Updated October 2026 · By Farrukh Moin Ahmed, digital marketing expert
Short answer: Performance Max for ecommerce is a strong tool but a weak strategy. When one campaign runs the whole catalogue, you lose visibility into what sells, control over the funnel and say over where budget goes. Give it narrow jobs instead: product-specific campaigns, Shopping-only asset groups, new-customer settings and a deliberate funnel around it. That approach took Petoi from a 2.0 to a 4.1 ROAS.
Key takeaways
- Performance Max is good at finding conversions, poor at explaining them.
- One campaign for the whole catalogue favours returning visitors, brand searches and best sellers.
- Split it by product line or margin, and judge capture and creation campaigns separately.
- Petoi: ROAS 2.0 to 4.1, CPC $2.30 to $1.00, purchases 31 to 165 over October to December.
What is Performance Max, and why do ecommerce accounts lean on it?
Performance Max is a Google Ads campaign type that runs across Google’s inventory from a single campaign: Search, Shopping, YouTube, Display, Discover, Gmail and Maps. You give it assets, a product feed and a conversion goal, and Google’s automation decides where to show ads, to whom and at what bid.
For an online store, that’s attractive. One campaign, one budget, products shown everywhere Google can show them. It’s also why so many ecommerce accounts I audit end up with most of their spend in Performance Max. It was easy to set up, it reported good numbers early on, and over time it quietly became the whole strategy.
Performance Max is very good at one thing: finding conversions inside Google’s inventory. It’s much less good at telling you why it found them, or whether those conversions would have happened anyway.
What do you lose when Performance Max runs everything?
When one Performance Max campaign carries most of an account, three things quietly disappear.
1. Visibility
You can’t easily see which products, audiences, searches or placements are doing the work. Reporting has improved, but a single campaign covering the whole catalogue still blends best sellers with products that never earn their spend.
2. Funnel control
Prospecting and remarketing blend together. A campaign that reports a strong return might be mostly reaching people who already visited your site or searched your brand name. That looks efficient on a dashboard but does little to grow the business.
3. Spend allocation
Budget flows toward whatever converts most easily. That’s often returning visitors, branded searches and your top two or three products. New products, higher-margin lines and genuinely new customers get whatever is left.
None of this means Performance Max is bad. It means it shouldn’t be the strategy. It should be one tool inside one.
What are the two jobs every ecommerce account has to do?
Every online store’s ad account is doing two different jobs at once:
| Capturing demand | Creating demand | |
|---|---|---|
| Who you reach | People already searching for what you sell | People who don’t know they need it yet |
| Where it happens | Search and Shopping results | YouTube, Display, Discover, Demand Gen |
| How fast it converts | Quickly | Slowly, often through a later search |
| How to judge it | Cost per purchase and ROAS | New customers, branded search growth, assisted sales |
These need different campaigns, different budgets and different success metrics. When both are hidden inside one Performance Max campaign, the automation will always favour capture, because it converts faster and makes the campaign’s numbers look better. Demand creation gets starved, and growth eventually stalls.
How do you give Performance Max narrower jobs?
The fix isn’t to switch Performance Max off. It’s to give it narrower, clearer briefs, so you keep the automation and get your visibility back.
- Split by product line or margin. Instead of one campaign for the whole catalogue, build product-specific Performance Max campaigns for your priority lines. Custom labels in your feed, such as margin band or best seller, make this easy. I cover them in the Shopping feed guide.
- Use Shopping-only asset groups where you need clarity. An asset group that relies on the product feed, without a full set of creative assets, behaves much more like Shopping, which makes product-level performance easier to read.
- Set targets by what each line can afford. A high-margin product line can run at a lower return target than a low-margin one. One campaign with one target forces them to compete on the same terms.
- Use customer acquisition settings deliberately. Google’s customer lifecycle goals let Performance Max, Search, Shopping and Demand Gen bid higher for new customers, or only for new customers, using your customer list to identify existing buyers. That directly addresses the problem of automation chasing people who would have bought anyway.
How do you build a funnel around Performance Max?
Narrow Performance Max campaigns work best inside a deliberate funnel, where each stage has a campaign whose job is clear:
- Top of funnel: top-of-funnel Search for problem and category searches, prospecting Display, YouTube and Demand Gen to reach new audiences.
- Mid funnel: in-market audiences and product-specific Performance Max to move interested shoppers toward a decision, with Shopping-only asset groups where you need product-level clarity.
- Bottom of funnel: remarketing to bring back people who visited but didn’t buy.
This structure is more work to set up than one big campaign. In return, every stage can be judged on its real job, and budget can be moved between stages on purpose instead of by the algorithm’s preference for easy conversions.
What happened when Petoi made this change?
Petoi is a US robotics brand selling programmable robot kits. Robotics kits are a considered purchase: many buyers need to understand the product before they’re ready to buy. Its account relied heavily on Performance Max, which left limited visibility into what drove sales, little control over the funnel, and no clear say over where spend went.
I rebuilt it around the buyer journey: top-of-funnel Search, prospecting Display, YouTube and Demand Gen at the top; in-market audiences, product-specific Performance Max and Shopping-only asset groups in the middle; and remarketing to close. Over the October to December reporting window:
| Metric | Before | After |
|---|---|---|
| ROAS | 2.0 | 4.1 |
| Average CPC | $2.30 | $1.00 |
| Purchases | 31 | 165 |
| Conversion value | $11,000 | $50,000 |
The lower cost per click matters as much as the higher return. When campaigns have clear jobs and reach the right people at the right stage, Google doesn’t need to bid as aggressively to find buyers. More on selling products that need explaining is on the consumer tech and robotics page.
When is a single Performance Max campaign fine?
There are accounts where one Performance Max campaign is a reasonable choice:
- A small catalogue where every product has similar margins.
- A budget too small to feed several campaigns with enough conversion data.
- A strong, clean feed and reliable conversion tracking with real purchase values.
- A business that’s happy to judge results on blended store revenue rather than detailed channel reporting.
Even then, I’d keep brand search reported separately and check new versus returning customers, so a single campaign can’t take credit for sales your brand would have made anyway.
What should you check in Performance Max reporting?
Whatever the structure, these are the numbers I look at every week in an ecommerce account running Performance Max:
- Conversion value by product. Which products earn their spend and which are carried by the rest of the catalogue.
- Search term insights. What people actually searched before buying, which often shows how much spend is going to branded searches.
- New versus returning customers. Whether the campaign is growing the customer base or recycling existing buyers.
- Return against breakeven. Your breakeven ROAS is 1 divided by your gross margin. A campaign reporting a 3.0 ROAS on a 30% margin product is barely profitable before any other costs.
- Store revenue. Google-reported conversions compared with actual orders in your store, so attribution doesn’t flatter the account.
How do you restructure Performance Max without losing sales?
Tearing down a campaign that’s producing most of your revenue is risky. Bidding systems lose their learning, and a sudden structural change can cause a dip before results recover. I restructure in stages instead:
- Fix measurement first. Confirm purchases and purchase values are tracked correctly and match store orders. Restructuring on bad data just moves the problem around.
- Label the feed. Add custom labels for margin band, best sellers and new arrivals, so products can be grouped by business value.
- Carve out one product line. Move a single priority line into its own Performance Max campaign or Shopping-only asset group, and exclude it from the original campaign. Watch it for a few weeks before moving the next.
- Add the missing funnel stages. Top-of-funnel Search, YouTube or Demand Gen for demand creation, and remarketing for people who didn’t buy, each with its own budget.
- Shift budget on evidence. Move spend toward the campaigns and product lines whose results hold up against store revenue, not just platform ROAS.
Done this way, the account keeps most of its learning while you get visibility back one piece at a time. The goal isn’t fewer automated campaigns. It’s automated campaigns with clear, measurable jobs.
The same principle applies on Meta. When Woda Bags’ Meta account was split into too many campaigns, each one starved of data. There, the fix ran the other way: consolidation. On both platforms, the structure should match how much data each campaign can actually learn from.
A quick check for your own account
If you can’t answer at least three of these, Performance Max is probably running your account rather than the other way round:
- What share of your Google Ads spend sits in Performance Max?
- Which five products drive most of your conversion value?
- Do you have a budget specifically for reaching new customers?
- Is remarketing measured separately from prospecting?
- How much of your Performance Max spend goes to searches for your own brand?
- Do Google-reported conversions match the orders in your store?
Automation should execute a strategy. It shouldn’t replace one.
For how Performance Max fits into a complete Google setup for online stores, see Google Ads for ecommerce. If Meta Ads runs alongside it, Google Ads vs Meta Ads for ecommerce explains how the two divide the work.
If you want help splitting Performance Max into narrower jobs, that is what I do as a Google Ads expert for ecommerce.
Frequently asked questions
Is Performance Max good for ecommerce?
Yes, as part of an account. It’s strong at finding conversions across Google’s inventory. It works best with narrow briefs, such as one product line or margin band per campaign, rather than running the whole catalogue.
Where do Performance Max ads show?
Across Google’s inventory from one campaign, including Search, Shopping, YouTube, Display, Discover, Gmail and Maps.
Can Performance Max focus on new customers?
Yes. Google’s customer lifecycle goals let Performance Max bid higher for new customers, or only for new customers, using a customer list to identify existing buyers.
Should I run Standard Shopping alongside Performance Max?
Often, yes. Standard Shopping gives clear product and search term reporting. Many accounts use both, with each campaign given a distinct set of products or a distinct job.
Why is my Performance Max ROAS high but sales are flat?
Often because the campaign is reaching people who would have bought anyway: returning visitors and branded searches. Check new versus returning customers and branded search share, and compare with store revenue.
