Ask any performance marketer about Performance Max for lead generation and you'll get one of two reactions: enthusiastic adoption or flat refusal. The refusal camp has a case — there are enough real examples of PMax campaigns generating high volumes of junk leads that the scepticism is understandable. But the diagnosis is usually wrong.
PMax doesn't fail at lead generation. Smart Bidding without closed-loop data fails at lead generation. And because PMax is the most algorithmically dependent campaign type Google has ever shipped, the consequences of running it without proper conversion data are more severe than in any other format. The platform amplifies both good and bad signal at scale.
Here's what's actually happening when PMax "doesn't work" for lead gen — and what you need to do before it can.
The real problem: what you tell Google to optimise for
Every Google Ads campaign — PMax, Search, Demand Gen — is ultimately optimising towards whatever conversion action you've defined. If you've only set up form submissions or phone calls as your conversion goals, you've told Google: "find me people who submit forms or call." And Google will. Efficiently. At scale.
The problem for home services and trades businesses in Australia is that not every form submission is a good lead. Not every call becomes a job. Someone calling from the wrong suburb, requesting a service you don't offer, looking for a price you'd never quote — these all count as conversions in Google Ads if that's how you've set things up. The algorithm thinks it's winning. Your sales team knows it isn't.
This problem exists in Search campaigns too — but PMax makes it worse because the algorithm has so much more latitude in how it finds and targets audiences. When a Search campaign targets a tightly defined keyword theme, at least there's intent-based filtering happening. PMax finds audiences across every Google surface — Search, Display, YouTube, Gmail, Discover, Maps — using ML to identify patterns in your conversion data. Give it bad data and it's finding people who look like your bad leads.
The tyre-kicker amplification loop
Here's the cycle that plays out in most PMax lead gen accounts that aren't running closed-loop conversion data:
-
1
You launch PMax with form submit as your only conversion
Google needs 30–50 conversions per month to exit the learning period. Form submits are easy to generate — many are low quality, but the volume is there.
-
2
Smart Bidding learns what a form submitter looks like
The algorithm identifies demographic and behavioural signals common to everyone who converted — including the tyre-kickers. It can't distinguish between them because you haven't told it which ones became customers.
-
3
PMax finds more people who "look like" your converters
Across all placements — Search, YouTube, Display, Discover — the campaign finds more people matching the broad profile. Volume increases. Lead quality doesn't.
-
4
You pause PMax and blame the campaign type
"PMax doesn't work for lead gen." But the campaign was doing exactly what you asked. You asked for form submitters. It found them.
This is the closed-loop problem. And it's not unique to PMax — it's the reason lead gen campaigns of all types underperform. PMax just makes the consequences more visible because the algorithm is more aggressive in its audience expansion.
What closed-loop conversion data actually means
Closed-loop conversion data means importing what happens to a lead after the form is submitted back into Google Ads as offline conversion events. Not just "form submitted" — but also:
- →MQL (Marketing Qualified Lead): the lead has been reviewed and meets your basic qualification criteria — right location, right job type, realistic budget.
- →SQL (Sales Qualified Lead): your sales team has engaged, a quote has been issued, or the lead has moved to active consideration.
- →Confirmed Sale / Booked Job: the lead became a customer. This is what you actually want Google to find more of.
When these signals are fed back into Google Ads — typically through Google Ads' offline conversion import feature — the algorithm can see the full journey. It learns that certain types of people, at certain times, searching certain queries, convert at each stage of your pipeline. Smart Bidding can then start optimising for confirmed customers rather than form submitters.
This is what makes PMax genuinely powerful for lead gen. Once the algorithm has quality conversion signal, its ability to find similar audiences across multiple placements becomes an asset rather than a liability. The same ML that was amplifying your tyre-kicker problem is now finding people who look like your best customers — and finding them everywhere Google operates.
How to implement offline conversion imports for lead gen
The mechanics of offline conversion imports vary by CRM, but the principle is consistent:
Step 1: Capture the GCLID at form submission
Google Click ID (GCLID) is a parameter appended to your landing page URL when someone arrives via a Google Ad. You need to capture this in your CRM against each lead record. Most CRMs (HubSpot, Salesforce, ActiveCampaign) support this natively or via a hidden field on your form.
Step 2: Tag each lead as it progresses through your pipeline
When a lead is qualified as an MQL, your team marks it in the CRM. When a quote is sent, it becomes an SQL. When a job is booked, it's a confirmed sale. These stage transitions are the events you'll import back to Google.
Step 3: Export and import via Google's offline conversion upload
Google provides a template for uploading offline conversions: GCLID, conversion name, conversion time, and optionally a conversion value. This can be done manually via CSV upload or automated via the Google Ads API. Most mid-tier CRMs have native integrations or Zapier workflows that handle this automatically.
Step 4: Set conversion value weighting
Assign relative values to each conversion stage. If a confirmed sale is worth $5,000 to your business, you might assign a form submit a value of $10, an MQL $50, an SQL $200, and a confirmed sale $5,000 (or a fraction of the job value). This guides Smart Bidding to weight the pipeline stages appropriately in its optimisation.
PMax vs Search: When to use which
Start with Search when
- –Your account is new and has fewer than 30–50 conversions per month
- –You don't yet have offline conversion imports set up
- –You need full visibility into which search queries are converting
- –Your budget is limited and you need precise control over where spend goes
Add PMax when
- –You have closed-loop offline conversion data flowing in reliably
- –Your Search campaigns are volume-constrained and you want additional reach
- –You have creative assets (images, video) to populate the asset groups effectively
- –You're running across multiple locations and want the algorithm to find hyper-local demand signals
For most home services businesses, the path is: fix tracking → build Search campaigns → establish conversion history → add PMax once the data foundation is solid. PMax is not a shortcut; it's a multiplier. It multiplies whatever signal you give it — good or bad.
The asset group quality problem
Conversion data aside, the other major reason PMax underperforms for lead gen is poor asset quality. Google's guidance on PMax is to provide as many creative assets as possible — multiple headlines, descriptions, images, and ideally video. When these assets are thin or generic, the algorithm has less to work with, and performance suffers.
For a solar installer or HVAC business, this means:
- →Headlines that speak to specific intent — "HVAC Installation Sydney", "Split System Air Con Installed This Week" — not generic brand messaging.
- →Images that show real work — completed jobs, before/after, equipment — not stock photography of people on laptops.
- →Separate asset groups for different service lines if your business covers multiple categories (installation vs. service/repair, for example).
Asset quality and conversion data quality are the two inputs PMax needs to work properly. Most accounts that struggle with PMax are deficient in at least one.
"PMax is not a campaign type problem. It's a data quality problem wearing a campaign type hat."
What good looks like: the closed-loop PMax account
In accounts where offline conversion imports are set up correctly and PMax is running alongside Search, the typical outcome over 90–180 days is:
- →Declining cost per confirmed sale — as the algorithm builds understanding of what a good customer looks like, cost-per-acquisition at the revenue-generating stage tends to fall even if cost-per-form-submit appears to rise slightly.
- →Improved lead quality scores — your sales team starts noting that leads are better qualified, arriving with more context, more realistic about timeline and budget.
- →Broader geographic reach without quality degradation — PMax finds demand in suburbs and postcodes that Search alone might not have covered, without the quality tradeoff that typically comes with geographic expansion.
The 90–180 day timeframe is significant. Google's Smart Bidding learning period for PMax is real — the algorithm needs time and volume to build reliable conversion models. Patience, and resisting the urge to make major changes during the learning period, is part of making it work.
The bottom line
Performance Max is not the right campaign type for every lead gen account at every stage. For accounts without proper conversion infrastructure, it's genuinely the wrong choice — start with Search, build your conversion foundation, and add PMax once the data is flowing.
But for accounts that have done the work — GA4 properly configured, call tracking with duration thresholds, CRM integrated with offline conversion imports feeding MQL, SQL, and confirmed sale data back to Google — PMax becomes one of the most effective tools available for scaling lead generation across multiple locations without proportional cost increases.
The goal isn't to avoid PMax. The goal is to get good enough data that PMax can do what it's actually designed to do.
Want the full framework — including the tracking setup, CRM integration steps, and when to introduce PMax into your account structure? Download our free guide: The Complete Guide to Lead Generation with Google Ads.