touched since.
That's like spending £20k on petrol and never servicing the engine. The fuel isn't the problem. The engine is choking on bad inputs.
The Upstream Investment
Redirecting 5-10% of ad spend toward systematic feed improvement typically returns more ROAS improvement than the equivalent budget increase ever could. You're not spending more. You're spending on the right layer.
Next Steps
If your response to declining ROAS is always "cut budget," you're solving the wrong problem. The feed is the foundation. Fix it first. Then decide whether budget is really the issue.