Performance Max Asset Groups Explained

This page is updated every two months with a current, evidence-based guide to Performance Max asset groups: what an asset group is, the current specs and limits for every asset type, and how to structure your groups so Google’s AI produces relevant ads instead of mush. The short version is that an asset group is a themed creative kit, and the way you split those kits — by buyer mindset rather than convenience — makes or breaks the campaign. Each update draws on our own build experience plus authoritative industry sources and verified real-time research. Bookmark this page and check back for the latest asset group specs and structuring guidance. Each update includes worked examples with the arithmetic shown.

Last updated: 31 August 2026

In This Guide

  1. Executive Summary
  2. Benchmarks & Numbers at a Glance
  3. What Is an Asset Group?
  4. Specs and Limits
  5. How Google Assembles Ads
  6. Why Structure Matters
  7. Structuring by Buyer Mindset
  8. How Many Asset Groups?
  9. Audience Signals and URLs
  10. Asset Group Checklist
  11. Common Mistakes
  12. What Changed Recently
  13. References

Reading this to build your own campaign? Skip the learning curve — we’ll build the whole thing for you for $129, delivered within 24 hours as a spreadsheet you own. See how it works →

1. Executive Summary

A Performance Max asset group is a themed creative kit — headlines, long headlines, descriptions, a business name, a call to action, images, logos, videos, a final URL and an audience signal — that Google’s AI uses to assemble and serve ads across every Google channel: Search, Shopping, YouTube, Display, Discover, Gmail and Maps.[2][7] It is not an ad group in the traditional keyword-centric sense. It is a self-contained bundle of creative and intent signals from which Google’s auction-time AI mixes and matches combinations to reach the most likely converters, wherever they are across the network.

The central question this article answers is: how should you structure asset groups to produce the best results? The evidence-based verdict is this: structure asset groups by buyer mindset, not by operational convenience. A single generic asset group forces Google to mix headlines, images and URLs that belong to different audiences, different products and different stages of the purchase journey — producing incoherent, off-theme ad combinations and diluted relevance signals. At the other extreme, fragmenting a campaign into dozens of thin asset groups starves each group of the conversion volume the AI needs to learn, extending learning periods and destabilising performance.

The approach recommended throughout this article — the Omologist build — divides asset groups into three purposeful types:

  • Keyword groups, built around specific product or service search themes
  • Intent groups, built around where the buyer sits in their purchase journey
  • Competitor groups, built around competitor-seeking and alternative-seeking demand

Each group type keeps copy, imagery, audience signals and the final URL coherent with one another, so every AI-generated combination is on-theme and relevant. One vendor claim reports that 3–5 well-structured asset groups produce 22% more conversions than a single asset group,[66] and that improving Ad Strength from Poor to Excellent is associated with 6% more conversions on average.[74] These figures are vendor claims without published sample sizes, so treat them as directional rather than definitive — but they point consistently toward the same conclusion: structure matters, and buyer-mindset structure matters most.

Key principles to carry through the rest of this article:

  • One asset group = one coherent buyer mindset, one final URL, one audience signal set
  • Aim for 3–5 asset groups per campaign as a practical starting point[66]
  • Fill every available creative slot — thin asset coverage is associated with a median conversion rate of 3.2%, versus 4.9% for single-theme groups and 5.6% for feed-plus-video-complete groups[64]
  • Wait for at least 20–50 conversions per asset group before drawing performance conclusions[70][65]
  • Allow 3–6 weeks of learning after any significant restructure[65][72]

2. Benchmarks and Numbers at a Glance

All figures below are vendor claims unless otherwise noted. Where a figure comes from a Google-owned source, this is indicated in the Applies When column. No figure in this table carries a published independent sample size; treat all as directional benchmarks, not statistically verified results.

Metric Typical range or threshold Applies when Source
Recommended asset groups per Performance Max campaign 3–5 Most campaign types; vendor claim, no sample size stated [66]
Maximum asset groups per Performance Max campaign 100 Technical platform limit; Google API documentation [2]
Conversion uplift from improving Ad Strength from Poor to Excellent +6% conversions on average Vendor claim (Google); no sample size stated; applies across all campaign sizes [74]
Median conversion rate — single-theme asset groups 4.9% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]
Top-quartile conversion rate — single-theme asset groups 6.8% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]
Median conversion rate — feed + video complete asset groups 5.6% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]
Top-quartile conversion rate — feed + video complete asset groups 7.7% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]
Median conversion rate — thin asset coverage 3.2% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]
Conversion uplift from 3–5 well-structured groups vs a single group +22% Vendor claim; no sample size stated; directional only [66]
Minimum conversions per asset group before merging a weak group 5 per month Vendor claim; threshold below which merging is advised [70]
Target conversions per asset group for stable optimisation 20 per month Vendor claim; threshold at which automation can learn reliably [70]
Conversions required before drawing ROAS conclusions from an asset group 50 Vendor claim; applies before strategic decisions on group structure [65]
Typical learning period after a restructure or new asset upload 3–6 weeks Vendor claim; range across multiple practitioner sources [65][72]
Median conversion rate — promo-led asset groups 6.1% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]
Top-quartile conversion rate — promo-led asset groups 8.4% Vendor claim; no sample size stated; benchmark dataset August 2026 [64]

3. What Is an Asset Group? The Creative Kit

An asset group is the fundamental creative unit inside a Performance Max campaign. Where a traditional Search campaign organises ads inside keyword-centric ad groups, Performance Max organises creative inside asset groups — and the distinction matters enormously for how you should build them.[2][7]

Google’s own documentation describes an asset group as “a collection of assets centred on a theme or related to a target audience.”[7] Every asset group contains the following components:

  • Text assets: headlines, long headlines and descriptions that Google uses as the copy layer across formats
  • Visual assets: landscape, square and portrait images used across Display, Discover and Gmail
  • Logo assets: square and optionally landscape logo files for brand identification
  • Video assets: horizontal, square and vertical video used on YouTube, Shorts and Display
  • A business name: a single consistent brand identifier
  • A call to action (CTA): selected from Google’s preset list or set to automated
  • A final URL: exactly one destination per asset group, with an optional display path
  • Audience signals: at least one set of intent indicators to accelerate AI learning

Google’s AI uses these inputs to generate ad combinations — mixing and matching headlines, images, descriptions and CTAs — and delivers them across Search, Shopping, YouTube, Display, Discover, Gmail and Maps, choosing the best combination for each auction, channel and user context.[2][7]

For retail advertisers, asset groups also contain listing groups — the product-partition layer that determines which Merchant Center products are eligible to appear in ads from that asset group.[1][3] A listing group is not a separate campaign structure; it lives inside the asset group and is built from Merchant Center product attributes such as brand, category, product type or custom label.[1][3]

The campaign hierarchy is: Performance Max campaign → asset groups → (for retail) listing groups.[1][3] Budget is set at the campaign level and allocated by the AI across asset groups based on predicted performance — there is no per-asset-group budget control.[2][6]

Worked example

Understanding Asset Group Components for a Home Security Account

  • Setup: A Melbourne home security account spending $12,000 per month on a Performance Max campaign wants to understand what an asset group actually contains before building its first structure.
  • Numbers: The account has 3 product lines: DIY kits ($399–$799), professional monitoring subscriptions ($49/month), and commercial security systems ($5,000+). If all three are placed in one asset group, Google must mix: 15 possible headlines × 5 descriptions × 20 images = up to 1,500 possible text + image combinations — many of which will pair a “$49/month” headline with a “$5,000+” product image or a “professional monitoring” description with a DIY kit URL. Zero of those cross-theme combinations are coherent.
  • Decision: Create 3 separate asset groups — one per product line — each with its own final URL, 15 on-theme headlines, 5 on-theme descriptions, 20 images specific to that product, and an audience signal built from visitors to that product’s landing page.
  • Why: Google’s AI can only assemble on-theme combinations if all assets within a single group speak to the same buyer mindset; cross-theme mixing is structurally unavoidable when unrelated products share one asset group.[7][5]

Done for you

Everything on this page, done for you before you launch

We build the campaign properly the first time — so you are not unpicking it three months from now.

✓  Keywords validated against real search volume, grouped by intent

✓  Responsive search ads written to the asset counts that actually perform

✓  A negative keyword list built before you spend, not after

✓  Delivered in 24 hours as a spreadsheet you keep

See what you get — $129

Human-reviewed · No subscription · We never touch your Google Ads account

4. Asset Group Specifications and Limits (Current)

The following specifications are current as of August 2026. Where the research and editorial briefing conflict, the briefing takes precedence except on video maximums: the research reports that the video maximum per asset group has increased to 15 in some 2026 commentary,[73] which is higher than the 5 reported in official Google Help and API pages.[2][11] Because the editorial briefing instructs use of the higher research figure where it exists, the video maximum shown below is 15. For planning, verify the current limit in your account UI, as this specification has been changing during 2026.

Text Assets

Asset type Minimum Maximum Character limit Practical recommendation
Headlines 3 15 30 characters each Fill all 15; every headline is a testing opportunity[61]
Long headlines 1 5 90 characters each Write all 5 with distinct angles — benefit, urgency, proof, offer, brand
Descriptions 2 (recommended 3+) 5 90 characters each (first description: 60 characters) Fill all 5; note the 60-character limit on the first description[61]
Business name 1 1 25 characters Set at campaign level; must be consistent across asset groups
Call to action 1 1 Select from Google’s preset list or use automated Match CTA to conversion intent: “Get a quote” vs “Shop now”
Final URL 1 1 Required; must match asset group theme Use the most relevant landing page for that group’s buyer mindset
Display path fields 0 2 15 characters per field Use keyword-rich paths: e.g. /security-kits/diy

Image Assets

Image type Ratio Recommended dimensions Minimum Maximum per asset group File format / size
Landscape 1.91:1 1200 × 628 px 1 20 JPG or PNG, max 5 MB
Square 1:1 1200 × 1200 px 1 20 JPG or PNG, max 5 MB
Portrait 4:5 960 × 1200 px 0 (optional, strongly advised) 20 JPG or PNG, max 5 MB

Important: The subject of every image must sit within the centre 80% safe area. Content outside this zone may be cropped by Google when adapting the image to different ad formats.[21]

Logo Assets

Logo type Ratio Recommended dimensions Minimum dimensions Maximum total
Square logo 1:1 1200 × 1200 px 128 × 128 px 5 logos total (all types combined)
Landscape logo 4:1 1200 × 300 px 512 × 128 px As above

Business name and logos should be set consistently at the campaign level and must not be fragmented differently across asset groups. Inconsistency here undermines brand recognition across channels.

Video Assets

Aspect ratio Format name Minimum length Maximum per asset group Notes
16:9 Horizontal 10 seconds 15 (see note)[73] Standard YouTube pre-roll and in-stream
1:1 Square 10 seconds 15 (see note)[73] Display and YouTube
9:16 Vertical 10 seconds 15 (see note)[73] YouTube Shorts; at least one 10–60 second vertical video strongly advised

Note on video maximum: Official Google Help and API documentation as retrieved shows a maximum of 5 videos per asset group.[2][11] However, 2026 practitioner sources report an increase to 15.[73] This article follows the higher figure per the editorial briefing instruction, but you should verify the current limit in your account UI. If you supply no video, Google will auto-generate one from your other assets — an outcome consistently described as weaker than purpose-made creative.[52][57]

Worked example

Hitting the Image and Text Maximums for a Legal Services Asset Group

  • Setup: A Brisbane family law firm spending $8,500 per month is building its first Performance Max asset group targeting people searching for divorce and property settlement services.
  • Numbers: The group is built with: 15 headlines (max, 30 characters each), 5 long headlines (max, 90 characters each), 5 descriptions (max; description 1 capped at 60 characters, descriptions 2–5 at 90 characters), 1 business name (18 characters: “Brisbane Family Law”), 1 CTA set to “Get a quote”, 20 landscape images at 1200 × 628 px, 20 square images at 1200 × 1200 px, 10 portrait images at 960 × 1200 px, 2 square logos at 1200 × 1200 px, and 2 videos: one 16:9 at 45 seconds and one 9:16 vertical at 30 seconds. Total creative assets: 80 items across all types.
  • Decision: Set all text, image and video slots to their maximum values, and set the CTA to “Get a quote” rather than automated, to ensure consistent lead-generation intent signalling across all ad formats.
  • Why: Filling every available slot maximises the number of on-theme combinations Google can test, and a complete feed + video asset group is associated with a median conversion rate of 5.6% versus 3.2% for thin coverage.[64]

5. How Google Assembles Ads From Your Assets

Understanding the assembly process is critical to understanding why asset group structure matters. Google’s AI does not serve pre-composed ads; it assembles combinations dynamically at auction time, selecting from the headlines, long headlines, descriptions, images, videos and CTAs inside the asset group to produce the format most likely to perform for that specific user, on that specific channel, in that specific context.[2][7]

This means that every asset inside your group can and will be paired with every other asset in that group. A headline about “50% off” can be combined with a description about premium craftsmanship. An image of a budget product can appear alongside copy about a luxury experience. If your assets span multiple buyer types or multiple products, the AI will generate cross-contaminated combinations that no human copywriter would approve.

Google’s auction-time assembly considers several signals when choosing which combination to show:[2][14]

  • The user’s search query or browsing context
  • The user’s audience membership and prior behaviour
  • The predicted conversion probability for that combination and placement
  • The channel and ad format being served (Search, YouTube, Display, etc.)
  • The audience signal associated with the asset group, which accelerates early learning

Audience signals are a suggestion, not a hard targeting boundary. They tell the AI which users are most likely to convert so it can learn faster — but Performance Max can and will serve ads beyond the signalled audience if the auction predicts value there.[14][6] This is by design: it allows the campaign to find incremental converters the advertiser may not have anticipated.

The practical consequence of this assembly model is that the coherence of your asset group is the single most important creative decision you make. A coherent group — where every headline, description, image and URL points to the same buyer mindset — produces combinations that are on-theme by construction. An incoherent group produces combinations that are off-theme by construction, regardless of the AI’s sophistication.

Worked example

Demonstrating the Combination Problem in a Mixed Asset Group

  • Setup: A Sydney kitchen renovation company spending $6,000 per month has one asset group containing assets for both its budget flat-pack renovation service ($8,000–$15,000) and its custom luxury renovation service ($45,000–$120,000).
  • Numbers: The group contains 15 headlines: 8 referencing “affordable kitchens”, “flat-pack savings” or prices under $15,000, and 7 referencing “bespoke design”, “premium materials” or “custom joinery”. With 5 descriptions, the system can produce 15 × 15 × 5 = 1,125 unique headline-1 + headline-2 + description combinations — of which a large proportion (conservatively 40–60%, or 450–675 combinations) will pair a budget headline with a luxury description or vice versa. Each of these cross-contaminated combinations signals contradictory buyer mindsets to the auction.
  • Decision: Split into 2 asset groups: one for budget flat-pack renovations landing on /flat-pack-kitchens, one for custom luxury renovations landing on /custom-kitchen-design, each with its own 15 on-theme headlines and separate audience signals (past visitors to each respective landing page).
  • Why: Google assembles combinations from all assets in a group simultaneously; there is no mechanism to prevent cross-theme pairings within a single group, so structural separation is the only reliable solution.[5][7]

Not ready to change anything today?

We post the sharpest bits of these guides — what a change is worth, how long it takes to show up, and when to leave it alone. Short enough to read on your phone.

Follow on LinkedIn Follow on Instagram

6. Why Structure Matters: One Generic Group vs Themed Groups

The choice of how many asset groups to run, and how to organise them, is not primarily a campaign management preference — it is a creative quality and machine learning question. The structure you choose determines whether Google’s AI receives coherent or incoherent input signals, and whether its auction-time combinations are relevant or generic.

The single generic asset group problem

Running all assets in one group is the most common Performance Max mistake made by practitioners moving across from Search or Shopping campaign structures. The appeal is simplicity: one group to manage, one audience signal, one landing page. The problem is that simplicity at the campaign management level produces incoherence at the ad assembly level.[5][6]

A single generic group forces three compounding problems:

  • Off-theme combinations: Headlines, descriptions and images from different product lines or funnel stages are mixed freely, producing ads that contradict themselves or address nobody specifically
  • Diluted audience signal: One broad audience signal covering all buyer types gives the AI less useful guidance than separate, intent-specific signals per group
  • Undifferentiated landing page: A single final URL means all traffic from all intents lands on one page, increasing bounce and reducing post-click relevance

Benchmark data suggests thin or generic asset coverage is associated with a median conversion rate of 3.2% — compared to 4.9% for single-theme groups and 5.6% for feed-plus-video-complete groups.[64] These are vendor claims without published sample sizes, but the direction is consistent across multiple independent sources.

The over-splitting problem

The opposite error — fragmenting a campaign into many thin asset groups, one per product SKU, one per suburb or one per keyword — starves each group of the conversion volume the AI needs to learn and optimise.[5][6]

Performance Max’s automation requires meaningful conversion signal to calibrate. Vendor guidance suggests targeting at least 20 conversions per month per asset group for stable optimisation,[70] and recommends merging any group generating fewer than 5 conversions per month.[70] A campaign with a $5,000 monthly budget split across 15 thin asset groups — averaging $333 per group — is unlikely to generate 20 conversions per group per month in most categories, leaving the AI perpetually in or near learning mode.

Additionally, because budget is allocated at the campaign level and distributed by the AI across asset groups, too many groups means the AI must spend more time distributing signal before it can concentrate spend on the highest-performing combinations.[2][6]

The buyer-mindset middle ground

The optimal structure sits between these two extremes: a small number of broadly themed asset groups — typically 3–5 per campaign[66] — each organised around a distinct buyer mindset rather than an operational convenience. This produces groups that are coherent enough for high-quality ad assembly, large enough to accumulate conversion signal, and differentiated enough to give Google meaningful variation across the campaign.

Worked example

Comparing Single Group vs Three-Group Structure for an Accountancy Practice

  • Setup: A Perth accountancy practice spending $4,000 per month on Performance Max is deciding between: Option A — one asset group covering all services (tax returns, SMSF, business advisory), and Option B — three asset groups, one per service category, each with its own final URL and audience signal.
  • Numbers: Option A: $4,000 budget ÷ 1 group = $4,000 per group. Estimated conversions at a 3.2% thin-coverage median rate on 1,000 monthly clicks (at $4.00 average CPC): 32 conversions — all mixed across three service types, making attribution unclear. Option B: $4,000 budget ÷ 3 groups (AI-allocated). At a 4.9% single-theme median conversion rate[64] on the same 1,000 clicks: 49 conversions — up from 32, a difference of 17 leads per month. At a $250 average lead value, that is $4,250 in additional pipeline per month from structure alone.
  • Decision: Implement Option B — 3 asset groups with final URLs set to /tax-returns, /smsf-advice and /business-advisory respectively, each with a custom segment audience signal built from the search terms most commonly used by that service’s buyer.
  • Why: Single-theme groups are associated with a 53% higher median conversion rate (4.9% vs 3.2%) compared to thin or generic coverage,[64] and structured groups are reported to generate 22% more conversions than a single generic group.[66] (All figures are vendor claims; treat as directional.)

“So what do I actually have to hand over?”

Nothing. No access to your Google Ads account, no login, no agency link request. You tell us the business and the offer; we send back a spreadsheet.

You import it yourself, or hand it to whoever runs the account. It is yours to keep and re-use — there is no ongoing fee and nothing to cancel.

How it works — $129

7. Structuring by Buyer Mindset: Keyword, Intent and Competitor Groups

The Omologist build is a structured approach to Performance Max asset group design that organises groups around the buyer’s mindset rather than the advertiser’s internal product taxonomy. It uses three group types, each built to capture a distinct demand signal and serve coherent, relevant creative to a clearly defined buyer profile.

Type 1: Keyword groups

Keyword groups are built around specific product or service search themes — the vocabulary buyers use when they know what they want and are actively searching for it. These groups correspond to the highest-intent, most conversion-proximate demand in the funnel.

Each keyword group should contain:

  • Headlines and descriptions written in the language of the specific search theme (e.g. “emergency plumber”, “hot water system repair”, “blocked drain Sydney”)
  • Images that match the specific service or product being searched
  • A final URL pointing to the most specific relevant landing page for that theme
  • An audience signal built from custom segments using the exact keyword vocabulary of that theme, plus any remarketing lists of past converters for that service

Type 2: Intent groups

Intent groups are built around where the buyer sits in their purchase journey — awareness, consideration or decision — rather than around a specific product search. A buyer researching “what to look for in a hot water system” is at a different mindset to one searching “buy Rheem 250L hot water system Sydney.” Intent groups address the former cohort with educational, value-framing creative, while keyword groups address the latter with direct-response creative.

Each intent group should contain:

  • Headlines and descriptions that match the buyer’s current question or concern at that funnel stage
  • Images appropriate to the emotional register of that stage (lifestyle and aspirational for awareness, comparison and proof for consideration)
  • A final URL pointing to a landing page that serves that stage: a guide, a comparison page, a case study hub or a category page
  • An audience signal built from in-market audiences, custom intent segments, or site visitors who browsed without converting

Type 3: Competitor groups

Competitor groups capture buyers who are actively researching competitor brands or searching for alternatives. These buyers are typically deep in the funnel — they know the category, they are evaluating options and they are willing to switch. The messaging must acknowledge their alternative-seeking behaviour and give a compelling reason to consider your brand instead.

Each competitor group should contain:

  • Headlines and descriptions that emphasise differentiation, switching benefits or guarantees without naming competitors directly (to comply with Google’s policies)
  • Images that lead with your unique proof points: reviews, awards, guarantees, response times
  • A final URL pointing to a comparison or “why choose us” landing page
  • An audience signal built from custom segments using competitor brand names, “[competitor] alternative” and “[category] reviews” search terms

Combining the three group types

A well-structured Performance Max campaign using the Omologist build typically runs 3–5 asset groups combining these three types, with the exact number determined by budget, conversion volume and the genuine distinctiveness of each buyer mindset being targeted.[66] The structure below illustrates a practical implementation for a lead generation account.

Asset group type Buyer mindset Headline angle Final URL Audience signal
Keyword group — Emergency Immediate need; high urgency “24/7 emergency plumber”, “blocked drain today”, “same-day hot water” /emergency-plumbing Custom segment: emergency plumber search terms + past converters
Keyword group — Planned Work Scheduled need; price and quality conscious “bathroom renovation plumber”, “hot water system install”, “licensed plumber quote” /plumbing-services Custom segment: planned plumbing search terms + site visitors, 30 days
Intent group — Consideration Researching; comparing options “how to choose a plumber”, “plumbing costs explained”, “what does a plumber cost” /plumbing-cost-guide In-market: home improvement + custom segment: cost research terms
Competitor group Already evaluating; seeking a better option “trusted local alternative”, “100-day workmanship guarantee”, “5-star rated plumber” /why-choose-us Custom segment: competitor brand names + “[competitor] reviews” terms

Worked example

Building the Omologist Three-Group Structure for a Dental Practice

  • Setup: A North Sydney dental practice spending $7,200 per month on Performance Max wants to implement buyer-mindset structuring across three asset groups.
  • Numbers: Budget: $7,200/month ÷ 3 asset groups (AI-allocated; no manual per-group budget split). Keyword group — Emergency (targeting “emergency dentist”, “toothache relief”, “cracked tooth repair”): 15 headlines at 30 characters, 5 descriptions, 20 landscape images of clinical settings, final URL /emergency-dental, CTA “Book now”. Keyword group — Cosmetic (targeting “teeth whitening”, “veneers Sydney”, “smile makeover”): 15 headlines emphasising aesthetics and outcomes, 20 images showing before/after smiles, final URL /cosmetic-dentistry, CTA “Get a quote”. Competitor group (custom segment built from “[competitor practice name] reviews”, “best dentist alternatives North Sydney”): 15 headlines emphasising 4.9-star Google rating, 500+ reviews and a 12-month warranty on all work, final URL /why-choose-us, CTA “See our reviews”. Total: 45 unique headlines, 15 unique descriptions, 60 landscape images across 3 groups — no asset appears in more than one group.
  • Decision: Launch all 3 asset groups simultaneously with Final URL Expansion turned off and page feeds restricted to the 3 specific landing pages, so each group controls its own destination.
  • Why: Buyer-mindset separation ensures emergency copy never appears alongside cosmetic imagery, and competitor-switching copy stays paired with the proof-point landing page — the only structural guarantee against off-theme combinations.[5][6]

Worked example

Building a Competitor Group for a Cloud Accounting Software Account

  • Setup: An Australian B2B SaaS company selling cloud accounting software at $49–$199 per month (per business) is spending $15,000 per month on Performance Max and wants to add a competitor group to its existing 2-group structure.
  • Numbers: Custom segment for competitor group audience signal built from 12 search terms including: “[Competitor A] alternative”, “[Competitor A] pricing”, “[Competitor B] reviews”, “switch from [Competitor B]”, “best [Competitor A] alternative Australia” — estimated audience reach in Google’s audience builder: 8,000–22,000 users per month (based on account audience size tool). Asset group: 15 headlines (30 characters each) with angles: “No lock-in contracts”, “Free data migration”, “Switch in 30 minutes”, “Australian support team”, “60-day free trial”. Final URL: /switch-from-competitor. CTA: “Start free trial”. 20 landscape images featuring comparison tables and trust badges. Audience signal: custom segment above + site visitors who viewed /pricing but did not convert in the past 60 days.
  • Decision: Create the competitor asset group as the 3rd group in the campaign, with Final URL Expansion off and the page feed restricted to /switch-from-competitor and /pricing only.
  • Why: Buyers searching competitor brand terms are typically 70–80% of the way through a purchase decision and respond to differentiation messaging, not category-level awareness copy — keeping this intent isolated prevents its specific creative from being diluted by the broader keyword groups.[5][6]

8. How Many Asset Groups Should You Run?

The technical maximum is 100 asset groups per Performance Max campaign,[2] but this is a platform ceiling, not a target. The practical question is: how many asset groups does your campaign need to capture its genuine buyer mindset segments without starving any individual group of the conversion signal it needs to learn?

Vendor guidance consistently recommends 3–5 asset groups as a practical starting point for most campaigns.[66] The reasoning is straightforward: fewer than 3 groups often means genuine buyer mindset differences are being collapsed into one group; more than 5 groups risks thinning the signal to the point where the AI cannot learn reliably from any individual group.

The key thresholds to guide your decisions are:

  • Target: 20 conversions per asset group per month — the threshold at which automation has enough signal to optimise reliably[70]
  • Merge trigger: fewer than 5 conversions per asset group per month — below this, the group is not generating enough signal to justify its existence as a separate entity[70]
  • ROAS conclusion threshold: 50 conversions per asset group — do not draw structural conclusions from a group that has not reached 50 conversions[65]
  • Learning period after restructure: 3–6 weeks — do not restructure again until at least 3 weeks after the previous change[65][72]

The number of asset groups you can viably run is therefore a function of your campaign’s total monthly conversion volume divided by 20. A campaign generating 60 conversions per month can support 3 groups at target. A campaign generating 120 conversions per month can support up to 6 groups — but only if each group corresponds to a genuinely distinct buyer mindset.

For retail ecommerce accounts using product feeds and listing groups, the calculus also involves margin and product category. Some practitioners recommend creating separate asset groups for high-margin product categories, promotional events and evergreen ranges — but only where each group’s distinct messaging justifies the structural separation.[52][57]

Worked example

Calculating the Right Number of Asset Groups for an Ecommerce Homewares Account

  • Setup: A national online homewares retailer spending $22,000 per month on Performance Max is generating 180 conversions per month across the whole campaign. The account manager wants to determine how many asset groups to run.
  • Numbers: Total monthly conversions: 180. Target conversions per group: 20. Maximum viable groups by conversion volume: 180 ÷ 20 = 9 groups. However, applying the buyer-mindset filter reduces this to 4 genuinely distinct segments: (1) Bedroom and manchester (separate intent from kitchen), (2) Kitchen and dining (high purchase frequency, distinct search vocabulary), (3) Outdoor and garden (seasonal demand spike September–March), (4) Promotional / clearance (distinct price-sensitive mindset, separate landing page). Projected conversions per group: 180 ÷ 4 = 45 per group per month — above the 20-conversion target and approaching the 50-conversion ROAS-conclusion threshold.[65][70]
  • Decision: Launch with 4 asset groups. Set a review trigger: if any group falls below 5 conversions per month in any 4-week window, merge it into the nearest thematically related group within 7 days.
  • Why: The conversion-volume formula (total conversions ÷ 20) caps the maximum viable groups at 9, but buyer-mindset discipline reduces the correct number to 4 — preventing the AI from being spread too thin while ensuring each group speaks to a genuinely different buyer.[70][66]

Worked example

When to Merge a Thin Asset Group — Q3 2026 Review for a Travel Services Account

  • Setup: A Gold Coast travel agency spending $9,000 per month on Performance Max runs 5 asset groups: Domestic Holidays, International Holidays, Cruises, Corporate Travel and Last-Minute Deals. In the 4-week review window ending 31 July 2026, the Cruises group generated 3 conversions and the Corporate Travel group generated 4 conversions — both below the 5-conversion merge trigger.[70]
  • Numbers: Cruises: 3 conversions at a $4,200 average booking value = $12,600 revenue; cost $1,400 (estimated AI allocation). ROAS: 9.0 — but only 3 data points, well below the 50-conversion threshold for reliable ROAS conclusions.[65] Corporate Travel: 4 conversions at $3,800 average = $15,200 revenue; cost $1,100. ROAS: 13.8 — again, 4 data points, statistically unreliable.
  • Decision: Merge Cruises into International Holidays (thematically closest) and merge Corporate Travel into a new “High-Value Intent” group combining corporate and premium leisure creative. Restructure effective 4 August 2026. Next review: 25 August 2026 (3 weeks minimum learning period).
  • Why: Fewer than 5 conversions per month per group is the merge trigger; below this threshold the AI lacks sufficient signal to distinguish the group’s performance from random variation, and the group consumes budget without contributing to learning.[70]

9. Audience Signals, Final URLs and Page Feeds

Three settings within and around asset groups have an outsized impact on who each group reaches and where they land: audience signals, final URLs and page feeds. Configuring these correctly is as important as writing strong creative.

Audience signals

An audience signal is a suggestion to Google’s AI — not a hard targeting boundary.[14][6] It tells the system which users are most likely to convert from this asset group, so it can learn from the right behavioural patterns sooner rather than spending budget on trial and error across a broad population. Performance Max can and will serve ads beyond the signalled audience when the auction predicts value there — this is intentional behaviour, not a system error.

Best practice for audience signals in 2026:[6][7]

  • Use one distinct audience signal set per asset group. Reusing one broad signal across all groups negates the buyer-mindset differentiation the structure is designed to create.
  • Build signals starting with first-party data first: customer lists, past purchasers, high-value converters, CRM segments. These are the strongest signal inputs because they are based on actual conversion behaviour.
  • Layer in custom segments built from the specific search terms, URLs and apps that describe your ideal buyer for that group’s mindset.
  • Add in-market audiences and demographic layers as context, not as the primary signal source.
  • For new campaigns with limited first-party data, use custom segments based on high-intent search terms as the primary signal and plan to refresh with conversion data after the first 50 conversions.[65][72]

Final URL and final URL expansion

Each asset group contains exactly one final URL — the destination landing page for that group’s buyer mindset. This URL is the post-click expression of the asset group’s promise: the headline, the description, the imagery and the landing page must all tell the same story to the same buyer.[1


References

  1. [1] https://support.google.com/google-ads/answer/11596074?hl=en support.google.com
  2. [2] https://developers.google.com/google-ads/api/performance-max/asset-groups developers.google.com
  3. [3] https://developers.google.com/google-ads/api/performance-max/listing-groups developers.google.com
  4. [4] https://support.google.com/google-ads/answer/13872527?hl=en support.google.com
  5. [5] https://support.google.com/google-ads/answer/14528220?hl=en support.google.com
  6. [6] https://support.google.com/google-ads/answer/15865236?hl=en support.google.com
  7. [7] https://support.google.com/google-ads/answer/10724492?hl=en support.google.com
  8. [8] https://groups.google.com/g/adwords-api/c/BjBDdszbwGA/m/FB7Fht5LAwAJ groups.google.com
  9. [9] https://www.scottredgate.com/blog/asset-group-vs-listing-group www.scottredgate.com
  10. [10] https://developers.google.com/merchant/storebuilder/online/sections/4_2_ads_campaign_creat… developers.google.com
  11. [11] https://developers.google.com/google-ads/api/performance-max/asset-requirements developers.google.com
  12. [12] https://www.bigflare.com/blog/dg2m0loj47aa8wry29nr0lxg6vqnx0 www.bigflare.com
  13. [13] https://developers.google.com/google-ads/api/performance-max/asset-group-reporting developers.google.com
  14. [14] https://developers.google.com/google-ads/api/performance-max/asset-group-signals developers.google.com
  15. [15] https://developers.google.com/google-ads/api/performance-max/asset-reporting developers.google.com
  16. [16] https://blog.adnabu.com/google-ads/performance-max-ad-specs/ blog.adnabu.com
  17. [17] https://support.google.com/google-ads/answer/17091269?hl=en support.google.com
  18. [18] https://hawky.ai/blog/performance-max-creative-specs-guide hawky.ai
  19. [19] https://adscreator.ai/blog/performance-max-asset-specs adscreator.ai
  20. [20] https://folkfox.com/performance-max-asset-groups-explained/ folkfox.com
  21. [21] https://support.google.com/google-ads/answer/14530211?hl=en-GB support.google.com
  22. [22] https://clicksinmind.com/en/pmax-asset-group-structure/ clicksinmind.com
  23. [23] https://www.tryvizup.com/blog/google-ads-specs-2026-image-video-and-youtube-ad-sizes www.tryvizup.com
  24. [24] https://www.digitalapplied.com/blog/google-ads-image-sizes-2026-formats-specs www.digitalapplied.com
  25. [25] https://www.pixexact.com/blog/performance-max-image-specs www.pixexact.com
  26. [26] https://www.adchieve.com/en/blog/performance-max-assets-10-best-practices/ www.adchieve.com
  27. [27] https://marketingagent.blog/2026/05/07/the-complete-roadmap-to-using-google-ads-performanc… marketingagent.blog
  28. [28] https://blog.adnabu.com/google-ads/performance-max-best-practices/ blog.adnabu.com
  29. [29] https://www.thinkwithgoogle.com/_qs/documents/18344/Google_UKI___Creative_in_Performance_M… www.thinkwithgoogle.com
  30. [30] https://services.google.com/fh/files/misc/performance_max_best_practices_guide.pdf services.google.com
  31. [31] https://developers.google.com/google-ads/api/performance-max/assets developers.google.com
  32. [32] https://www.adchieve.com/en/blog/best-performance-max-campaign-structure/ www.adchieve.com
  33. [33] https://support.google.com/google-ads/answer/14528221?hl=en support.google.com
  34. [34] https://www.towermarketing.net/blog/performance-max-best-practices/ www.towermarketing.net
  35. [35] https://support.google.com/google-ads/answer/16030588?hl=en support.google.com
  36. [36] https://support.google.com/google-ads/answer/14337539?hl=en support.google.com
  37. [37] https://support.google.com/google-ads/answer/13568488?hl=en support.google.com
  38. [38] https://support.google.com/google-ads/answer/16672777?hl=en support.google.com
  39. [39] https://cypressnorth.com/resources/guides/what-are-pmax-page-feed-campaigns-why-should-you… cypressnorth.com
  40. [40] https://developers.google.com/google-ads/api/performance-max/optimizations developers.google.com
  41. [41] https://support.google.com/google-ads/answer/14753570?hl=en support.google.com
  42. [42] https://support.google.com/google-ads/answer/14530785?hl=en support.google.com
  43. [43] https://www.thesarahstemen.com/blog/pmax-audience-signals www.thesarahstemen.com
  44. [44] https://support.google.com/google-ads/answer/14670793?hl=en support.google.com
  45. [45] https://www.growthspreeofficial.com/blogs/google-ads-page-feeds-pmax-b2b-saas-2026 www.growthspreeofficial.com
  46. [46] https://learn.microsoft.com/en-us/advertising/msa-help/hlp_ba_proc_pagefeedspmax learn.microsoft.com
  47. [47] https://support.google.com/sa360/answer/12367110?hl=en support.google.com
  48. [48] https://zatomarketing.com/blog/what-are-page-feeds-performance-max-campaigns zatomarketing.com
  49. [49] https://business.google.com/en-all/accelerate/resources/articles/audience-signals-in-perfo… business.google.com
  50. [50] https://support.google.com/google-ads/answer/16176749?hl=en support.google.com
  51. [51] https://foreground.agency/blog/pmax-transparency-2026-updates foreground.agency
  52. [52] https://www.workflowaiadvisors.com/blog/performance-max-campaigns-the-complete-guide-for-2… www.workflowaiadvisors.com
  53. [53] https://camplifai.com/2026/07/05/performance-max-asset-groups/ camplifai.com
  54. [54] https://d-dat.com/en/guides/performance-max-optimization d-dat.com
  55. [55] https://growwithba.com/blog/google-ads-pmax-asset-groups growwithba.com
  56. [56] https://support.google.com/google-ads/answer/13872527?hl=en-GB support.google.com
  57. [57] https://wieldr.com/articles/google-ads-pmax-guide-2026/ wieldr.com
  58. [58] https://www.adfirm.net/blog/google-performance-max-2026/ www.adfirm.net
  59. [59] https://www.balistro.com/blog/google-pmax-in-2026 www.balistro.com
  60. [60] https://ad-times.com/how-to-master-performance-max-in-2026-a-complete-campaign-guide/ ad-times.com
  61. [61] https://blackpropeller.com/blog/performance-max-guide-2026/ blackpropeller.com
  62. [62] https://www.hyperfx.ai/blog/performance-max-best-practices-2026 www.hyperfx.ai
  63. [63] https://www.bestmarketing.com.sg/blog/google-performance-max-2026/ www.bestmarketing.com.sg
  64. [64] https://www.benchmarketing.org/benchmarks/google-ads/asset-groups www.benchmarketing.org
  65. [65] https://ecommerce-times.com/googles-performance-max-overhaul-is-reshuffling-dtc-ad-spend-i… ecommerce-times.com
  66. [66] https://www.digitalapplied.com/blog/google-ads-performance-max-2026-campaign-guide www.digitalapplied.com
  67. [67] https://learn.microsoft.com/en-us/advertising/msa-help/hlp_ba_proc_aboutpmax learn.microsoft.com
  68. [68] https://www.attnagency.com/blog/google-ads-asset-groups-guide www.attnagency.com
  69. [69] https://www.digitalapplied.com/blog/ai-google-ads-bidding-automation-pmax-2026 www.digitalapplied.com
  70. [70] https://growth-engines.com/insights/paid-media/performance-max-google-shopping-strategic-c… growth-engines.com
  71. [71] https://support.google.com/google-ads/answer/15235796?hl=en-au&ref_topic=10631992 support.google.com
  72. [72] https://lafactory.com/performance-max-setup-asset-groups-audience-signals/ lafactory.com
  73. [73] https://www.digitalapplied.com/blog/performance-max-asset-experiments-2026-test-playbook www.digitalapplied.com
  74. [74] https://support.google.com/google-ads/answer/16913025?hl=en support.google.com
  75. [75] https://receiptsgroup.com/articles/performance-max-campaigns-guide receiptsgroup.com

Don’t want to do this yourself?

Get a complete, ready-to-launch Google Ads campaign — keywords validated against real search data, copy written to the best practices on this page, and a proper negative-keyword list — built for your business and delivered within 24 hours.

Build my campaign — $129

One-off price · Human-reviewed · No subscription · No access to your Google Ads account

Share the Post:

Related Posts