Performance Max (PMax) is one of the most talked about campaign types in paid search right now. It promises access to the full Google inventory, smart automation, and better results with less manual work. For a lot of advertisers, it does deliver on that. However, it also comes with a specific set of problems that, if left unchecked, can quietly drain budget and give you a misleading picture of how things are actually performing.
Hopefully, by the end of this blog, you will be able to spot these issues in your own campaigns and know what to do about them.
Brand terms are eating into your budget
This is probably the most common PMax issue we see when auditing accounts. It happens when your PMax campaign starts showing ads to people who have already searched for your brand by name. These are users who were almost certainly going to click through anyway, so paying to capture them through PMax is not moving the needle. It is just inflating your reported figures with conversions you were always going to get.
Because PMax takes priority over Standard Shopping campaigns in Google’s auction, it will happily intercept branded traffic if you let it. Your PMax numbers look great. Your brand campaign quietly loses impression share. The two are connected.
How to fix it
The fix is straightforward: apply brand exclusions at the campaign level. Go into your PMax campaign settings, find Brand Exclusions, and add your brand name plus any common variants or misspellings. This stops the campaign from serving on branded queries and pushes that traffic back to a dedicated brand search campaign, where you have more control and a lower cost per click.
Running a separate brand campaign alongside PMax is best practice. It keeps branded and non-branded performance clearly separated, which makes it much easier to understand what PMax is actually contributing in terms of new demand.
Using existing warm traffic to pad performance
This one is closely related to the issue above but goes a bit further. Even without serving on branded searches, PMax can still lean heavily on existing warm audiences. If your remarketing lists, customer match data, and recent website visitors are all accessible to the campaign, Google’s algorithm will gravitate towards them because they convert at higher rates. This looks great in your reporting but does not represent genuine new customer acquisition.
The algorithm is doing exactly what it is designed to do: find conversions efficiently. The problem is that from a business perspective, spending budget on people who were already in your funnel is not the same as winning new customers.
How to fix it
Upload your first party data audiences and apply them as exclusions within the campaign. The most useful exclusion audiences to build are:
- Recent purchasers (last 30, 60, or 90 days depending on your purchase cycle)
- Existing customer lists uploaded via Customer Match
- Recent website visitors who are already in your remarketing pool
Once these exclusions are in place, you are effectively pushing PMax up the funnel. Rather than mopping up warm demand, it now has to go and find genuinely new audiences. This is where the campaign operates best: at the top of funnel (TOF), reaching people who have not yet interacted with your brand and introducing them to it across Search, Display, YouTube, and more.
It is worth thinking about how this fits into the wider structure of your account. PMax with first party exclusions applied is a prospecting tool. It should not be expected to close the sale on its own. That is where your search campaigns come in. Search is much better suited to capturing lower funnel intent, where someone is actively researching a product or ready to buy. PMax brings new people into the funnel at the top, search picks them up further down. The two work together, but only when PMax is set up to stay in its lane.
Keep your customer lists refreshed on a regular basis. Monthly updates are a sensible cadence for most businesses. For more on how to use audience targeting effectively across your campaigns, we have covered this in detail elsewhere on the blog.
Pulling in low-quality leads
For lead generation businesses, this is one of the most frustrating things about PMax. Because the campaign is optimising towards whatever conversion action you have set up, and because it is casting a wide net across a huge range of placements and audiences, it often pulls in a significant volume of leads that simply do not go anywhere. Form fills from people who had no real interest, enquiries that never pick up the phone, contacts that your sales team immediately discard as unqualified.
The issue is that Google does not know the difference between a lead that turns into a customer and one that goes nowhere. As far as the algorithm is concerned, a conversion is a conversion. So it keeps doing what it thinks is working, optimising towards more of the same low quality submissions, and your cost per actual sale quietly climbs while your reported cost per lead looks fine.
How to fix it
The answer is offline conversion imports. Rather than telling Google that a lead is a success the moment someone fills in a form, you feed qualified conversion signals back into the platform after the fact, once your sales team has had a chance to assess the lead. This gives the algorithm a much more accurate picture of what a valuable conversion actually looks like, and over time it will start to find more of them.
Offline conversion tracking works by capturing the Google Click ID (GCLID) at the point a user submits a form. This ID ties that user back to the specific ad interaction that brought them to your site. When a lead is later marked as qualified in your CRM, that GCLID gets imported back into Google Ads alongside a conversion event, telling the platform that this particular click was worth something. For a fuller explanation of how the process works, our offline conversion tracking guide covers it in detail.
Setting it up with Zapier and your CRM
You do not need a developer to get this working. If your business uses a CRM like HubSpot, Salesforce, or Pipedrive, you can connect it to Google Ads via Zapier and automate the whole process. Here is how it works in practice:
- Make sure your forms are capturing and storing the GCLID. This usually involves a hidden field on your form and a small piece of JavaScript that passes the GCLID from the URL into that field when someone lands on your page.
- When a lead is created in your CRM, the GCLID gets stored against that contact record.
- In Zapier, set up a trigger based on a deal stage or lead status in your CRM. For example, when a lead is marked as Qualified or a deal moves to Proposal Sent.
- The Zap then fires a conversion event to Google Ads using that stored GCLID, importing the conversion at the point the lead reaches that stage rather than at form fill.
- In Google Ads, set the imported conversion as your primary optimisation goal for the campaign and move the original form fill to a secondary conversion so it is still tracked but no longer used for bidding.
Once this is in place, PMax is no longer rewarded for generating junk. It is only receiving positive signals when a lead has genuinely progressed. The algorithm adjusts accordingly and the overall quality of what comes through improves. If you want to go further and track performance all the way to keyword level within your CRM, our blog on improving lead quality using CRM keyword tracking is worth a read.
Not providing audience signals
Performance Max does not use keywords in the traditional sense. Instead, it relies on audience signals to understand who your ideal customers are before it goes and finds more of them. Audience signals are optional, which means a lot of advertisers skip them or add just a couple of broad in market audiences and call it done.
Without strong signals, the algorithm is essentially starting from scratch. It will figure things out over time, but you will burn budget during the learning phase while it does. The better the signals you provide up front, the faster and more efficiently the campaign learns.
How to fix it
Layer in several types of signals rather than relying on just one. Useful options include:
- Customer match lists from your CRM or email platform. These are the strongest signal you can give Google because they represent people who have already bought from you.
- Website visitor audiences, especially people who have reached a product page or got as far as checkout without converting.
- Custom segments built around relevant search terms your ideal customers are likely to have used, or competitor and category URLs they might visit.
- In market audiences where they are genuinely aligned with your customer profile.
Audience signals are not hard targeting constraints. Google can still serve beyond them, but they give the algorithm a meaningful head start. For more on building effective PPC audiences, take a look at our dedicated guide.
Overlapping with your other campaigns
PMax has a broader tendency to step on the toes of other campaigns across your account. Because it takes priority over Standard Shopping campaigns in the auction, and because it can go up against Search campaigns on non branded queries too, launching PMax without thinking about your existing campaign structure can lead to a redistribution of conversions rather than any actual growth.
You launch PMax, conversions start coming in, and it looks like a success. Meanwhile your Shopping or Search campaigns lose impression share and their numbers drop. The total account performance might be roughly the same as before. You have just added a new layer of opacity on top.
How to fix it
Before launching PMax, document the performance of every other campaign in your account so you have a proper baseline to compare against. After launch, monitor impression share changes in your Search and Shopping campaigns over the first 60 days. A drop that correlates with the PMax launch date is a signal worth investigating. Running a regular PPC audit is one of the most effective ways to catch this early.
URL exclusions are also useful here. If you have well performing Search or Shopping campaigns targeting specific product categories, you can exclude those URLs from PMax to reduce direct competition between your own campaigns.
Conclusion
Performance Max is not a campaign you can set up and leave running. The issues above are not inevitable, but they are common, and they all come from the same place: not giving the campaign the right structure, signals, and oversight to do its job properly. Get those foundations right and PMax can be a genuinely powerful part of your paid media strategy. Get them wrong and it will look like it is working when it is not.
If you want support with auditing or setting up your Performance Max campaigns, get in touch with the Embryo PPC team and we will take a look.
