Site icon Oxedent

Performance Max strategy explained for ecommerce managers

Biplab Poddar
Decorative title card illustration for ecommerce PPC article
Rate this post

A Performance Max (PMax) strategy is a goal-based Google Ads campaign type that uses Google AI and Smart Bidding to serve ads across every Google channel from a single campaign, with the aim of maximising conversions or conversion value at your target cost. Rather than managing separate Search, Shopping, YouTube, Display, Discover, Gmail, and Maps campaigns, you feed one campaign with assets, a product feed from Google Merchant Center, audience signals, and a bidding objective, then let the algorithm allocate budget across inventory in real time. Advertisers who adopt Performance Max see an average increase of 27% more conversions or conversion value at a similar CPA or ROAS, according to Google.

The four components you will act on immediately:

Key takeaways

A Performance Max strategy delivers its best results when conversion tracking, feed quality, and creative assets are all in place before spend is scaled.

Point Details
Inputs drive outcomes Feed quality, conversion values, and audience signals matter more than manual bid adjustments.
Asset counts matter Google recommends 15 headlines, 5 descriptions, 7 images, and at least one uploaded video per asset group.
Learning phase patience Allow 1–6 weeks after launch or major changes; avoid frequent pausing or large budget shifts.
Structure by margin, not channel Segment asset groups by product margin and audience intent; let the algorithm choose the channel.
Oxedent for managed scale Oxedent builds and manages PMax campaigns for ecommerce brands using feed engineering and value-led bidding.

Table of Contents

What is a Performance Max strategy and how does it work?

Performance Max operates on a simple input-to-output logic: you supply conversion goals, creative assets, a product feed, and audience signals; Google’s AI then enters auctions across its entire inventory and allocates impressions to the placements most likely to generate a conversion at your target cost. The quality and breadth of those inputs determine outcomes far more than any manual bid adjustment you could make, because the algorithm has no other levers to pull.

Asset groups are the creative containers at the heart of the campaign. Each group holds a set of headlines, descriptions, images, and videos, plus a set of audience signals. Google’s AI mixes and matches those assets to build ads suited to each placement, from a text ad on Search to a skippable video on YouTube. You can run multiple asset groups within one campaign, which lets you tailor creative and signals to different product categories or audience intents without multiplying campaign overhead.

Performance Max uses asset groups, search themes, and final URL expansion to extend targeting beyond explicit keywords and to combine creatives across Google’s inventory for incremental reach. Search themes are short phrases you provide to indicate the intent you want the campaign to capture; they are not keywords in the traditional sense, but they steer the algorithm toward relevant queries. Final URL expansion allows Google to send traffic to the most relevant page on your site rather than the single URL you specify, which can lift conversion rates but requires careful governance.

Smart Bidding sits beneath all of this. The algorithm reads hundreds of contextual signals at auction time, including device, location, time of day, and browsing behaviour, and adjusts bids in real time. Your choice of bidding objective tells the algorithm what to optimise for. Maximise Conversions pushes volume; Maximise Conversion Value pushes revenue or profit, depending on how you configure your conversion values.

The conceptual flow looks like this:

What are the essential campaign components and controls?

Getting the inputs right is the entire job. The algorithm can only work with what you give it, so a weak feed, misconfigured conversion tracking, or thin creative assets will cap performance regardless of budget.

Essential inputs to configure:

Campaign-level controls to monitor:

Performance Max supports only two Smart Bidding strategies: Maximise Conversions (with optional target CPA) and Maximise Conversion Value (with optional target ROAS); portfolio bid strategies are not supported. In practice, most ecommerce accounts start with Maximise Conversion Value and introduce a tROAS target once the campaign has accumulated enough conversion data to sustain a constraint without throttling delivery.

Pro Tip: Set up conversion value rules before you scale spend. Weighting high-margin product categories or repeat-purchase customers above your blended average gives the algorithm a profit signal rather than a revenue signal, which is the difference between a campaign that grows your top line and one that grows your bottom line.

How to set up a Performance Max campaign: a practical checklist

Accurate conversion tracking is the single most important setup item. Without it, the algorithm is optimising toward a proxy that may not reflect real business value, and every subsequent decision compounds that error.

Setup steps in order:

  1. Confirm conversion tracking fires correctly and records values, not just counts; use ecommerce conversion tracking to verify tag implementation
  2. Connect Google Merchant Center and resolve all feed errors before launch
  3. Link your Business Profile if you have physical locations or use local inventory ads
  4. Build your first asset group with the recommended asset counts (see below)
  5. Upload audience signals: Customer Match list, website remarketing, and at least one custom intent audience
  6. Set your bidding objective; start with Maximise Conversion Value if you have 30+ conversions in the last 30 days
  7. Configure brand exclusions at campaign level to protect branded Search traffic
  8. Review final URL expansion settings and add URL exclusions for pages you do not want traffic sent to (checkout, account login, etc.)
  9. Set a daily budget that allows a reasonable number of conversions per week during the learning phase

Recommended asset counts per asset group:

Google recommends at least 20 text assets (15 headlines, 5 descriptions), at least 7 image assets (3 landscape, 3 square, 1 portrait), and at least one video asset; uploading your own video is advised to avoid lower-quality auto-generated video.

For ecommerce specifically, check that your Merchant Center feed includes GTINs, accurate pricing, availability status, and product-type taxonomy. Feeds with missing attributes or stale pricing are one of the most common causes of poor Shopping performance within PMax.

Audience signals work best when your remarketing lists have at least 100 active users in the last 30 days. If your lists are smaller, upload a Customer Match list from your CRM to supplement them.

Pro Tip: Always upload your own video asset, even a simple 15-second product or brand video. If you do not, Google will auto-generate one from your image assets and landing page, and the quality is rarely good enough for YouTube placements. A weak video asset suppresses YouTube delivery and limits the campaign’s reach across one of Google’s highest-intent channels.

Advanced ecommerce optimisation tactics for Performance Max

The single most effective optimisation for an ecommerce PMax campaign is the combination of a well-engineered product feed, value-based bidding, and diverse creative assets. Each element amplifies the others: a richer feed gives the algorithm better product signals; accurate conversion values give it a profit target; strong creative assets give it more combinations to test across channels.

Step-by-step tactics:

  1. Enrich your feed titles and descriptions. Include brand, material, colour, size, and model number in product titles. The algorithm uses feed data to match products to queries on Shopping and to inform ad copy on other channels. A title like “Men’s Blue Merino Wool Crew-Neck Jumper” outperforms “Men’s Jumper” in both relevance and conversion rate.

  2. Apply conversion value rules by product category or margin tier. If your accessories carry a 60% margin and your hardware carries 20%, assign a value multiplier to the accessories category. The algorithm will then prioritise those products in auctions where it has a choice, shifting spend toward profit rather than revenue.

  3. Segment asset groups by audience intent and product margin. Group high-margin, high-intent products together with audience signals that reflect purchase-ready behaviour (recent cart abandoners, Customer Match purchasers). Lower-margin or awareness-stage products can sit in a separate asset group with broader signals.

  4. Use inventory-based bidding signals. If you have seasonal stock or clearance lines, adjust conversion values temporarily to accelerate sell-through. This is cleaner than budget manipulation and does not restart the learning phase.

  5. Run structured asset experiments. Google Ads allows you to run experiments comparing asset group variations. Test one variable at a time (a new video versus the existing one, or a revised headline set) and allow at least two weeks per test before drawing conclusions. Asset-level performance reports in the UI show which assets are rated “Best”, “Good”, or “Low” — replace “Low” assets promptly, but avoid deleting and rebuilding entire asset groups, which resets learning.

Feed engineering combined with value rules is the lever most ecommerce managers overlook. Practical ecommerce optimisation often combines feed enrichment with value rules and inventory-based bidding to prioritise high-margin SKUs. The accounts that do this consistently tend to see the clearest ROAS improvement over time.

Pro Tip: Structure asset groups by audience intent and product margin, not by channel. PMax decides the channel; you decide the creative context and the profit priority. An asset group built around “high-margin accessories, purchase-intent audience” will outperform one built around “all products, broad audience” because the algorithm has a tighter signal set to work with.

How does Performance Max interact with your other campaigns?

Search campaigns with exact-match keywords take priority over PMax for matching queries. This is the most important prioritisation rule to understand, because it means a well-structured Search campaign acts as a floor that PMax cannot undercut on your most valuable branded and non-branded terms.

Where overlap and cannibalisation become problems:

Brand exclusions are your primary defence. Apply them to every PMax campaign to prevent the algorithm from consuming branded traffic that your Search campaigns already capture efficiently. For non-branded overlap, use account-level negative keyword lists to exclude irrelevant queries that PMax surfaces in its search themes report.

The Performance Max vs Standard Shopping decision is one of the most common structural questions for ecommerce accounts. PMax gives broader channel coverage; Standard Shopping gives more granular bid control and clearer reporting. Many mature accounts run both, with PMax handling prospecting and Standard Shopping handling high-priority SKUs.

Recommended account structure by monthly ad spend:

On the question of how many PMax campaigns to run: fewer campaigns with more asset groups is almost always better than many campaigns with thin asset groups. Each campaign needs its own learning budget and conversion data. Splitting too early fragments signals and slows learning across the board.

Reporting, the learning phase, and troubleshooting

Expect the learning phase to run for 1–2 weeks in most cases, though Google advises allowing up to 6 weeks and avoiding frequent or large changes to budget, bidding strategy, or campaign status during that period. Patience here is not passive; it is a deliberate performance decision.

What to monitor in reporting:

Troubleshooting checklist:

Practitioners note that PMax is designed to be always-on, and frequent pausing restarts algorithmic learning and harms long-term performance. If you need to pause for operational reasons (a stock outage, a site issue), do so briefly and accept a short re-learning period on restart.

When Performance Max is not the right choice

PMax is not ideal when you lack reliable conversion measurement or need precise placement control. The algorithm cannot optimise toward a goal it cannot measure, and it has no mechanism to exclude specific placements within a channel (individual YouTube channels, for example) beyond the standard brand safety controls.

Red-flag scenarios where PMax may not be appropriate:

Short mitigations:

An honest view on when to hire an expert

Hiring an agency for Performance Max is worth it when margin optimisation, feed engineering, and scale are all in play at the same time. At that point, the complexity of managing conversion value rules, feed taxonomy, asset testing cadences, and campaign structure interactions exceeds what most in-house teams can sustain alongside their other responsibilities.

What good agency management should actually deliver is specific: feed engineering that goes beyond basic attribute completion to include custom labels for margin tiers, seasonal flags, and inventory status; conversion-value modelling that ties bidding targets to gross margin rather than revenue; structured experiment design that tests one variable at a time without breaking the learning phase; and account governance that prevents budget waste from cannibalisation, poor ad strength, or feed errors compounding over time.

The proof points worth requesting from any agency are straightforward. Ask for asset-level performance reports showing how creative decisions changed outcomes. Ask for a feed audit showing what was changed and why. Ask for experiment results, not just before-and-after revenue figures, which attribution models can distort. Transparent billing with no long-term contract requirement is a reasonable baseline expectation.

PMax rewards patience and input quality above all else. An agency that rushes to scale spend before the learning phase is complete, or that launches with thin assets and no value rules, is not adding value. The discipline of doing the groundwork first is what separates accounts that compound performance over time from those that plateau after the first month.

Oxedent’s approach to Performance Max for ecommerce

Oxedent builds and manages Performance Max campaigns for established ecommerce brands that need more than a basic setup. The work starts with feed engineering and conversion-value modelling, not creative uploads, because those inputs determine what the algorithm can achieve before a single pound is spent.

The core services Oxedent delivers for PMax:

If you are ready to scale Performance Max with a team that works exclusively in ecommerce PPC, talk to Oxedent about managed PMax campaigns or review the eCommerce PPC management service to see whether it fits your account.

Sources

The following official and practitioner resources are worth bookmarking if you are building or optimising a Performance Max campaign:

Exit mobile version