PMax - good or bad? Real talk.

harper-local

I posted recently about how my ex-Google Ads manager sabotaged my account, and when I mentioned I was running PMax, I got some funny looks. The theory behind letting AI choose everything sounds decent, but the reliability is dreadful. Don't trust it blindly.

The biggest issue for me? PMax forces you onto the Display Network. Google's display is riddled with fake traffic and fraudulent clicks, and Google knows it - they rely on it for revenue. You can't opt out. If you've got e-commerce with checkout data feeding back into Google, you might be fine. but for lead gen? It's a nightmare.

i hit a wall trying to set audience targeting - couldn't even pick an age group. Some people swear by classic Shopping campaigns where you tweak every setting manually. Does that make PMax bad? Not entirely, but if you're doing lead gen, you need a system to feed bad leads back into the platform. Otherwise you'll just get junk leads you can't reach by phone or email.

so far I'm leaning away from it for local campaigns. Too much waste, not enough control.

logan-lifecycle

Honestly, the concept of feeding AI your audience segments and letting it optimise the journey is fine in theory. But the automation logic is brittle - triggers misfire, segments bleed, and the reliability is absolute garbage. Do. Not. Trust.

ctrjunkie

PMax isn't bad per se - the real issue is people jump in before they've got proper controls in place. The reach across Google's inventory is genuinely massive, and the automation can sniff out conversions you'd never spot manually. Works a treat for e-commerce brands with a solid product feed, and you barely have to touch it.

But the trade-off is brutal. You get almost no visibility into where the budget actually goes. Audience, location, query data? Basically a black box. If your tracking or feed quality is shaky, it'll burn money fast. Google loves to over-expand targeting, so you end up with branded traffic inflating your results, remarketing carrying the load, and low-quality Display or YouTube views mixed in. You can't even see the search terms that triggered it.

That's why most experienced advertisers I know still run standard Shopping and Search campaigns alongside it, keeping branded and non-branded separate. PMax only really works well when your conversion tracking is rock solid, you've got historical data to feed it, the product feed is clean, and the account structure is mature.

On a brand new account with hardly any conversion data and weak creatives? It's a fast track to wasting money. So no, the AI isn't the enemy. The problem is automation without clean data and strategic control - that's how you end up with a very expensive lesson.

storyweaver

oh, absolutely. PMax is like a hyperactive puppy - great fun until it pisses on your rug because you didn't watch it closely.

the thing with Standard Shopping is you're in control - ads only fire when someone's actually looking for your product. PMax? It's showing up everywhere: Shopping, Search, YouTube, Gmail, Maps, and every random Display pop-up that'll take your money. Without a tight leash, the AI treats your budget like a buffet - shoves it into low-intent impressions that look great in reports but do bugger all for revenue.

You already saw that with the age-targeting mess. PMax is basically flying blind - won't tell you which search terms drove sales, won't split budget by channel, won't give you a clue where your money actually went. It's a black box with a smiley face on it.

verdict: it's not evil. For big e‑commerce feeds, it can be bloody brilliant. but you need to hover over it like a helicopter parent, or it'll blow your budget on garbage traffic and leave you wondering where the conversions went.

zoecreative

PMax is basically the display network in a trench coat, and Google knows the display network is riddled with fraudulent traffic. They depend on it to hit quarterly numbers, so of course every "AI-powered" solution includes display with no opt-out. Shocking.

If you've got a checkout funnel feeding purchase data back into Google Ads, you're fine - the algorithm will optimise toward people who actually pay. For lead gen, you need a system that feeds bad leads back into the machine, otherwise you'll just get garbage leads you can never reach by phone or email.

miaaffiliate

PMax can work, but only if the foundations are solid. You need 50+ conversions per month, accurate tracking, and tight audience signals - customer list plus website visitors - alongside a full asset group with 15+ images, 5 videos, and strong headlines. Miss any of those, and it'll underperform hard.

The most common failure i see: launching with no audience signals at all. Then the budget gets torched on Display and YouTube placements where nobody's ready to buy. Before you go live, shove your customer list and warm website visitors into the audience signal settings in the asset group.

If you're not sure your account is ready, there's a free health check at adpredictor.ai/en/tools/account-health that spots set-up gaps in about 60 seconds. worth running before you risk your budget.

maya-b2b

for me, PMax is a double-edged sword. it works brilliantly when you've got solid conversion history and updated customer lists-basically, you're feeding Google enough data so it knows who to chase. without that, it'll burn budget on random clicks. i tend to use PMax for brand awareness or top-of-funnel plays rather than chasing high intent. search campaigns still outperform for that last-click, ready-to-buy audience.

jamesaso

I've run Pmax for app install campaigns, not ecom. A mate with a health app is dropping $1k+ a day on it and seeing solid ROAS. The key is feeding it good first-party data - blast generic creative in there and it'll waste budget on garbage placements. works if you keep a tight leash on asset groups

growthgeek

PMax isn't bad because it's a bad campaign type - it's bad because Google will happily burn your budget on bottom-feeder Display placements the second you give it half-empty asset groups. The platform is a lazy algorithm: if you don't spoon-feed it every single slot, it assumes you don't care and shoves out whatever generic auto-generated garbage it pleases.

The creative bottleneck is real. Fifteen images and five videos across multiple aspect ratios? that's a full production shoot for most teams. i've been there - spending thousands on lifestyle photography just to test variations that might not even work.

i've been using an AI tool to solve it, uploading flat product shots and letting it churn out lifestyle product images and short video ads in all the ratios PMax demands. It means i can max out the asset group in an afternoon instead of a month. that stops Google from serving your audience the ugly auto-generated trash it loves to fall back on.

but honestly, the fact that you need an external tool just to stop the algorithm from wasting your money says everything about how broken PMax is right now.

buggedout

PMax success is entirely down to the guardrails you set and the quality of your conversion tracking. without those, you're just burning budget.

Lower lead quality compared to standard Search is real - not just low intent, but outright bot traffic and spam.
Cost per click on Search is climbing steadily while margins stay flat, so PMax is becoming inevitable for most verticals.

i run Google Ads for service businesses, and we're using PMax more and more. The catch is that to make it work you need proper campaign structures and ongoing system management - not something a freelancer or small business can just switch on and expect results.

dailygrind

PMax works well when conversion data is sufficient. Without that, Google shifts spend to Display and YouTube instead of Search and Shopping. From my experience in FinTech growth, that's consistent.

ranktracker

that 'sabotaged by your ex' bit is mental, but I've seen it happen more than you'd think. had a client where someone from the agency side 'helped' by flipping feed settings and swapping entire asset groups without a word. Performance went off a cliff overnight. The real kicker? Nobody thought to check change history and line up the timestamps with the performance drop. You can spot it every time if you just dig into the logs.

harper-local

Yeah, that's pretty much where I'm at with Pmax too. for local lead gen, it's a black box that loves burning budget on branded searches or some random display placement that never converts. You can't even see the search terms properly half the time. i've tested it on a few local service clients - foot traffic and call tracking just went muddy. Standard search campaigns still give me way more control for map pack dominance. Pmax feels like Google's way of automating waste

harper-local

I've been down that exact road - that trap is real. With $45 AUD a day, PMax will burn through it fast on low-quality broad traffic. if you haven't even launched a standard Shopping campaign, kill the PMax. For an online shop, Shopping should be your bread and butter, not Search. Search eats budget on branded clicks while Shopping actually shows products to people who want to buy. run a standard Shopping campaign first, then maybe add a tight Search campaign later once you've got conversion data. Keep it simple - don't split that tiny budget across both

harper-local

i hear you. Getting a grip on PMax takes some doing. here's how i lock it down for local lead gen.

first, asset groups are your friend. create one per service area or location, not a blanket campaign. Strip out all the broad keywords Google wants to throw in-use exact match negatives from day one. Turn off "Display Network" expansion if you can (sometimes hidden under advanced options). Then, set a strict conversion action: phone calls with call tracking only, not any random form fill.

also, add location exclusions for areas you don't serve. and don't let it run on "maximise conversions" out the gate-start with a target CPA that's actually profitable. google will overspend if left to its own devices.

if it still feels loose, pause the campaign for a day, create a new one with exactly those settings, and give it zero room to "learn" beyond real conversions. works for me

harper-local

right, so it's an online shop, not lead gen. For a straight ecommerce setup, keeping a Pmax running for five months without knowing the numbers is risky. have you actually looked at the conversion data? if it's driving profitable sales, don't bin it just because some thread told you Pmax is garbage. but if you're seeing low ROAS or it's cannibalising organic search, then yes - swap to standard Shopping campaigns. Run a proper audit first: check the product feed, search term reports, and segment by device and time of day. pull the plug only if the data says it's underperforming compared to what you'd get from a manual Shopping setup.

harper-local

Mate, i get the wariness around links - too many dodgy ones floating about. but you're spot on with the theme: Pmax is built for shops that already have a bit of traction.

You've got over 300 sales - that's a solid base. For customer lists, pull them from your order history. export from your e-commerce platform (Shopify/WooCommerce/whatever you're on) as a CSV. Then upload that list into Google Ads as a customer match audience. That'll give Pmax a decent seed to work with instead of starting blind from zero data.

One thing nobody tells you: clean the list first. remove duplicates, bad emails, and any incomplete addresses. Helps the algorithm not waste time on duds.

harper-local

I get where you're coming from. Pmax can be a black box, especially with zero historical data. My rule of thumb for local clients? Pmax needs at least 60 days of conversion data before it stops doing weird stuff. For a store five months in that's probably still building a solid audience list, it'll swing wildly.

I'd bin the Pmax and go with two standard Shopping campaigns: one for your high-margin bestsellers, one for everything else. Monitor search term reports like a hawk. Pmax hoovers up garbage clicks when it doesn't know what to optimise for. You'll get more predictable spend and you can actually see what's working.

If you're chasing foot traffic attribution or phone calls (which is my jam), Pmax is even worse. Standard Shopping at least lets you layer on call extensions properly. Ditch the black box.

storyweaver

Oh bless, PMax and its delusions of grandeur. first thing's first: go to Settings > Brands and create an exclusion list for your own name. Stop that algorithm from patting itself on the back every time someone already had you in mind. You want it to actually find new customers, not just claim credit for foot traffic.

Also, shove a negative keyword list in there - block 'how to' and 'free' searches. those tyre-kickers will drain your budget faster than a bad influencer deal. works wonders