Google Shopping isn’t dead. Lazy PPC strategy is.
Putting every product into Performance Max is not a strategy. Standard Shopping still gives commercially minded retailers the control to separate valuable searches from expensive, noisy traffic.
You could be forgiven for thinking Standard Shopping had been quietly killed off.
Speak to enough PPC agencies and the recommendation is increasingly predictable: put the entire product feed into Performance Max, give Google enough data and let the algorithm do the rest.
It is simple. It is scalable. It is also very convenient for the agency.
But convenient account management and good commercial management are not the same thing.
Automation is a tool. It is not the strategy.
Performance Max can absolutely work. This is not an argument for switching off automation or pretending a human can process auction-time data faster than Google. It is an argument against surrendering commercial judgement to a platform whose primary objective is to sell more advertising.
When every product, audience and stage of the buying journey is blended into the same campaign, the headline return can look healthy while hiding what is actually happening underneath.
Brand searches, repeat customers and high-intent product searches can flatter performance. Prospecting can consume spend without being held to the same commercial standard. Search, Shopping, YouTube, Display and other inventory can be combined into one number that is easy to report and difficult to interrogate.
What did we actually pay Google to influence?
A customer searching for a specific premium appliance by brand and model is not the same as somebody searching for a broad inspiration term. The intent, conversion rate, likely margin and cost of acquiring that sale can be completely different.
Treating both searches as interchangeable because they eventually appear inside the same automated campaign is not sophisticated. It is commercially lazy.
A practical example: Smeg Shopping
Take a retailer selling Smeg appliances. There is a meaningful difference between searches such as “Smeg range cooker”, “Smeg 50s style fridge freezer” or a specific Smeg model—and broader searches such as “best fridge freezer” or “retro kitchen ideas”.
The first group contains shoppers who already know the brand and may know the exact product family they want. The later searches sit further up the funnel and are likely to be noisier, less predictable and more price-sensitive.
Both groups may be worth targeting. They should not automatically be bought in the same way or judged against the same target.
Campaign one: valuable brand and product intent
The first Standard Shopping campaign is designed to capture high-value Smeg searches. The product feed is carefully segmented. Search terms are reviewed daily and irrelevant or weak-intent queries are excluded through negative keywords.
Generic research traffic is pushed out so the campaign budget is concentrated on searches with a strong likelihood of producing profitable revenue. That control gives us the confidence to increase exposure deliberately.
We can monitor Search impression share, lost impression share from budget and rank, click costs, conversion rate, revenue and spend as a percentage of turnover. Where the economics remain sound, we keep pushing for a greater share of the searches that matter.
The objective is not cheap traffic. It is profitable dominance of commercially valuable demand.
Campaign two: acquire new demand at a different price
The same relevant Smeg inventory can then sit in a separate Standard Shopping campaign aimed at broader, non-brand searches. This campaign has a different role: introduce Smeg products to people searching by category, style or need rather than by brand.
Because this traffic is further from purchase, it should usually be acquired at a lower CPC and managed against a return requirement that reflects the greater risk. The tROAS targets are deliberately staggered to reflect the economics of each traffic group.
Campaign priority and negative keywords route searches intentionally and reduce cross-pollination. Simply copying a campaign does not create control. The routing logic, exclusions, budgets and bidding targets have to work together.
| Campaign | Commercial role | Primary control |
|---|---|---|
| Smeg brand and product | Convert existing high-intent demand and grow impression share profitably | Daily query review, generic negatives, appropriate tROAS and budget |
| Smeg non-brand | Acquire genuinely new category and style demand | Brand negatives, lower CPC ambition, separate budget and stricter return target |
The controlled alternative
Separate intent
Do not allow easy brand demand to conceal the true cost of prospecting.
Route queries
Use campaign priority and negatives so each search enters the right commercial lane.
Price the risk
Set budgets and tROAS targets around the economics of each traffic group.
Scale evidence
Increase impression share where revenue, margin and cost of sale support it.
Control is not the enemy of automation
There is a tendency to talk about hands-on search-term management as though it belongs to an earlier era of PPC. It does not.
Automation is only as valuable as the commercial boundaries placed around it. There are searches a business should not pay for, irrespective of how confident an algorithm appears to be.
Daily negative keyword work is not about blocking every imperfect query. It is about spotting patterns, protecting budgets and teaching the account which traffic belongs in which commercial lane.
ROAS alone can hide the truth
An account can report a good return while making poor decisions. If brand demand is blended with generic acquisition, the easier sales can make the total look efficient. If returning customers are mixed with new customers, reported revenue can be mistaken for incremental growth.
This is why the commercial scorecard has to extend beyond a single platform ROAS:
- Spend as a percentage of turnover
- Gross margin and contribution after advertising cost
- Search impression share on priority terms
- New versus returning customer performance
- Brand versus non-brand revenue
- Product-level profitability and stock position
- Genuinely incremental revenue
Google Ads should be managed as a commercial system, not a dashboard competition.
Where PMax does belong
Performance Max is not inherently bad. Used properly, it can uncover additional demand, combine signals efficiently and expand reach beyond conventional Shopping inventory.
But it should have a defined job. That might mean testing a selected product group, supporting new-customer acquisition, expanding into additional Google inventory or finding incremental volume after the core high-intent demand is already being captured efficiently.
It should be tested against a meaningful baseline, with brand controls, clean conversion data and clear expectations for incrementality. If it cannot demonstrate a better commercial outcome than the controlled alternative, its convenience is irrelevant.
The agency’s job is to make decisions
The role of a PPC agency is not to repeat Google’s latest recommendation deck. It is to understand the client’s margins, stock, customer journey, appetite for growth and acceptable cost of sale—and then build an account that reflects those realities.
Sometimes PMax will be the right answer. Sometimes Standard Shopping will offer better control. Often the strongest account will use both, each with a distinct role and a transparent measurement framework.
The important thing is that the choice is made for a commercial reason.
If your paid-search plan can be summarised as “put everything into PMax and let Google learn”, you do not have a strategy yet. You have a setting.