Performance Max gives Google more control than any campaign type in history. That’s either powerful or terrifying — depending on how well you understand what it’s doing with your budget.
If you’re running PMax and feeling vaguely unsettled about where your money’s going, you’re not paranoid. You’re paying attention.
This post covers what Performance Max actually does, what the data says about ecommerce results in 2026, and the specific steps that separate the advertisers getting real returns from those funding Google’s quarterly results.
What Is Performance Max (and Why Did Google Push It So Hard)?
Performance Max — or PMax — is Google’s fully automated campaign type. One campaign. One budget. One AI making decisions across every Google property simultaneously: Search, Shopping, YouTube, Display, Gmail, and Maps.
Google launched it in 2021 and by 2022 had effectively forced advertisers to migrate from Smart Shopping campaigns. If you’re running Google Ads for an ecommerce store today, you almost certainly have a PMax campaign running somewhere.
The pitch is appealing: hand Google your assets (headlines, images, video, product feed), set a target ROAS, and let machine learning find the best converting placements across all surfaces.
The reality is more complicated.
How Performance Max Actually Works
PMax uses auction-time bidding with signals pulled from every Google property simultaneously. It’s not just search intent — it’s YouTube watch history, Gmail behaviour, Maps activity, Display engagement. The AI builds a picture of likely converters and bids accordingly.
For a well-funded account with strong conversion history, this can be genuinely powerful. PMax has access to signals that no human-managed campaign can replicate.
For a UK ecommerce SMB spending £50–100/day? The picture is more nuanced.
The problem is that PMax optimises for conversions — and branded queries convert at 8–15x the rate of non-brand terms. Non-brand typically converts at 2–5x. When the AI is chasing easy wins, it finds them in branded searches.
Which means you may be paying Google to intercept traffic you’d have got for free.
The Brand Cannibalisation Problem (With Real Numbers)
This is the issue that doesn’t get talked about enough at the SMB level.
According to Google’s own data and independent audits from PPC platforms including Optmyzr and Smarter Ecommerce, PMax accounts running without brand exclusion typically see 15–35% of conversions attributed to branded queries. For SMBs with any brand search volume, that figure rises to 20–40% — because PMax aggressively targets the easiest conversions first, and branded searches are the easiest there is.
The tell: pull a Search Terms report after 30 days. If you see your brand name dominating the top queries, you’re not running incremental advertising. You’re paying Google to intercept your own organic brand traffic.
What happens to ROAS when you apply brand exclusion?
Your dashboard ROAS drops. Typically 30–50%.
That sounds catastrophic. It isn’t. The incremental revenue stays the same — you’ve simply stopped attributing natural brand traffic to the PMax campaign. The ROAS figure was inflated because branded queries were doing the heavy lifting.
Here’s the real-world version: a UK ecommerce retailer spending £2,000/month on PMax found that 38% of conversions were branded. After applying brand exclusion, reported ROAS dropped from 6.2x to 3.8x. Actual incremental revenue — the revenue that wouldn’t have happened without the ads — was unchanged. They’d been paying to intercept their own customers.
The honest metric for PMax is incremental ROAS: how much new revenue did the campaign drive that wouldn’t have happened organically? If you can’t answer that, your PMax data isn’t telling you what you think it is.
How to Set Up Brand Exclusion (Do This First)
Before anything else, exclude your brand terms at the account level.
- In Google Ads, go to Campaigns → Performance Max campaign
- Click Settings → Brand exclusions
- Add your brand name and all variants (misspellings, with/without Ltd, trading names)
- Apply at campaign level — this is now available natively in PMax (added late 2025)
If you’re running a separate branded Search campaign (which you should be — more on this below), the PMax brand exclusion ensures the two campaigns aren’t bidding against each other.
Recommended setup for UK ecommerce SMBs:
- PMax campaign: non-branded terms, product feed, with brand exclusion
- Separate branded Search campaign: exact match brand terms, manual or Target Impression Share bidding
This gives you full control over branded spend (and usually much lower CPCs on brand than PMax would bid) while letting PMax do what it’s actually good at: finding new customers you weren’t already reaching.
What’s Working for UK Ecommerce in 2026
The accounts getting the best results from PMax share a consistent set of behaviours. These aren’t theoretical — they’re what works.
Asset quality matters more than anything else
The assets you upload — headlines, descriptions, images, videos — are what Google uses to create ads across every channel. Weak assets mean weak ads across every channel, simultaneously. Invest time here.
For ecommerce, this means:
- Product-specific image assets (not generic lifestyle photography)
- At least one video asset, even a simple 15–30 second product demo. Accounts without video get limited YouTube inventory.
- Headlines that lead with the specific product category or USP, not the brand name
- Descriptions that name the benefit, not just the product
Feed quality is the foundation for Shopping placements
Most of your PMax conversions for an ecommerce store will come from the Shopping inventory — it’s the highest-intent placement. The quality of your Merchant Centre feed directly determines how well Google can match your products to search queries.
If your product titles are vague (“Blue Shirt — Size M”), your Shopping results will be too. Clean up titles to lead with the product type, key attributes, and brand in that order. That’s what searchers type, and what Google matches against.
Asset groups by product category, not one catch-all
One of the biggest PMax mistakes: a single asset group with all your products lumped together. Google then decides which assets match which products, which products match which searches. It doesn’t do this well.
Set up asset groups aligned to your product categories. Use audience signals for each asset group to tell Google the type of customer you’re targeting. This doesn’t restrict who sees your ads — Google will still go broader — but it gives the algorithm a starting point that actually reflects your business.
Give it 4–6 weeks before judging
PMax’s learning period is real. Cutting a campaign after 2 weeks because the ROAS looks poor will give you a false negative. Set your tROAS based on your historical Shopping performance — not an aspirational target — and let the campaign accumulate data before making adjustments.
What’s Not Working (And What to Watch)
PMax in low-conversion-volume accounts
If your Google Ads account generates fewer than 30–50 conversions per month across all campaigns, Performance Max doesn’t have enough data to optimise meaningfully. It will default to broader, less targeted placements that look active but don’t convert.
In these cases, Standard Shopping with manual bidding or a basic tROAS — and a solid negative keyword list — will outperform PMax until you have the volume to support automation.
Insufficient budget for the learning period
PMax needs budget to learn. If you’re running it on £10/day, you won’t accumulate enough data to exit the learning phase in any useful timeframe. A rough guide: your daily PMax budget should be at least 10x your target CPA. If your target CPA is £50, you need at least £500/day to give the algorithm room to work.
For most UK SMBs, that’s not realistic — which is exactly why Standard Shopping often makes more sense at the early stage.
Ignoring the insights tab
The PMax insights tab is the one place Google gives you back some visibility. It shows you top-performing search categories, audience insights, and asset performance. Check it weekly. If your video assets are rated “Low,” fix them. If search categories are irrelevant to your products, tighten your asset groups and product feeds.
Treating PMax as a set-and-forget campaign
It’s not. Google’s automation handles the bidding and placement decisions, but you still need to:
- Review and refresh asset groups every 4–6 weeks
- Monitor your feed quality in Merchant Centre
- Check for and apply brand exclusions
- Review the insights tab for anomalies
- Ensure your conversion tracking is accurate and your tROAS reflects reality
What Budget Does PMax Actually Need to Work?
Below a certain threshold, PMax doesn’t have enough data to optimise properly. The algorithm needs conversion volume to learn. Without it, you’re paying for an AI that’s essentially guessing.
For UK ecommerce SMBs, the thresholds are:
- Minimum viable: £30–50/day (£900–1,500/month) — below this, PMax rarely accumulates enough conversion data to exit the learning phase meaningfully
- Sweet spot for early testing: £50–75/day (£1,500–2,250/month) — enough volume for the algorithm to start making genuine decisions
- Realistic starting point if budget is limited: Consider delaying PMax and running Standard Shopping + Search until you have 30+ conversions/month. Smart Bidding (including PMax) performs significantly better with conversion history.
If your account is under 30 conversions per month, tROAS targets are also likely to cause problems — the algorithm doesn’t have enough signal to trust. In that scenario, Maximise Conversion Value without a target ROAS is often more effective to start.
The Single-Surface Test: PMax Without the Black Box
One of the most underused PMax tactics is running a campaign that behaves more like Standard Shopping — without the opacity of full multi-surface PMax.
How it works: Create a PMax campaign with an asset group that contains only your product feed. No headlines. No images. No video. No description text beyond feed attributes.
When PMax has no creative assets for Display, YouTube, or Gmail, Google can only serve Shopping and Search ads. You get PMax’s cross-channel bidding intelligence while limiting placements to the surfaces you actually understand.
This is particularly useful as a test phase:
- Run feed-only PMax for 4 weeks with brand exclusion applied
- Review Search Terms report for placement quality
- Only add image/video assets once you’re confident in the baseline performance
You lose some of PMax’s theoretical reach — but you gain clarity. And for most ecommerce SMBs, Shopping + Search is where the money is anyway.
What the Search Terms Report Is (Now) Telling You
One of the biggest improvements to PMax in 2025 was improved Search Terms reporting at the asset group level. You can now see more of the actual queries your PMax campaign is triggering — not everything (Google still aggregates low-volume terms), but enough to identify problems.
Check monthly for:
- Brand terms appearing — apply brand exclusion immediately if you see them
- Irrelevant categories — use negative keywords at campaign level (now available for PMax)
- Query intent mismatches — informational queries converting poorly are a budget drain
- Geographic skew — if you’re UK-only, confirm your location settings aren’t capturing international traffic (this is more common than it should be)
PMax does now support campaign-level negative keywords. Use them. They’re the primary control mechanism you have over what PMax actually serves.
Performance Max vs Standard Shopping: Which Should You Run?
This isn’t an either/or for most accounts. Many experienced ecommerce advertisers run both — using Standard Shopping for their highest-margin, best-converting product categories (where they want granular control and bidding precision), and PMax for broader product ranges and prospecting.
Use PMax when:
- You have 30+ conversions/month (enough data for Smart Bidding to work properly)
- You want to expand reach beyond Shopping and Search to YouTube/Display
- Your product feed is clean and your assets (images, headlines) are high quality
- You have brand exclusion set up and a separate branded campaign running
Stick with Standard Shopping when:
- You’re under 30 conversions/month — PMax won’t outperform with thin data
- You want granular control over match types and placements
- You’re testing a new product range and want visibility into which products are serving
- Your budget is below £30/day — PMax needs volume to learn
Neither is universally better. PMax works when it has what it needs. When it doesn’t, it’s an expensive black box.
If you’re starting fresh, Standard Shopping first — until you have the conversion history to feed PMax properly.
The Realistic 2026 Verdict
Performance Max is a mature product now. The early-adopter mess — where advertisers handed Google a blank cheque and got confused results — is settling into something more manageable.
The advertisers winning with PMax in 2026 share a few things in common:
- Brand exclusion is non-negotiable. If you haven’t set it up, do it today.
- They watch the Search Terms report. Monthly minimum. Weekly if you’re spending £100+/day.
- They started with a feed-only asset group. Full creative came after baseline performance was established.
- They have separate branded campaigns. PMax handles prospecting. Search handles brand.
- They measure incremental ROAS, not dashboard ROAS. The vanity metric is the enemy.
PMax isn’t magic. But it’s also not broken — it’s just misused by accounts that handed it the keys without understanding what it does with them.
Frequently Asked Questions
What is Performance Max and how is it different from Shopping campaigns?
Performance Max (PMax) is Google’s fully automated campaign type that serves ads across all Google properties simultaneously — Search, Shopping, YouTube, Display, Gmail, and Maps — from a single campaign and budget. Standard Shopping campaigns only serve on Google Shopping and Search. PMax uses AI to allocate budget across surfaces automatically, which gives it broader reach but less advertiser control. For ecommerce, PMax replaces the older Smart Shopping campaign format.
Does Performance Max cannibalise brand search traffic?
Yes — and this is one of the most significant issues for ecommerce advertisers. Without brand exclusion applied, PMax typically attributes 15–40% of conversions to branded queries. Since branded searches convert at much higher rates than non-brand, the algorithm gravitates toward them. The fix is applying brand exclusions at the campaign level and running a separate branded Search campaign.
How much budget does Performance Max need to work properly?
PMax needs enough conversion volume for its Smart Bidding to function effectively. For UK ecommerce SMBs, a minimum of £30–50/day (£900–1,500/month) is typically required to exit the learning phase. Below this, consider Standard Shopping campaigns until your conversion volume grows to 30+ per month.
What is Performance Max brand exclusion and how do I set it up?
Brand exclusion prevents PMax from bidding on searches containing your brand name. Set it up in Google Ads under Campaign → Settings → Brand exclusions. Add your brand name and all variants. This is especially important if you’re running a separate branded Search campaign — without brand exclusion, PMax and your branded campaign compete with each other, inflating costs.
How do I know if my Performance Max campaign is actually working?
The most reliable measure is incremental ROAS — the revenue that wouldn’t have happened without the campaign, excluding branded conversions that would have come through organically. Pull a Search Terms report monthly to review what queries are triggering your ads. If branded terms dominate and your ROAS drops significantly after brand exclusion, the campaign wasn’t adding as much value as the dashboard suggested.
Should I use Performance Max for ecommerce, or is it not right for my business?
PMax is not the right fit for every ecommerce business. Avoid it — or delay it — if: your account generates fewer than 30 conversions per month (Smart Bidding won’t have enough data to work properly); your daily budget is below £30 (insufficient volume for the algorithm to learn); you need granular placement control for compliance or brand safety reasons; or you’re launching a new product range where you want to monitor exactly which products and queries are driving results. In those situations, Standard Shopping campaigns give you more control and visibility while you build the conversion history PMax needs to deliver results.
Why is my Performance Max campaign not performing as expected?
The most common reasons PMax underperforms for ecommerce stores are: insufficient conversion volume (fewer than 30–50/month makes automation difficult), poor product feed quality in Merchant Centre, weak creative assets (especially missing video), tROAS targets set too aggressively too soon, or brand search terms cannibalising credit from organic traffic. A Google Ads audit can identify which of these is the main issue in your account.
Ready to Review Your Google Ads Performance?
If you’re unsure whether your PMax campaign is working as hard as it should, or whether it’s just taking credit for your brand’s organic traffic, we can help.
PalMultimedia.com manages Google Ads for UK ecommerce businesses — from account audits to full campaign management. Every recommendation comes from hands-on experience with Magento, Shopify, and Linnworks-integrated accounts.