Performance Max Campaigns: Budget Allocation Without the Guesswork

The Real Problem With Budget Allocation in Google Ads

 Or they throw everything at it, cannibalise their Search and Shopping campaigns, and end up with a lower overall account ROAS than they started with.

The thing is, neither of these approaches makes sense. Performance Max has become the default for Google Ads enterprise advertisers, who now allocate 70% or more of their total spend to it, but the nuance of how to allocate that budget is where most accounts fall apart. You need a framework, not a guess.

This guide walks through how to structure your PMAX budget so you’re optimising across channels, protecting your best performers, and actually knowing whether your campaigns are working.

What We Know About PMAX in 2026

Before we get into allocation, let’s establish the baseline. The data is clear: accounts running a hybrid strategy (PMAX plus Search, plus selective Shopping campaigns) consistently outperform accounts running PMAX alone.

The hybrid approach across Search, Shopping, Display, YouTube, and Discover delivers significantly higher conversion rates than relying on a single campaign type. Early adopters of a full-funnel structure see 25-35% improvements in conversion rates compared to PMAX-only accounts.

Here’s why: PMAX lacks search query transparency, it offers limited channel-level budget control, and it struggles when you have new products without historical performance data. It’s brilliant at what it does, cross-channel optimisation at scale, but it’s not designed to handle every scenario your business throws at it.

The second key finding: Target ROAS bidding outperforms manual CPC bidding by 38% on average. This matters because it tells you that once your campaigns are mature enough, your bid strategy choice matters as much as your budget allocation choice.

The 70-20-10 Framework: Your Starting Point

Most mature Google Ads accounts in 2026 run a consistent structure:

• 70% of the budget to Performance Max (your conversion engine)
• 20% of the budget to branded Search campaigns (protection and brand control)
• 10% of the budget to experimental or specialist campaigns (testing, remarketing, new products)

This framework exists because it works. It gives PMAX enough budget to accumulate meaningful conversion data (typically 30+ conversions per month to optimise effectively), whilst protecting your brand terms from price wars and maintaining control over your highest-intent traffic.

For e-commerce brands specifically, PMAX typically sits at 50-60% of total budget. For B2B and service businesses, it’s often 30-40%. The variance depends on your business model and how much control you need over lead quality. A law firm might run a lower PMAX allocation because every lead matters and needs to be qualified. An e-commerce store can afford to let the algorithm loose because purchase intent is clearer.

The key principle: if your PMAX campaign isn’t generating at least 30 conversions per month, it’s data-starved. A well-fed, smaller campaign will outperform a larger but under-optimised one every single time.

Segmenting PMAX Budgets: The ROAS-Based Approach

Once you’ve established your overall PMAX allocation, the next layer is segmenting within PMAX itself. Not all products perform equally, and not all campaigns should receive the same budget.

The ROAS-based approach allocates budget based on historical performance, creating campaign tiers:

Tier 1: High-ROAS Campaign (60% of PMAX budget).

These are your best performers. Set a Target ROAS of 500% or higher if that’s what your historical data supports. Feed these campaigns aggressively. They’ve proven they convert efficiently, so the algorithm will work with proven patterns.

Tier 2: Medium-ROAS Campaign (30% of PMAX budget). 

These are decent performers with room to improve. Set a Target ROAS between 301% and 499%. This tier is where testing happens; you’re validating whether certain products, creatives, or audience signals can move upmarket. Many of your Tier 1 campaigns probably started here.

Tier 3: Low-ROAS Campaign (maximum 10% of PMAX budget).

These are the underperformers or new products. Cap spend here. You want them visible enough to gather data, but not so visible that they waste budget. Set a Target ROAS of 200-300% and monitor whether they improve over time or plateau.

The rationale is straightforward: you’re letting data, not opinion, drive allocation. If your leather jackets consistently return 6:1 ROAS and your denim returns 2.5:1 ROAS, it makes sense to feed the leather jackets and constrain denim until something changes.

Profit Margin Matters (More Than You Think)

There’s a catch with pure ROAS-based budgeting: it ignores margins. A product with 5:1 ROAS on a 15% margin is actually less profitable than a product with 3:1 ROAS on a 40% margin. If you’re running an e-commerce business with variable product margins, this distinction matters a lot.

Some businesses build in a second layer using gross profit margin (GPM) to refine their allocation:

• High-GPM products (40%+ margins): Can tolerate lower ROAS targets because the absolute profit per sale is high. Allocate accordingly.
• Medium-GPM products (20-40% margins): Require balanced ROAS targets to remain profitable.
• Low-GPM products (<20% margins): Need higher ROAS targets to justify the spend.

This is particularly relevant for brands that bundle high-margin and low-margin products. You might suppress spend on a low-margin item that carries decent ROAS, and increase spend on a high-margin item with lower ROAS, because the bottom line matters more than the multiplier.

Setting Up Your Campaigns: The Bid Strategy Transition

New PMAX campaigns need a specific setup sequence to avoid wasting budget during the learning phase.

Start all new campaigns on Maximize Conversions. This gives the algorithm room to explore without aggressive CPA or ROAS constraints that might throttle delivery when it has limited data. Run on Maximize Conversions until you’ve accumulated 30-50 conversions.

Only after that threshold do you transition to Target CPA or Target ROAS. Setting aggressive targets on a cold campaign is the fastest way to constrain delivery and extend your learning phase. You’ll see lower overall conversion volume and take longer to reach profitability.

Once you’ve hit the conversion threshold, watch your asset performance. Google assigns Low, Good, and Best ratings to individual creative assets. Replace Low-rated assets immediately, they’re pulling down your quality score and wasting impressions. Study your Best-rated assets to understand what’s working: What messaging resonates? What image style? What video length? Use those insights to inform your next creative batch.

Budget Scaling: The 7-14 Day Rule

One of the most common mistakes is adjusting PMAX budgets too frequently. The algorithm needs time to stabilise after changes.

Scale in increments of 15-20%, then wait 7-14 days before reassessing. Every significant adjustment restarts the algorithm’s learning phase. If you increase the budget on Monday, decrease it on Wednesday, and increase it again on Friday, you’re constantly resetting the optimisation timeline. That’s why performance looks inconsistent.

Set calendar reminders if you need to, but the discipline is simple: adjust, then monitor without acting for two weeks. It’s the hardest part of PMAX management because you want to respond to data immediately. Resist that urge.

For budget decreases, the same principle applies. If a campaign isn’t hitting its target ROAS, reducing spend by 30% and then checking back in three days won’t help. It will just create more volatility. Make a deliberate downward adjustment, commit to it for two weeks, then assess.

Protecting Your Existing Performance

This is where internal alignment matters. If your PMAX campaign is pulling budget from Search, Search performance will decline. That’s not a PMAX failure; it’s a reallocation issue.

The hybrid model protects against this by maintaining separate, well-funded campaigns for your highest-intent traffic (branded Search), whilst giving PMAX room to work on demand generation and consideration traffic across all Google inventory.

In practice:

• Branded Search (your 20%) captures users actively looking for you by name. Protect this budget fiercely. These conversions typically have the lowest CPA and highest quality because intent is unambiguous.
• PMAX (your 70%) captures everything else: generic keywords, Shopping queries, Display placements, YouTube, and Discover.
• Experimental campaigns (your 10%) test new audiences, new products, or new strategies without impacting your core revenue drivers.

The risk of allocating 90% to PMAX is that you eventually lose visibility on branded terms to competitors, your Search CTR drops as competition drives up CPCs, and you’ve put all your eggs in a black box you can’t fully control. The 70-20-10 model avoids that risk.

Scaling PMAX When It’s Working

If your PMAX campaigns are consistently hitting or exceeding target ROAS, the question becomes: how much can you scale?

The answer depends on whether you’re scaling PMAX itself or growing your overall ad budget. If you’re growing your budget and PMAX is your best performer, you might increase its allocation from 70% to 75-80% of total spend. That’s reasonable if your branded Search and experimental campaigns are also performing.

If you’re scaling PMAX at the expense of other channels, slow down. Most accounts that have tried cutting Search to feed PMAX see account-level ROAS decline within 60 days because they’ve lost visibility on high-intent traffic. The algorithm is efficient, but it’s not magic. Efficient scaling respects channel balance.

Measurement: Know What’s Working

The only way to know if your budget allocation is right is to measure it properly. This isn’t complicated, but it’s easy to get wrong.

Track these metrics in your account:

Per-Campaign Level:

• Conversions per month (is this campaign data-rich enough?)
• Cost Per Acquisition (is it meeting your target?)
• ROAS (is it hitting your assigned tier?)
• Impression Share (is there room to scale?)

Account Level:

• Total conversions (month-over-month trend)
• Blended ROAS across all campaigns (hybrid model should protect this)
• Cost Per Conversion (trending down or up?)
• Channel mix (are you maintaining your intended 70-20-10, or has it drifted?)

Use Google Search Console alongside your Google Ads account to spot which pages are driving volume, which aren’t, and whether your landing pages are actually converting or just generating impressions.

Most importantly: measure incrementally. If you restructure your PMAX setup, give it 4-6 weeks of data before concluding whether it worked. Algorithm changes, seasonal patterns, and competitive shifts all add noise to the data. One week of poor performance doesn’t mean your strategy is wrong. One month of trend data does.

Common Mistakes to Avoid

Setting ROAS targets without historical data. If you don’t know what your actual ROAS is, setting a 500% target is a guess. Use historical data to inform targets, then adjust gradually based on performance.

Treating PMAX as black box automation. It’s not. You still need to monitor creatives, assess asset performance, and make deliberate budget decisions. The algorithm optimises; you strategise.

Underfeeding PMAX because you don’t understand it. A well-optimised PMAX campaign with sufficient data typically outperforms a hand-managed Standard Search campaign. Give it room to work.

Ignoring your highest-intent traffic. Protecting branded Search isn’t paranoia; it’s protecting the most valuable traffic you own. Don’t sacrifice it.

Scaling based on one week of good data. Wait. Patterns emerge over time. Anomalies happen every day.

PMAX Budget Allocation: The Framework You Can Use

Here’s the practical checklist for your next restructure:

1. Define your overall allocation: 70% PMAX, 20% branded Search, 10% experimental. Adjust based on your business model, but don’t deviate wildly.
2. Segment PMAX by performance: Create campaign tiers (high/medium/low ROAS) and allocate 60-30-10 within PMAX.
3. Set appropriate bid strategies: Maximize Conversions for new campaigns until 30-50 conversions, then Target ROAS or Target CPA based on your business model.
4. Document your targets: Write down what ROAS or CPA you’re targeting in each campaign tier, and why. This becomes your decision-making framework.
5. Scale incrementally: 15-20% changes, 7-14 day waiting periods. Discipline over impatience.
6. Monitor blended metrics: Watch your account-level ROAS and channel mix. Make sure the optimisation in PMAX isn’t cannibalising performance elsewhere.
7. Review monthly, adjust quarterly: Monthly data informs where attention is needed. Quarterly data informs structural changes.

The businesses running 400%+ ROAS on PMAX aren’t guessing. They’ve built frameworks, respected the data, and maintained discipline. That’s replicable.

If you’re looking at your current PMAX setup and realising it’s been running on fumes and guesswork, our Google Ads specialists can audit your account structure and build a data-driven allocation strategy tailored to your business model. We’ve restructured accounts across ecommerce, B2B, and lead generation, and the pattern is consistent: deliberate allocation beats random distribution every single time.