Performance Max Campaigns: The Complete Optimisation Guide for 2026

Performance Max now accounts for 45% of all Google Ads conversions — yet it remains the most misunderstood and poorly optimised campaign type in most advertisers' accounts. The reason is structural: PMax was designed to hand control to Google's AI, and most advertisers interpreted that as a reason to stop optimising. That interpretation is costing them money every single day.

This guide is written for practitioners managing real Performance Max accounts — marketers and agencies who need more transparency, more control, and better results from a campaign type that Google is pushing harder than ever. If you're newer to Google Ads strategy, you may want to start with our Digital Marketing Strategy Guide for 2026, which covers the full paid search landscape before diving into campaign-type specifics.

The good news: 2024 and 2025 updates fundamentally changed what's possible inside PMax. Search themes replaced the old category insights. Campaign-level brand exclusions arrived. Asset group reporting now breaks down by channel. Customer Match lists no longer require a minimum audience size. These changes mean advertisers who know how to use the controls now have meaningful leverage — and those who don't are still flying blind while Google spends their budget across its entire network.

What Performance Max Actually Is in 2026

Performance Max is a goal-based campaign type that places ads across every Google-owned surface simultaneously: Search, Shopping, YouTube, Display, Discover, Gmail, and Maps. From a single campaign, Google's machine learning allocates budget, selects audiences, chooses ad formats, and adjusts bids in real time based on your conversion goals.

That breadth is its core value proposition. A traditional Search campaign reaches users who are actively searching. A Shopping campaign surfaces product listings. A YouTube campaign builds awareness. PMax is supposed to do all of this in concert, following users across their journey from first exposure to final conversion.

In practice, this doesn't always unfold the way Google implies. A large-scale study by Adalysis covering 3,300 campaigns found that when both PMax and Search campaigns were eligible for the same search terms, Search campaigns consistently had higher conversion rates. This isn't a failure of PMax's technology — it's a signal about how the campaign behaves when left ungoverned. Without deliberate asset group architecture, audience signal guidance, and structural controls, PMax defaults to doing what's easiest: claiming brand search traffic and cheap Display impressions, then reporting impressive aggregate numbers that mask where the real value is coming from.

The practitioner's job in 2026 is to understand this dynamic and use every available lever to guide PMax toward genuinely incremental performance — new audiences, new queries, and cross-channel reach — rather than cannibalising your existing Search campaigns and brand traffic.

Over 1 million advertisers globally now use Performance Max, and Google's own data claims an average 18% lift in conversions at similar cost-per-action for advertisers using PMax alongside their existing campaigns. Adoption jumped from 60% of surveyed advertisers in 2024 to 71% in 2025. Whether those numbers reflect real incrementality or structural cannibalisation depends entirely on how the account is set up.

The 2024–2026 Control Evolution: What Changed and Why It Matters

To understand how to optimise PMax in 2026, you need to understand the control timeline. Early PMax was genuinely opaque — advertisers provided assets and a budget, Google did everything else, and the only feedback was aggregate campaign metrics. Critics called it a black box, and that criticism was largely fair.

The 2024 and 2025 update cycles changed the equation significantly:

Search Themes (2024): Replaced the old Search Category Insights with an actionable input. Advertisers can now add up to 25 keyword-like themes per asset group that tell the PMax algorithm which search intent to prioritise. These behave like broad match keywords — they guide the algorithm toward related query clusters rather than triggering exact matches — but they provide meaningful directional control that was entirely absent in earlier versions. The Google Ads help documentation confirms you can add up to 25 search themes per asset group, with exclusions and negative keywords still applying on top.

Campaign-Level Brand Exclusions (2024): One of the most important additions. Advertisers can now prevent PMax from serving on their own brand queries at a campaign level, separate from account-level negative keyword lists. This is critical for accurate ROAS measurement — without brand exclusions, PMax will absorb brand search traffic that would have converted anyway, inflating its reported numbers and making it appear more effective than it actually is.

Channel-Level Asset Group Reporting (2025): The asset group insights section now shows conversion contribution by channel — Search, Display, YouTube, Gmail, Discovery, and Maps — broken down per asset group. This was the transparency upgrade practitioners needed most. If 90% of your PMax conversions come from Search and you have no video assets, your campaign is essentially functioning as a less-controlled Search campaign. That's a solvable problem once you can see it.

Customer Match Without Minimums (2025): Previously, Customer Match lists required a minimum of 1,000 matched users before they could be used as audience signals. That minimum was removed, making Customer Match viable even for smaller B2B accounts with modest CRM lists.

URL Expansion Controls: URL expansion — where Google automatically creates ads pointing to pages on your site other than your specified landing page — is enabled by default. Google's own support documentation outlines how to add URL exclusion lists to block specific pages (career pages, blog posts, privacy policy, discontinued product pages) from being served, or to disable URL expansion entirely for campaigns where landing page specificity is critical to conversion quality.

Asset Group Architecture: The Foundation of PMax Performance

The asset group is the fundamental unit of Performance Max. Where traditional campaigns have ad groups with keyword lists, PMax has asset groups — collections of headlines, descriptions, images, videos, and audience signals that the algorithm combines and serves across channels. Getting asset group architecture right is the highest-leverage structural decision you'll make.

The core principle is thematic coherence: every asset in an asset group should belong together. When your headlines, descriptions, images, and video all communicate the same specific value proposition to the same specific audience, the algorithm can create better combinations and the resulting ads will be more relevant, more engaging, and more likely to convert.

Asset Group Strategy by Business Type

Ecommerce brands should build asset groups around product categories, not around campaign objectives. Separate asset groups for women's running shoes, men's casual footwear, and accessories — not one group for all footwear. This allows the headlines and product images within each group to be genuinely specific, and the search themes can be tailored to the intent patterns relevant to each category. An asset group mixing multiple product categories forces the algorithm to write generic copy that performs worse across all categories than specific copy would perform for each.

B2B and service businesses should structure around buyer personas or service lines. A digital marketing agency might build separate asset groups for: enterprise businesses seeking paid search management; ecommerce brands seeking shopping campaign management; and professional services seeking lead generation. Each group gets headlines addressing that persona's specific pain points, images relevant to their context, and search themes reflecting how they describe their problems.

Multi-location businesses benefit from location-based asset group segmentation when there are meaningful differences in service offerings, pricing, or competitive landscape by geography. A national trades business with premium pricing in Auckland versus standard pricing in provincial cities should not be serving the same assets with the same ROAS targets across both markets.

Asset Requirements and Quality Standards

Google's asset requirements for a complete PMax asset group include: 15 headlines (30 characters each), 5 long headlines (90 characters), 5 descriptions (60 characters), 5 long descriptions (90 characters), 20 images in multiple aspect ratios, 5 logos, and 5 videos. Meeting the minimum isn't enough. Google's internal testing shows 25–40% better performance for campaigns with comprehensive video libraries compared to image-only asset groups — the algorithm increasingly favours video-enabled groups for Discovery, YouTube, and Demand Gen inventory.

Every asset in your group has a performance label: Best, Good, Low, or Learning. Low-rated assets aren't just underperforming — they're actively dragging down the group's performance because the algorithm is still required to use them in combinations. Audit asset labels weekly and replace Low-rated assets with new variants. If all your assets eventually reach Good or Best, your asset group performance floor rises significantly.

Headlines should not all say the same thing. A common mistake is writing 15 headlines that are variations of the same core message. Google combines these into ad combinations — if all your headlines communicate the same benefit, you lose the ability to test different value propositions and the algorithm can't identify which messaging resonates with different audience segments. Write headlines that cover: your core offer, a key differentiator, a social proof element, a price/value signal, and an urgency or recency hook. That variety lets the algorithm find the optimal combination for each user context.

Search Themes: The Most Powerful New Control in PMax

Search themes are the most operationally significant new feature in Performance Max. Before they existed, you could provide audience signals and hope the algorithm found the right searches. Now you can directly influence which search intent the algorithm pursues for each asset group.

Each asset group supports up to 25 search themes — keyword-like inputs that function similarly to broad match keywords in how they expand to related queries, but which also influence non-search inventory by informing the AI about what your campaign is conceptually about. The Google Ads Help Centre documentation describes them as "optional and additive" — the algorithm still pursues queries and placements it predicts will convert based on your assets and feeds, but search themes provide additional guidance toward queries you know are valuable.

The right way to think about search themes is as thematic intent signals, not keyword lists. Don't add 25 near-identical keyword variants. Add 8–12 themes that represent different facets of your audience's intent: a job-to-be-done theme ("improve Google Ads ROAS"), a solution category theme ("Performance Max management"), a pain point theme ("wasted Google Ads spend"), and a competitor displacement theme ("Google Ads agency alternative"). Each theme should open up a cluster of related queries rather than targeting one specific phrase.

Practical Search Theme Guidelines

Start with high-intent commercial themes. Themes containing words like "buy", "price", "near me", "service", or specific product/service names signal to the algorithm that you want purchase-ready traffic. Broad educational themes ("what is Performance Max") will attract informational traffic that's unlikely to convert, increasing costs without proportional returns.

Use Google Keyword Planner to validate themes. Before finalising your search themes, run them through Keyword Planner to check search volume and competition. A theme that represents a real search pattern will perform better than one you invented that doesn't match how your customers actually phrase their searches.

Align themes with asset group content. If your search theme is "Performance Max optimisation agency" but your asset group headlines are about general digital marketing services, there's a relevance gap. The theme signals intent, but the assets need to deliver on that intent. Misaligned themes and assets produce lower Quality Scores across Search inventory and weaker relevance signals for Display and YouTube.

Limit themes to 5–10 per asset group to start. Too many themes can dilute the signal and slow the algorithm's learning. Begin with your strongest 5–10 themes based on historical search term data from existing campaigns, then expand as the campaign matures and you can see which theme clusters are driving conversions.

Audience Signals: Guiding Without Restricting

Audience signals in Performance Max occupy a unique position: they inform the algorithm's targeting decisions without restricting them. This is fundamentally different from audience targeting in other campaign types. You're not telling PMax to only show ads to people on your audience list — you're telling it that people like the ones on your list are valuable, so start there and learn from who converts.

This distinction matters for audience signal strategy. Because signals guide rather than restrict, the quality of your signal matters more than its size. A Customer Match list of 500 of your highest-value clients will produce better audience signal quality than a broad interest audience of 500,000 people who sort of match your target market. The algorithm learns the conversion-predictive characteristics from your signal, then finds similar people across Google's full audience graph.

Audience Signal Hierarchy

Build audience signals in priority order. The highest-quality signals come from your own data:

Tier 1 — Customer Match lists: Upload your existing customers, your highest-value clients, and your closed-won deals from CRM. These are the people who actually bought from you — the algorithm's job is to find more people like them. Since the minimum list size requirement was removed in 2025, even accounts with 100–200 contacts can use Customer Match effectively.

Tier 2 — Website Visitor Audiences: Segment your website remarketing lists meaningfully before using them as signals. A single "all website visitors" audience is less informative than separate audiences for product page visitors, checkout abandoners, or blog readers. The algorithm learns different things from each segment.

Tier 3 — Custom Intent Audiences: Build custom audiences based on the search queries your highest-intent prospects use. These bridge the gap between your Customer Match data and the broader audience pools Google can access.

Tier 4 — In-Market Audiences: Use Google's predefined in-market audiences as a supplementary signal when your first-party data is thin. These are less precise but still more informative than no signal at all.

It's important to understand that the algorithm will expand beyond all of these signals if it finds conversions elsewhere. PMax regularly converts users who don't match any of your audience signals — that's the feature, not a bug. The signals establish the starting point for learning; they don't define the ceiling of your reach.

ROAS Benchmarks by Campaign Type and Industry

One of the most common PMax frustrations is not knowing whether your ROAS is good, average, or poor relative to comparable accounts. The answer depends significantly on industry, campaign type mix, and how well brand exclusions are set up.

Aggregate data from Focus Digital covering 2025 full-year performance shows Performance Max delivering a median ROAS of 2.57x — lower than Search campaigns at 5.17x but significantly ahead of Smart campaigns (1.72x) and standard Display (0.12x). However, the PMax number requires an important caveat: campaigns without brand exclusions are crediting PMax with conversions that came from branded queries, systematically inflating the true incremental ROAS figure. Studies consistently find that well-governed PMax accounts with brand exclusions report meaningfully different (lower) headline ROAS but demonstrate more genuine incremental revenue than ungoverned accounts with higher nominal ROAS.

Industry variance in PMax ROAS is substantial. Legal services and healthcare verticals — which typically operate on high-intent search queries rather than broad discovery — show Google Ads ROAS of 4.21x and 3.64x overall, but PMax's cross-channel reach adds less value in high-intent verticals where Search already captures the conversion-ready audience. Ecommerce retail and physical product categories tend to benefit more from PMax's Shopping and Display reach. Professional services and B2B accounts should treat PMax ROAS numbers with particular scepticism without brand exclusions in place.

Bidding Strategy: When to Use What, and How to Graduate

Performance Max's bidding strategy should evolve with campaign maturity. The two primary options are Maximise Conversion Value (or Maximise Conversions for lead gen) and Target ROAS (or Target CPA for lead gen). The right choice depends on how much conversion data your campaign has accumulated.

New PMax campaigns should always start with Maximise Conversion Value without a ROAS target. Setting a ROAS target too early — before the algorithm has enough conversion data to understand your account's performance patterns — causes the algorithm to operate in a constrained state, limiting its ability to explore and learn. The Channable best practices guide explicitly recommends starting with Maximise Conversion Value to speed learning, then introducing a ROAS target once volume is stable.

The graduation to tROAS should happen when your campaign is generating 30+ conversions per month consistently. Set your initial tROAS target at 10–20% below your current actual ROAS, not at your aspirational target. This gives the algorithm room to maintain volume while working toward profitability, rather than immediately constraining delivery and starving the campaign of the learning data it needs.

Budget and bidding should never be changed simultaneously. Every significant change to a PMax campaign — budget adjustments above 20%, bidding strategy changes, asset group additions, conversion action changes — resets the learning period. Advertisers who make multiple simultaneous changes prevent the algorithm from ever stabilising, creating a cycle of perpetual learning that never converts into mature performance.

The Learning Period Problem

PMax requires approximately 4–6 weeks of accumulated conversion data to exit the learning period and begin optimising efficiently. During this period, performance is volatile — CPAs will be higher than target, ROAS will be below expectations, and impression share will fluctuate. The instinct to intervene is strong, but premature changes extend the learning period and prevent the campaign from ever reaching stable performance.

The practical solution is to plan for the learning period in your client or stakeholder communications. Set expectations that the first 4–6 weeks of a new PMax campaign are a data collection phase, not an optimisation phase. Set a budget you're comfortable with for that period (lower than your intended steady-state budget if needed) and commit to not making structural changes during it. Evaluate performance at week 6–8, not week 2.

The Cannibalisation Problem: PMax vs Search

The most important structural risk in PMax accounts is brand and intent cannibalisation — PMax claiming credit for conversions that would have happened through your Search campaigns anyway. This matters because it inflates PMax's apparent ROAS while potentially degrading your Search campaigns' performance data, making it harder to optimise either campaign type accurately.

The Adalysis study of 3,300 campaigns found that Search campaigns consistently had higher conversion rates when both PMax and Search were eligible for the same queries. Google's auction prioritisation rules mean that PMax now competes with Search on a pure ad rank basis (as of late 2024 changes), rather than automatically losing to Search. The result is that without active brand exclusions and negative keyword management, PMax will win impressions on queries your Search campaigns were handling efficiently.

The Brand Exclusion Protocol

Every PMax campaign should have brand exclusions active by default, with no exceptions. The setup takes less than five minutes: navigate to Campaign Settings → Brand Exclusions → Add brands, then add your company name, product names, and any branded terms you don't want PMax to claim. This ensures that PMax's reported ROAS reflects genuinely non-brand performance.

Beyond brand exclusions, use account-level negative keyword lists to block queries that belong to dedicated Search campaigns. If you have a competitor campaign bidding on competitor brand terms, add those brands to your account-level negative list so PMax doesn't compete with your own competitor campaign for that traffic.

Search-PMax Campaign Architecture

The recommended architecture for accounts running both Search and PMax:

Exact match branded Search campaign: All your brand-term keywords, branded product names, and URL queries. High bid, high budget share, maximum control. PMax excluded from all these terms via brand exclusions.

Non-brand Search campaigns: Your core commercial keywords with phrase and exact match. These form the precision layer — high-intent, keyword-controlled, with full search term visibility.

Performance Max: Discovery layer. Uses search themes to target intent clusters adjacent to your core keywords, Customer Match and website visitor signals for audience guidance, and cross-channel reach for awareness and retargeting across Display, YouTube, Gmail, and Discovery. Brand exclusions active. URL expansion controlled.

This architecture gives you the control benefits of Search for your known, high-value queries, while letting PMax do what it's actually good at: finding new audiences and intent patterns you haven't explicitly targeted.

Reporting: What to Measure and How to Read PMax Data

PMax reporting requires a different analytical framework than traditional campaign types. The metrics are real but they tell an incomplete story without context about where they're coming from and what they're displacing.

The Three-Layer Measurement Framework

Layer 1 — Campaign-Level Metrics: What Google reports. Conversions, conversion value, ROAS, CPA, impression share, asset group performance, search term insights (limited). These are the starting point, not the end point of analysis.

Layer 2 — Cross-Campaign Impact: What's actually happening in your account. Changes in brand search volume and performance since PMax launched. Overall account conversion rates and trends. Whether non-brand Search campaign impression share has declined. New vs. returning customer ratios in conversions. Assisted conversion attribution from other campaigns. This layer reveals whether PMax is genuinely additive or primarily redistributive.

Layer 3 — Business-Level Validation: The ground truth. CRM-confirmed lead quality from PMax-attributed conversions versus Search-attributed conversions. Revenue from PMax-attributed customers versus comparable Search-attributed customers. Whether leads from PMax channels (Display, YouTube) convert to customers at similar rates to Search leads. This layer — which requires CRM data and offline conversion tracking — is where the real value assessment happens.

Channel Breakdown Analysis

Open your asset group insights and check the channel breakdown for each group. If your PMax campaign shows 85–90% of conversions attributable to Search with almost no contribution from YouTube or Display, your campaign is effectively functioning as an ungoverned Search campaign — all the risk of PMax's lack of keyword control without the benefit of cross-channel reach. The fix is usually video assets: campaigns without video creative cannot access YouTube and Discovery inventory, leaving large amounts of Google's network untouched.

A healthy PMax channel breakdown for an ecommerce account might show 40–50% Search, 20–30% Shopping, 15–20% YouTube/Display, and the remainder in Gmail and Discover. The exact mix varies by industry and objective, but any distribution heavily dominated by Search is a signal that your asset library needs attention.

Search Term Insights vs Search Term Reports

PMax provides Search Term Insights — an aggregated view of the queries triggering your ads, grouped into themes rather than individual terms. This is less granular than the full Search term report available in traditional campaigns, but it's meaningfully useful for spotting problematic theme clusters that should be added to your negative keyword list.

Review search term insights weekly. Look for theme clusters that represent your target market's problems but not your solutions (educational/research queries), competitor brand names appearing in PMax insights despite brand exclusion settings (a signal to check your negative keyword coverage), and geographic or demographic themes that suggest the algorithm is expanding targeting beyond your intended market.

Advanced PMax Techniques for 2026

Feed-Only Performance Max for Ecommerce

For ecommerce advertisers with large product catalogs, feed-only Performance Max — creating a PMax campaign without uploading any text or image creative assets — is a powerful technique that forces the algorithm to focus exclusively on Shopping and product-focused Display inventory. Without creative assets, PMax can't serve on YouTube, standard Display, or Gmail, which means 100% of budget is directed toward Shopping surfaces where product feed data drives ad creation.

This is particularly valuable for advertisers who want PMax's bid automation and audience expansion benefits without the risk of budget leaking to low-intent Display and YouTube impressions. It's the closest equivalent to running an AI-powered Smart Shopping campaign on a larger scale. The downside is limited cross-channel reach — if your objective includes awareness building or retargeting across non-search surfaces, you'll need separate asset-enabled campaigns for those goals.

The Profit Margin Architecture

One of PMax's most underutilised capabilities is the ability to assign different conversion values to different product categories or customer types, then set a campaign-level tROAS target that reflects profit margin rather than revenue. This allows the algorithm to optimise toward actual business profitability rather than gross revenue.

The implementation approach: use custom labels in your product feed to tag products by margin tier (high, medium, low). Create separate PMax campaigns for each tier. Set the highest tROAS target on your highest-margin products, and lower targets (or Maximise Conversion Value) on lower-margin products. The algorithm will naturally push budget toward inventory that earns against its targets, creating a self-managing profit-optimisation system.

Customer Match as a Suppression Signal

Customer Match is most commonly used as an acquisition signal — upload your best customers so PMax can find people like them. But it's equally valuable as a suppression signal. Upload your existing customers to a separate audience list and exclude them from prospecting asset groups. This ensures your acquisition budget is spent reaching genuinely new potential customers, not re-targeting existing ones through expensive paid inventory when they'd likely convert through cheaper direct or organic channels anyway.

The exclusion approach is particularly important for subscription businesses and SaaS companies where the acquisition and retention motions should be completely separate — different messaging, different offers, different channel mix, and different ROAS benchmarks.

Conversion Value Rules

Conversion Value Rules let you instruct PMax to weight certain types of conversions more highly than others based on characteristics of the user or conversion context. Common applications include: weighting conversions from mobile devices more highly if mobile users have higher LTV in your data; weighting conversions from specific geographic areas (Auckland vs provincial NZ) if those markets have different average order values; or weighting conversions from new customers more highly than returning customers if acquisition is your growth priority.

These rules don't change your actual reported conversion values — they only affect how the bidding algorithm weights different conversion opportunities in its real-time auction decisions. The effect is subtle but compounding: over thousands of daily auction decisions, conversion value rules meaningfully shift budget toward the audience segments and user contexts that matter most to your business.

Common PMax Mistakes (and How to Fix Them)

Mistake 1: No brand exclusions. The single most common and costly PMax error. If your PMax campaign's ROAS looks suspiciously high compared to your non-brand Search campaigns, brand exclusions are almost certainly not in place and PMax is absorbing branded conversions. Fix: Enable brand exclusions immediately in Campaign Settings.

Mistake 2: Single asset group for all products or services. One asset group for everything produces generic ad copy that performs worse than specific copy for each category. The algorithm can't write a compelling headline about running shoes and formal shoes from the same asset group. Fix: Create separate asset groups for each meaningful product category or service line with dedicated assets and search themes.

Mistake 3: No video assets. Without video creative, PMax cannot access YouTube, Discovery, or Gmail inventory — typically 20–35% of Google's available audience reach. Fix: Produce simple, direct video assets (even 15-second product demos or testimonial clips) and upload them to each asset group. This unlocks the full cross-channel reach PMax is designed to leverage.

Mistake 4: Daily optimisation changes. Every significant change resets the learning period. Advertisers making weekly creative swaps, bid adjustments, and audience changes keep their campaigns in perpetual learning, never reaching the stable performance state where real optimisation is possible. Fix: Establish a monthly optimisation cadence for structural changes, with weekly-only reporting reviews.

Mistake 5: Setting tROAS before sufficient data. Applying a Target ROAS target before the campaign has 30+ monthly conversions forces the algorithm into a constrained state where it reduces delivery to protect ROAS rather than accumulating conversion data. Fix: Start with Maximise Conversion Value, graduate to tROAS at 30+ conversions/month.

Mistake 6: URL expansion left uncontrolled. URL expansion will send traffic to your blog posts, careers pages, and other non-converting pages if not restricted. Fix: Add URL exclusion lists blocking non-commercial pages, or disable URL expansion for campaigns where landing page specificity is critical.

For a broader perspective on how Performance Max fits into your overall campaign portfolio, our Google Ads for Business Growth article covers the full paid search strategy framework. If you're managing ecommerce campaigns specifically, see our Google Ads for Ecommerce Strategy Guide. For professional services accounts where PMax use cases are more limited, our Google Ads for Professional Services guide covers the relevant campaign architecture.

Performance Max for Different Business Types

Ecommerce

PMax is most mature and most effective for ecommerce. The product feed provides structured data that helps the algorithm understand what you're selling, who it appeals to, and how to price it against competing options in Shopping auctions. The recommended ecommerce architecture: one PMax campaign per product margin tier, with asset groups organised by category, feed-only configuration tested as an alternative for high-volume catalog campaigns, and Standard Shopping running in parallel for price-sensitive high-conversion products where conversion rate stability matters more than incremental reach.

Ecommerce PMax accounts should target a conversion rate of 1.5–2.5% as a baseline benchmark, with healthy accounts achieving 2–4% on well-optimised Shopping placements. The PMax median ROAS of 2.57x across all verticals includes many poorly configured accounts — well-governed ecommerce PMax accounts with brand exclusions and segmented asset groups regularly achieve 3–5x ROAS on non-branded traffic.

Lead Generation

PMax for lead generation is more complex than for ecommerce because there's no product feed to anchor the algorithm's understanding of what you're selling. Audience signals, search themes, and asset group specificity all become more important as substitutes for feed data. The critical additional requirement is offline conversion tracking: if PMax is optimising for form fills and your form fills include a significant proportion of low-quality or unqualified leads, the algorithm will optimise for quantity rather than quality. Import qualified lead events from your CRM (MQL, SQL, Opportunity) and assign conversion values that reflect their relative quality.

For B2B lead generation with high-value deals, our recommendation remains Search campaigns as the primary conversion driver, with PMax in a supporting discovery role guided by strong audience signals and restricted URL expansion. See our Google Ads for B2B SaaS guide for the full B2B paid search architecture.

Local Services

Performance Max Local is a variant designed for location-based businesses. It emphasises Maps, local Search, and proximity-targeted Display and YouTube inventory. For local service businesses already running Local Services Ads, PMax Local fills the awareness and consideration gaps that LSAs don't address — building brand familiarity in the local market and reaching potential customers before they're actively searching. The campaign structure should be geographically restricted (service area only, not national targeting) with call and location extensions active to maximise conversion action diversity.

The Role of AI and Automation in PMax's Future

Google's 2025 roadmap signalled continued expansion of Performance Max's capabilities and transparency. Enhanced search term reporting with more granular query visibility, theme performance scoring showing individual theme effectiveness ratings, and AI-suggested themes based on your assets are all expected features in 2026. The trajectory is clear: PMax will become more transparent, more controllable, and more central to Google Ads strategy over the next 12–24 months.

The implication for advertisers is strategic: developing deep expertise in PMax optimisation now is an investment with compounding returns. The advertisers who understand how to govern the algorithm — asset group architecture, audience signal quality, brand protection, conversion data quality — will have a structural advantage as PMax takes a larger share of Google's ad inventory and conversion volume.

Understanding PMax doesn't exist in isolation from your overall digital marketing strategy. The campaign optimisation principles that apply to PMax — hypothesis-driven changes, statistical significance in testing, cross-channel attribution clarity — apply across your entire marketing mix. Our Campaign Optimisation Guide covers the systematic approach to improving performance across all channel types.

The data quality that drives PMax performance — clean CRM data, offline conversion tracking, accurate attribution — is the same foundation that supports effective reporting across your whole business. Our guide to Google Ads for B2B SaaS covers offline conversion tracking in detail for high-value, long-cycle accounts.

Getting Started: The 90-Day PMax Optimisation Sprint

If your current PMax campaigns are underperforming or you're setting up PMax for the first time, here's the 90-day roadmap:

Days 1–7 — Governance and Structure: Implement brand exclusions. Add URL expansion controls. Set up account-level negative keyword lists. Review existing asset groups and separate any single-group-for-everything setups into thematic segments. Upload Customer Match lists from CRM.

Days 8–21 — Asset Buildout: Audit all asset ratings and replace Low-rated assets. Build video assets for any asset groups that don't have them (even simple, direct 15-second clips qualify). Write new headline variants covering different value proposition angles. Ensure images are present in all aspect ratios.

Days 22–42 — Learning Phase: Launch restructured campaigns with Maximise Conversion Value (no tROAS). Set realistic budgets for the learning period. Review reporting weekly but make no structural changes. Monitor Search campaign performance for signs of cannibalisation.

Days 43–70 — First Optimisation Cycle: Review channel breakdown per asset group. Add new search themes based on Search Term Insights. Replace underperforming assets. If conversion volume is at 30+/month, graduate to tROAS set 15% below actual ROAS.

Days 71–90 — Measurement and Scaling: Run cross-campaign analysis to validate PMax's incremental contribution. Check CRM data for PMax lead quality vs Search lead quality. Adjust ROAS targets based on real performance data. Plan for next asset group expansion or new campaign segmentation.

Ready to see exactly how your current Google Ads campaigns are performing? The Campaign Optimiser benchmarks your account against industry standards, identifies the highest-leverage improvements, and gives you a prioritised action plan — across PMax, Search, and every other campaign type you're running. Run your free Campaign Optimiser assessment with Involve Digital.

Get Started Using The Form Below

Performance Max is one piece of a larger paid search strategy. For the complete picture of how PMax fits alongside Search, Demand Gen, and other Google campaign types, see our Digital Marketing Strategy Pillar Guide. If you're running Performance Max for an ecommerce store, our Google Ads for Ecommerce Strategy Guide covers the full Shopping campaign architecture. For broader campaign health and systematic optimisation methodology, our Campaign Optimisation Guide provides the diagnostic framework that applies across all campaign types.

‍