Tag: Positive Sparks News
Are you spending more on PMax… but not seeing more sales?
If Performance Max (PMax) feels like a “black box,” you’re not imagining it. You feed Google products, creatives, and budget… and it spends across Search, Shopping, YouTube, Display, Discover, and Gmail. Amazing when it works. Brutal when it doesn’t.
And here’s the uncomfortable truth: most “PMax isn’t working” situations aren’t platform problems: they’re control problems. The leaks usually come from:
- Unqualified traffic (wrong people, wrong locations, wrong placements)
- Weak assets (creative that doesn’t persuade)
- Unclear measurement (bad conversion data = bad optimisation)
So let’s plug the holes. Below are 7 control secrets we use with e-commerce brands to stop budget waste and turn PMax into a predictable growth engine.
1) Are you letting Google bid “however it wants”? Lock your bidding strategy.
PMax will happily chase volume if you don’t give it a strong constraint. So the question is: do we want efficiency (ROAS) or do we need cost control (CPA)?
What to do (pick one and commit for 2–4 weeks)
- Use Target ROAS if you have consistent conversion value tracking and enough purchases coming in (think: stable weekly volume).
- Use Target CPA if you’re tighter on budget, you’re still validating product-market fit, or your AOV is volatile.
Numbers to guide you
- If you’re getting fewer than ~30 conversions per month, Target ROAS can get erratic because the algorithm lacks signal.
- If your margins are thin, a ROAS target that’s too aggressive can choke volume: bear in mind PMax may simply stop spending.
A practical approach we like
- Start with Maximise conversion value for a short window (often 7–14 days) to gather signal, then graduate to Target ROAS once the campaign stabilises.
- Don’t “tweak daily.” Major bidding changes can reset learning, which often means paying for the same lessons twice.
Action checklist
- Confirm your primary goal (profitability vs acquisition volume)
- Set one bidding strategy and hold it steady for 2–4 weeks
- Adjust targets in small steps (e.g., 10–15% changes)
2) Are you leaking spend into the wrong places? Use placement and category exclusions.
Ever looked at your results and thought, “Why is my premium product being shown next to questionable content?” Or wondered why YouTube placements are eating spend with little to show for it?
PMax gives you reach, but not all reach is equal.
Where budget leaks happen
- Low-quality apps and sites on Display inventory
- Sensitive content categories that don’t match your brand
- Channels that generate clicks but don’t convert (especially when tracking is messy)
What to do right now
- Add placement exclusions and content category exclusions.
- Review placement insights regularly and keep pruning.
Action checklist
- Exclude sensitive categories that don’t fit your brand
- Review placements weekly for the first month
- Maintain a rolling “blocked placements” list
3) Are you paying for the wrong people to “discover” you? Tighten audience signals and exclusions.
Let’s face it: PMax will explore. That’s part of the promise. But exploration without boundaries is where budgets quietly bleed.
Here’s the nuance
Audience signals in PMax aren’t strict targeting the same way they are in some other campaign types. They’re guidance. But guidance matters: especially early on.
What to build into your audience signals
- Customer Match (past purchasers, email list)
- High-intent segments (cart abandoners, product viewers)
- Custom segments based on competitor terms or category intent
- Your best geo segments if you have regional performance differences
And yes: use exclusions where it makes sense
If you know certain demographics never convert, exclude them. If your product can’t ship to certain areas, exclude them. If you’re B2C, stop bleeding spend on irrelevant B2B intent.
Action checklist
- Upload Customer Match lists
- Add remarketing signals (site visitors, cart abandoners)
- Exclude non-serviceable geos and clearly irrelevant demos
- Revisit audience performance every 7–14 days
4) Are your assets “technically complete” but commercially weak? Manage asset performance like a creative lab.
A common e-commerce trap: you upload the minimum required assets, get the green tick, and move on. But PMax is creative-led. If your assets don’t do the heavy lifting, PMax will buy attention that doesn’t convert.
Google will label assets as Best, Good, or Low. Don’t treat that as decoration: treat it like a to-do list.
What we’ve seen work (consistently)
- Refresh creative every 4–6 weeks to fight ad fatigue and restart learning momentum.
- Build assets that answer real buying questions:
- What problem does this solve?
- Why is this better than the cheaper alternative?
- What risk does the customer feel: and how do we remove it?
A quick real-world example
One of our e-commerce clients in the wellness space had strong products but generic creative: “Natural. Effective. Trusted.” Nice words, no reason to buy today. We replaced that with:
- a specific benefit headline,
- a “results timeline” claim (truthful and compliant),
- and a simple offer-based CTA.
Result: CTR went up, and more importantly, conversion rate rose enough that ROAS improved without changing budget.
Action checklist
- Audit asset ratings weekly (Best/Good/Low)
- Replace “Low” assets with new angles, not tiny edits
- Rotate new images and headlines every 4–6 weeks
- Create at least 3 messaging angles (benefit, proof, offer)
5) Do you rely on PMax for everything? Pair it with manual Search to keep your “money keywords.”
PMax can capture demand. But if you hand it the keys to your highest-intent search terms with zero structure around it, you may lose the clarity you need to scale.
The question we ask is: “Which searches do we absolutely need to own?”
The hybrid approach
- Use manual Search campaigns (or standard Search) to:
- Defend brand terms
- Control spend on high-intent non-brand keywords
- Test new value propositions with cleaner query feedback
- Let PMax do what it’s good at:
- Shopping + multi-channel expansion
- Finding incremental pockets of demand
- Scaling when you’ve got strong signals and strong creative
Why this reduces waste
Because you regain visibility and intent control on your best keywords, while PMax explores in the background with boundaries.
Action checklist
- Separate brand search into its own campaign
- Build a non-brand Search campaign for your 10–30 highest-intent terms
- Use query insights to feed better PMax creative and landing pages
6) Is your budget too small (or too fragmented) for learning? Fund the algorithm properly: or simplify.
This is the one nobody wants to hear: underfunded PMax campaigns often waste more money than they save.
Machine learning needs volume. Without it, the system guesses. And guessing is expensive.
Benchmarks worth knowing
- A common lower-end minimum is around $45/day just to give the system enough signal to move.
- For more reliable optimisation, many e-commerce brands see stability closer to $300–$500/day depending on AOV, conversion rate, and competition.
- If you’re under $2,000/month total, you might be better off with simpler, more controllable campaign types until you’ve built momentum.
Another quiet leak: too many campaigns
If you split budgets across multiple PMax campaigns too early, each campaign can sit in permanent learning. The result? Lots of spend, not much certainty.
Action checklist
- Consolidate if you’re splitting budget too thin
- Aim for enough weekly conversions per campaign to stabilise
- Avoid major structural changes for 2–4 weeks once live
7) Is your tracking lying to Google? Fix conversion measurement before you optimise anything else.
If conversion tracking is off, PMax doesn’t just perform poorly: it optimises in the wrong direction.
So ask yourself: “If I doubled spend tomorrow, do I trust the numbers enough to stay calm?”
What “bad data” looks like in e-commerce
- Purchases duplicated (inflated revenue)
- Purchases missing (undercounted conversions)
- Wrong attribution windows causing weird spikes
- Data gaps from consent mode or tagging issues
- “Add to cart” accidentally set as a primary conversion
What to do
- Ensure Purchase is the primary conversion (and properly deduped).
- Validate that revenue matches your backend (Shopify / WooCommerce) within a reasonable variance.
- Use GA4 and Google Ads conversion tracking intentionally: don’t just “turn on everything.”
If you want a deeper measurement reset, our team also supports GA4-focused projects via our analytics work: https://positivesparks.com/google-analytics-agency
Action checklist
- Confirm Purchase is the only primary conversion (unless you have a clear strategy)
- Check revenue parity vs backend weekly
- Audit deduplication (especially with multiple tags/apps)
- Review attribution settings and conversion windows
A quick “budget leak” diagnostic: where would we look first?
If you’re unsure which lever matters most, here’s a simple priority order we use:
- Conversion tracking integrity (if this is wrong, everything is wrong)
- Bidding strategy + realistic targets
- Creative strength + asset refresh cadence
- Placements + category exclusions
- Audience signals + exclusions
- Budget sufficiency + consolidation
- Hybrid structure with manual Search
You don’t need perfection: you need progress. Plug the biggest hole first, then move down the list.
How we approach this at Positive Sparks (and why our model changes the conversation)
A lot of agencies get paid the same whether your ads perform or not. That’s not exactly a recipe for obsession, is it?
At Positive Sparks, we offer a commission-based model: meaning our incentives are aligned with yours. When performance improves, we win together. When it doesn’t, we don’t hide behind dashboards; we go hunting for leaks.
If you’re curious how that works in practice, you can see the breakdown here: https://positivesparks.com/how-our-fees-work
And if you’d rather just talk through your PMax setup and find the fastest win: https://positivesparks.com/speak-to-us
What are you going to tighten first?
- Which of the 7 leaks feels most “painfully familiar” right now?
- Are you more worried about efficiency (ROAS) or wasted volume (irrelevant traffic)?
- And be honest: do you fully trust your conversion tracking today?
If you want, share your current daily budget, AOV, and rough monthly conversions, and we can sanity-check which control lever will likely move the needle fastest.
