Our Approach to Performance Max for eCommerce

I’ve lost count of the number of eCom accounts I’ve audited where Performance Max is treated as a set-and-forget, multi-channel solution. The campaign’s running, the ROAS looks decent in the dashboard, and everyone assumes it’s doing the heavy lifting across Shopping, Search, Display, YouTube.

But when you actually dig into the data, the story is almost always different.

Performance Max is a powerful campaign type. But for ecommerce businesses, particularly those with a large number of SKUs, it needs to be structured with intent. Without that, you end up with bloated spend, cannibalised asset groups, and performance that looks good on the surface but falls apart under scrutiny.

This guide breaks down how I approach PMax for eCom brands at Amore Digital, the common pitfalls I see, and the levers that actually move the needle.

The PMax Misconception: “It Covers Everything”

One of the most common things I hear from ecommerce founders is: “We’re running PMax so we’re covered across all channels.”

In theory, that’s what Google wants you to believe. PMax serves ads across Shopping, Search, Display, YouTube, Discovery, and Gmail from a single campaign. Sounds efficient. But in practice, for most eCom businesses with a decent product catalogue, the overwhelming majority of Performance Max spend and revenue comes from one place: Google Shopping.

I’ve seen accounts where 80–90% of PMax conversions are driven by Shopping placements. Shopping is where transactional intent lives for product searches. But it does mean that treating PMax as a genuine multi-channel play is misleading. You’re essentially running a Shopping campaign with some extras bolted on.

The issue becomes when businesses don’t recognise this and fail to put the work into their product feed, Merchant Centre structure, and listing group segmentation. If Shopping is where the revenue comes from, then your feed is your foundation, and it needs serious attention.

The Brand Spend Problem: Your ROAS Isn’t What You Think It Is

This is probably the most overlooked issue I come across in PMax accounts.

You open the campaign, see a 6x, 8x, maybe even 10x ROAS, and think everything’s working brilliantly. But when you pull the search terms report (which Google has thankfully started surfacing more of), you’ll often find a huge chunk of that revenue is being driven by brand terms.

Someone searches your brand name, they already know you, they’re already coming to buy, and PMax is claiming that conversion. Your ROAS looks inflated because you’re paying for traffic that would’ve come to you organically or through a much cheaper branded search campaign.

This is why negative keywords are essential in PMax. Google now allows you to add up to 10,000 account-level negative keywords that apply to PMax campaigns. If you’re not using this, you need to start immediately.

Here’s what I’d recommend as a minimum:

• Negative your brand terms from PMax and run a dedicated branded search campaign instead. This gives you control over your brand traffic at a fraction of the cost, and forces PMax to work for its keep on non-brand, prospecting queries.
• Pull search term data regularly (via the Insights tab or Google Ads scripts) and identify any irrelevant queries that are eating budget.
• Add competitor brand terms as negatives unless you’re intentionally running a conquest strategy, and even then, that’s usually better served in a dedicated search campaign where you can control the messaging.

Once you strip out brand from PMax, you’ll get a much more honest picture of how it’s actually performing on non-brand, prospecting traffic. And that’s the metric that matters, because that’s where real growth comes from.

Asset Group Structure: Why Category-Level Segmentation Matters

I touched on this in a previous post, but it’s worth going deeper here because it’s one of the most impactful changes you can make to a PMax account.

Too many eCom accounts have asset groups that look well-organised on the surface, labelled by category, with tailored creative and copy, but when you check the listing groups, every single asset group is targeting the entire product range. Google doesn’t care what you’ve named your asset group. If the listing group says “all products,” that’s what it’ll serve.

This creates a few problems:

• Spend gets cannibalised across asset groups because they’re all competing for the same products.
• You lose any meaningful data at the category level because everything’s blended together.
• Creative differentiation becomes pointless. Google will prioritise Shopping inventory over your carefully crafted display assets in most cases.

The fix: split your asset groups by core product categories and ensure each one only targets the relevant products through listing groups. This requires work in Merchant Centre, your product types, custom labels, and category structure all need to be dialled in so you can segment properly.

Here’s the process I follow:

1. Audit your Merchant Centre feed. Make sure product types are accurate, product titles are descriptive custom labels are in place (I typically use these for margin tiers, seasonality, and bestseller status), and your category taxonomy is clean.
2. Create asset groups aligned to your core categories. For example, if you sell clothing, you might have separate asset groups for dresses, outerwear, accessories, etc. Each one should only target the products in that category via listing groups.
3. Tailor creative and copy per asset group. Now that each group targets a defined product set, your creative actually has context and relevance.
4. Monitor and iterate. Give each asset group enough data to learn (Google recommends 20–30 conversions per month per campaign for the algorithm to optimise properly). Watch which categories are performing and which are dragging.

Scaling Winners: When to Break Out a Campaign

This is where it gets interesting. Once you’ve got well-segmented asset groups and one starts showing strong, consistent performance, you’ve got a decision to make.

If your PMax campaign is capped by budget (meaning it’s consistently hitting its daily limit and could spend more) then simply increasing the budget applies to the entire campaign. Your top-performing asset group might get more spend, but so will underperforming ones.

The better approach is to break that winning asset group out into its own dedicated PMax campaign. This gives it its own budget, its own ROAS target, and its own learning data without being diluted by weaker categories. You can then scale it independently while keeping the original campaign focused on developing the remaining categories.

I’ve seen this approach unlock significant growth for eCom brands, particularly those with a few hero categories that drive the bulk of revenue. Giving those categories their own campaign means Google’s algorithm has a clear, focused objective and doesn’t have to balance competing priorities within a single campaign.

Understanding ROAS Targets: The Lever That Controls Your Spend

This is the area where I see the most confusion, and arguably where the biggest gains (or losses) are made.

In PMax, your target ROAS essentially tells Google how aggressively to bid. But the relationship isn’t always intuitive, below is a breakdown:

• A lower ROAS target gives Google more room to spend. You’re telling the algorithm you’re willing to accept a lower return per pound spent, so it bids higher, enters more auctions, and reaches a wider audience. Your CPCs go up, but so does your volume.
• A higher ROAS target restricts Google’s ability to spend. You’re saying you want a higher return, so the algorithm becomes more selective. It lowers bids, enters fewer auctions, and only shows your ads where it’s confident of a strong return. Your CPCs drop, but so does your reach and volume.

Think of it like a dial. Turn it down (lower ROAS target), and Google opens the taps. Turn it up (higher ROAS target), and it tightens.

The key insight here is that Google controls your CPC through the ROAS target. You don’t set CPCs directly in Pmax, the algorithm adjusts bids in real-time based on the target you’ve set and the conversion data it’s working with. So when you change your ROAS target, you’re not just changing a reporting metric, you’re fundamentally changing how aggressively Google bids in every single auction.

Finding the Balance

The mistake I see most often is setting a ROAS target that’s too high because it “looks good” in reporting. A 10x ROAS sounds brilliant on paper, but if your campaign is only spending £50 a day when you’ve got £500 of daily budget available, that target is choking your growth.

On the flip side, dropping your ROAS target too aggressively can burn through budget quickly on low-quality traffic. It’s a balancing act.

Here’s how I typically approach it:

• Start with your break-even ROAS. Work out what return you need to at least cover your costs (factoring in margins, shipping, overheads). That’s your floor, you never want to go below this.
• Set your initial PMax target slightly above break-even. This gives Google room to learn and spend while keeping you profitable.
• Adjust incrementally. If the campaign is spending well and performance is stable, you can gradually increase the target. If spend drops off and you’ve got room to be more aggressive, lower it slightly.
• Review weekly, not daily. PMax needs data to optimise. Making daily changes to ROAS targets disrupts the learning cycle and usually makes things worse.

The goal is to find the sweet spot where you’re spending as much as possible while maintaining a return that keeps you profitable. That number will be different for every business, and it’ll shift with seasonality, competition, and stock levels.

Bringing It All Together

If I could distil everything above into a practical checklist, it would look like this:

Recognise that PMax is primarily a Shopping campaign for eCom. Invest heavily in your product feed and Merchant Centre setup accordingly.

Implement negative keywords to remove brand traffic from PMax. Run brand through a dedicated search campaign and force PMax to earn its results on non-brand queries.

Segment asset groups by core product categories with proper listing group targeting, not just labels and creative.

Break out winning asset groups into dedicated campaigns when your main campaign is budget-capped, giving top performers their own runway to scale.

Use ROAS targets as your primary spend lever. Understand the inverse relationship between target and volume, and adjust incrementally based on your margins and growth objectives.

None of this is particularly complex in isolation. But the reality is that most eCom PMax accounts I audit are missing at least three of these five points. The brands that get these fundamentals right are the ones that scale efficiently, the rest end up overspending on brand traffic and wondering why their growth has plateaued.

If your PMax setup hasn’t been reviewed in a while, it’s probably worth a proper audit. The structure you set today determines the results you get for months to come.

If you would like a no-obligation audit of your Google ads campaign, feel free to request a growth plan, and I’ll be in touch to arrange an initial call. Link to growth plan > https://www.amoredigital.co.uk/growth-plan/