Performance Max is Google’s most powerful campaign type and the one we see misconfigured the most. For lead generation accounts specifically, the margin for error is razor thin because you are feeding an algorithm that was originally designed for ecommerce and telling it to find qualified leads.
When PMax works in lead gen, it works well. But the PMax mistakes we see in audits are not minor tuning issues. They are structural problems that cause the algorithm to optimize for junk leads, cannibalize brand traffic, and burn through budget on placements that never convert.
We have audited over 100 lead gen PMax accounts in the last year. These are the five mistakes that show up in nearly every underperforming account.
Mistake 1: Weak conversion actions poison the algorithm
This is the single most damaging PMax mistake in lead gen. The algorithm optimizes toward whatever conversion action you give it. If that action is a form submission with no quality signal, PMax will find the cheapest form fills on the internet. Most of those leads will never answer the phone.
What goes wrong
- Optimizing for form fills only: PMax finds low-intent users who fill out forms but never engage with sales
- Counting page views or "engaged sessions" as conversions: The algorithm gets rewarded for worthless signals
- Using multiple conversion actions with equal weight: PMax cherry-picks the easiest one, which is rarely the most valuable
What to do instead
- Import offline conversions from your CRM. Pass back which leads became qualified opportunities or closed deals. This gives PMax the signal it needs to find similar high-quality prospects.
- Use Enhanced Conversions for Leads (EC4L). This Google feature matches form submissions to downstream outcomes using hashed first-party data.
- Set primary vs. secondary conversion actions. Only your most valuable conversion action should be "primary." Everything else should be "secondary" (observable but not optimized toward).
- Implement value-based bidding. Assign different values to different conversion types based on their close rate. A phone call might be worth 3x a form fill.
Without strong conversion signals, everything else in your Performance Max strategy is built on a bad foundation.
Mistake 2: Asset group structure that defeats the purpose
Asset groups in PMax are supposed to organize your campaign by theme so each group serves the right creative to the right audience. Most lead gen accounts either throw everything into one asset group or create a fragmented mess.
Common asset group problems
| Problem | What Happens | Fix |
|---|---|---|
| One giant asset group | Algorithm mixes signals across all services | Separate by service line or offer |
| Too many asset groups (10+) | Budget fragments, learning phase never ends | Consolidate to 3-5 focused groups |
| Same audiences across groups | Groups compete against each other | Assign distinct audience signals per group |
| Weak creative variety | Algorithm has limited options to test | Provide 5+ headlines, 5+ descriptions, 3+ images per group |
| No final URL expansion control | Traffic lands on irrelevant pages | Set specific final URLs per asset group |
The right asset group architecture
For a lead gen account, we typically build 3-5 asset groups organized by:
- Service line: Each major service gets its own group with tailored creative and landing pages
- Audience intent: Separate groups for high-intent audiences (customer match, website visitors) vs. prospecting audiences
- Geographic focus: If performance varies significantly by region, split into geo-based groups
Each asset group should feel like a standalone mini-campaign with its own messaging, landing page, and audience signal strategy.
Mistake 3: No brand exclusions or guardrails
PMax will happily spend your entire budget on branded search if you let it. For lead gen accounts, this is particularly wasteful because brand searches are already converting through your Search campaigns at a lower CPA.
How brand cannibalization happens
PMax has auction priority over Standard Search campaigns. When someone searches your brand name, PMax bids on that query and claims the conversion. Your branded Search campaign loses impression share. Your PMax reports look great, but total lead volume does not increase.
This is the most common way PMax creates the illusion of performance. The cost per lead looks low because it is capturing your cheapest, highest-intent traffic.
How to fix it
- Add brand exclusions at the campaign level. Google now allows brand exclusion lists for PMax. Add your brand name, common misspellings, and variations.
- Monitor brand vs. non-brand split in PMax insights. If more than 20% of PMax conversions come from brand queries, the exclusions are not working well enough.
- Compare total branded search volume before and after PMax launch. If total branded conversions stayed the same but shifted from Search to PMax, that is cannibalization, not growth.
We consider brand exclusions a non-negotiable part of any lead gen Performance Max strategy. Without them, you cannot trust your PMax performance data.
Mistake 4: Ignoring placement quality
PMax runs across Search, Display, YouTube, Discover, Gmail, and Maps. Not all of those placements are equal for lead generation. Display and Discover placements in particular tend to generate high volume but low quality leads.
The placement quality problem
Google does not give you full control over where PMax spends. But you can influence it:
- Review the "Where ads showed" report. Identify placements that generate clicks but no conversions or only low-quality conversions.
- Exclude specific placements and URLs. Mobile game apps, parked domains, and irrelevant content sites are common offenders.
- Build exclusion lists proactively. Maintain a running list of 200+ placement exclusions for Display and YouTube.
- Strengthen audience signals. The better your audience signals (customer match lists, website visitors, custom segments), the less PMax relies on broad Display placements to find volume.
Placement quality by channel
For lead gen accounts, here is what we typically see:
- Search: Highest quality. Leads from Search placements close at the highest rates.
- YouTube: Mid-quality. Works well for remarketing, weaker for cold prospecting.
- Gmail: Variable. Can work for B2B but generates a lot of low-intent clicks for services.
- Display: Lowest quality. High impression volume, low conversion rate, poor lead quality.
- Discover: Variable. Mobile-heavy, impulse clicks, lower form completion rates.
You cannot turn off specific channels in PMax. But you can influence the mix by managing audience signals and exclusions aggressively.
Mistake 5: Reporting that hides the truth
The final mistake is treating PMax reporting at face value. Google’s default reporting for PMax is designed to make the campaign look successful. If you are not digging deeper, you are making decisions on incomplete data.
What default reporting misses
- Conversion quality: PMax reports conversion volume. It does not tell you how many of those leads were qualified, contacted, or closed.
- Incrementality: Was PMax the reason someone converted, or would they have converted through Search anyway? Default reporting assumes full credit.
- Cross-channel attribution: PMax claims conversions from multiple touchpoints. Check your CRM data for the real first-touch source.
How to build reporting you can trust
- Connect CRM data to Google Ads. Import pipeline stages so you can see PMax leads through to qualification and close.
- Run incrementality tests quarterly. Turn PMax off in a controlled geo for 2-4 weeks and measure Performance Max incrementality against the holdout.
- Build a custom dashboard that shows PMax performance alongside total account performance. If PMax CPA goes down but total account CPA stays flat, it is redistributing, not adding value.
- Track PMax leads separately in your CRM. Tag the source and measure close rate, average deal value, and time to close against other channels.
Frequently asked questions
What is the most common PMax mistake in lead gen?
Optimizing toward weak conversion actions like raw form submissions without any quality signal. The algorithm will find the cheapest form fills available, which are almost never your best leads. Importing offline conversion data from your CRM is the single most impactful fix.
Should I exclude brand terms from Performance Max?
Absolutely. PMax has auction priority over Search campaigns and will cannibalize your branded traffic. Add brand exclusion lists and monitor the brand vs. non-brand split in your PMax insights tab. If your brand conversions just shifted from Search to PMax, you gained nothing.
How do I know if PMax is actually adding incremental leads?
Run a geographic holdout test. Disable PMax in one region while keeping it active in another. Compare total lead volume (all channels) between the test and control groups. If total volume does not drop in the holdout, PMax was not adding incremental value.
How many asset groups should I have?
Three to five for most lead gen accounts. One per major service line or offer. Fewer than three limits the algorithm’s ability to segment audiences. More than seven usually fragments budget and extends the learning phase.
Stop guessing whether PMax is working
If your Performance Max campaign looks good on paper but your sales team says lead quality is declining, the structure is the problem. Book a Performance Max Strategy Session and we will audit your conversion actions, asset groups, brand exposure, and reporting to find exactly where the leaks are.
References
- Google Ads Help. Performance Max Campaign Best Practices. Google.
- Search Engine Land. Performance Max for Lead Generation: Common Mistakes and Fixes. Search Engine Land.
- Optmyzr. PMax Incrementality Testing Framework. Optmyzr Blog.
