One campaign, every Google channel.
Performance Max (PMax) is a goal-based Google Ads campaign type that runs ads across Search, Shopping, Display, YouTube, Gmail, Discover, and Maps from a single campaign. According to Google Ads documentation, PMax uses AI to optimize bids, placements, and creative combinations in real time — replacing the need to manage separate campaigns per channel. Google reports that advertisers who switch to PMax see an average 13% increase in total incremental conversions at a similar cost per action.
Performance Max campaigns replaced Smart Shopping and Local campaigns in September 2022. Instead of choosing where your ads appear, you provide Google with conversion goals, creative assets, audience signals, and a budget. Google's machine learning decides which channels, audiences, and creative combinations generate the most conversions.
For ecommerce brands, PMax is now the default campaign type for Google Shopping. You cannot create standalone Smart Shopping campaigns anymore. Every store running Google Shopping ads interacts with PMax whether by choice or by migration.
The critical distinction: PMax is not a "set it and forget it" solution. The algorithm depends on the quality of the inputs you provide — product feed data, creative assets, audience signals, and conversion tracking. Garbage in, garbage out applies directly. Brands that treat PMax as a black box tend to burn budget on low-intent Display and Discover placements. Brands that structure their PMax campaigns with intent see results that match or exceed legacy Shopping campaigns.
Understanding how PMax fits within your broader ecommerce KPI framework determines whether the automation works for or against your goals.
Standard Shopping campaigns let you control which products appear for which search queries through negative keywords and priority settings. PMax removes most of that manual control and replaces it with algorithmic optimization across all Google surfaces. The tradeoff: you gain reach and automation, but lose granular query-level visibility and bidding control, as outlined in Google's campaign type comparison.
The differences matter for budget allocation and reporting:
| Feature | Standard Shopping | Search Campaigns | Performance Max |
|---|
| Channels served | Shopping tab, Search | Search only | All 7 Google channels |
| Keyword targeting | Product feed matching | Manual keyword bids | Automated by AI |
| Negative keywords | Full control | Full control | Limited (account-level only) |
| Search term visibility | Full report | Full report | Partial (category-level insights) |
| Bid strategy | Manual or automated | Manual or automated | Automated only (tROAS or tCPA) |
| Creative control | Product feed only | Text ads | Assets across all formats |
| Audience targeting | Limited | Keywords + audiences | Audience signals (suggestions) |
| Reporting granularity | Product-level | Keyword-level | Asset group level |
Sources: Google Ads Help, Search Engine Land PMax analysis
The loss of search term visibility is the most debated change. With Standard Shopping, you could see exactly which queries triggered your ads and add negative keywords to block waste. PMax shows aggregated search categories instead of individual terms. Google has gradually improved this reporting, but it remains less transparent than legacy campaigns.
For many ecommerce advertisers, the practical approach is running PMax alongside a standard Search campaign with exact-match brand terms. This prevents PMax from spending budget on branded searches that would convert anyway — a common complaint among early adopters.
Setting up PMax for ecommerce requires five components: a connected Merchant Center feed, at least one asset group with creative assets, audience signals, a conversion goal (typically purchases), and a bid strategy (target ROAS or target CPA). Google recommends running PMax for at least six weeks before evaluating performance, per their best practices guide.
Step 1: Connect Google Merchant Center
Your product feed is the foundation. PMax pulls product images, titles, descriptions, and prices from Merchant Center to generate Shopping ads automatically.
Before launching PMax, audit your feed:
- Titles — Include primary keywords, brand, size, color, material. Front-load the most important attributes. "Men's Organic Cotton T-Shirt — Black, Crew Neck" outperforms "T-Shirt."
- Images — Use high-resolution product photos on white backgrounds. Add lifestyle images as additional product images.
- Descriptions — Include search terms buyers use, not marketing copy. Specifications, materials, and use cases belong here.
- Product categories — Use Google's taxonomy accurately. Incorrect categorization hurts ad matching.
- Price and availability — Keep these synced in real time. Stale pricing causes disapprovals.
Step 2: Create Asset Groups
Asset groups are the building blocks of PMax creative. Each asset group contains:
- Up to 20 images (landscape, square, portrait)
- Up to 5 videos (if none provided, Google auto-generates from images)
- Up to 5 headlines (30 characters max)
- Up to 5 long headlines (90 characters max)
- Up to 5 descriptions (90 characters max)
- 1 business name and logo
- Final URL
Structure asset groups by product category or audience segment — not by channel. One asset group for "running shoes" and another for "trail shoes" lets Google tailor creative to each product line.
Step 3: Add Audience Signals
Audience signals tell Google who your ideal customer is. These are suggestions, not restrictions — Google will go beyond your signals if it finds converting audiences elsewhere.
Effective audience signals for ecommerce:
- Customer lists — Upload your buyer email list. This is the strongest signal.
- Website visitors — Retarget people who viewed products or added to cart.
- In-market segments — Users Google identifies as actively shopping your category.
- Custom segments — Based on search terms people use or URLs they visit.
Step 4: Set Conversion Goals and Bid Strategy
For ecommerce, set "Purchases" as the primary conversion goal with conversion value tracking enabled. Use target ROAS bidding and start with a target 20–30% below your actual goal. If your target ROAS is 400%, set the initial PMax tROAS to 280–320%. This gives the algorithm room to learn before you tighten the target.
Use a ROAS calculator to determine your break-even and target ROAS before setting bid strategy parameters.
Step 5: Set Budget and Launch
Google recommends a daily budget at least 3x your target CPA. For ecommerce with a $30 target CPA, that means $90/day minimum. Underfunding PMax starves the algorithm of data and produces unreliable results.
What Are the Best Practices for PMax Asset Groups?
The highest-performing PMax campaigns segment asset groups by product category, use all available asset slots, and refresh creative every 4–6 weeks. Google's internal data shows that campaigns using "Excellent" ad strength ratings generate 12% more conversions than those rated "Poor," according to Google's asset best practices.
Structure Asset Groups by Intent, Not Channel
A common mistake: creating one asset group for "all products" with generic creative. This forces Google to use the same headlines and images for running shoes and dress shoes, reducing relevance across every placement.
Instead, create separate asset groups for:
- Top-selling product categories (dedicate budget to proven performers)
- Seasonal or promotional products (time-bound messaging)
- High-margin products (prioritize profit, not just volume)
Fill Every Asset Slot
Google tests combinations of your assets across channels. More assets mean more combinations. Campaigns using all available asset slots give the algorithm more room to find winning combinations.
Minimum viable vs. optimal asset count:
| Asset Type | Minimum Required | Recommended | Optimal |
|---|
| Images | 1 | 10+ | 20 (max) |
| Videos | 0 | 1+ | 5 (max) |
| Headlines | 3 | 5 | 5 (max) |
| Long headlines | 1 | 5 | 5 (max) |
| Descriptions | 2 | 4 | 5 (max) |
Provide Your Own Videos
If you do not upload videos, Google auto-generates them from your images and text. These auto-generated videos look templated and typically underperform custom creative. Even a simple product demo shot on a phone outperforms Google's slideshow generator.
Test different creative angles the same way you would test ad creative on other platforms. Problem-aware hooks, social proof callouts, and product demonstration videos each serve different stages of buyer awareness.
Audience signals are suggestions to Google's algorithm, not hard targeting restrictions. Google uses your signals as a starting point for optimization, then expands to audiences it identifies as likely to convert. According to Google, providing strong audience signals accelerates the learning phase and helps PMax find high-value customers faster — but the algorithm will always explore beyond your signals.
This distinction trips up advertisers coming from Facebook Ads, where audience targeting constrains delivery. In PMax, audience signals are training data, not boundaries.
The hierarchy of signal strength:
- Customer Match lists (strongest) — Your actual buyers. Upload purchase email lists segmented by LTV or purchase recency.
- Website visitor lists — People who already know your brand. Segment by action: product viewers, cart abandoners, past purchasers.
- Custom segments by search terms — People who searched specific queries on Google. Use the same keywords you would bid on in Search campaigns.
- In-market and affinity audiences — Google's pre-built audience categories. Useful but broad.
Weak signals (generic in-market audiences with no customer list) force PMax to spend more of your budget on prospecting and learning. Strong signals (buyer lists, high-intent search terms) compress the learning phase and direct spend toward audiences with proven conversion patterns.
Ready to maximize the performance of your PMax creative assets? ConversionStudio generates conversion-optimized ad copy, headlines, and landing pages for ecommerce brands — so every asset group in your Performance Max campaign starts with proven messaging frameworks instead of guesswork.
How Should You Structure PMax Campaigns for an Online Store?
The recommended structure for ecommerce PMax is one campaign per product category with dedicated asset groups for each subcategory or audience segment. This balances algorithm learning volume with creative relevance. Running a single PMax campaign for an entire store dilutes signals; running 20 campaigns fragments budget.
A practical structure for a store selling athletic footwear:
Campaign 1: Running Shoes (daily budget: $100)
- Asset Group: Men's Running Shoes (audience signal: male runners, trail running searches)
- Asset Group: Women's Running Shoes (audience signal: female runners, marathon training searches)
Campaign 2: Training Shoes (daily budget: $75)
- Asset Group: Cross-Training Shoes (audience signal: gym-goers, CrossFit searches)
- Asset Group: Weightlifting Shoes (audience signal: powerlifting, strength training searches)
Campaign 3: Catch-All / Discovery (daily budget: $50)
- Asset Group: New Arrivals (broad audience signals)
- Asset Group: Sale Items (price-sensitive audiences)
This structure lets you allocate budget by category priority, measure performance at the category level, and tailor creative messaging to each product line.
Brand Campaign Isolation
Run a separate standard Search campaign for branded keywords (your brand name, branded product names). PMax will happily spend budget on branded searches that convert at high rates — inflating your PMax ROAS while providing zero incremental value. Brand campaigns with exact-match keywords and low bids capture this traffic at a fraction of the cost.
PMax optimization centers on four levers: feed quality, creative refresh, audience signal refinement, and budget reallocation. Unlike Search campaigns where you optimize keywords, PMax optimization is about improving the inputs that the algorithm uses to make decisions. Google recommends waiting 6+ weeks before making significant changes, per their optimization guide.
Week 1–6: Learning Phase
Do not change anything significant during the first six weeks. The algorithm needs conversion data to calibrate. Monitor but do not react to daily fluctuations. Check that:
- Conversion tracking fires correctly on purchase confirmation pages
- Product feed has no disapprovals or warnings
- Budget is spending fully each day (underspend signals targeting or asset issues)
Week 6+: Iterative Optimization
1. Audit asset performance. Check which assets are rated "Best," "Good," or "Low" in the asset details report. Replace "Low" performing assets with new variations. Keep "Best" assets running.
2. Review search term insights. Navigate to Insights > Search term insights. While not as granular as Search campaign reports, PMax shows search categories and trending queries. Look for irrelevant categories consuming spend and add account-level negative keywords.
3. Refine audience signals. Update customer match lists monthly with new buyers. Remove audience signals that show low conversion rates in the Audience Insights report.
4. Reallocate budget. Shift budget from low-ROAS campaigns or asset groups toward high performers. Small, incremental shifts (10–15% at a time) prevent resetting the algorithm.
5. Refresh creative quarterly. Even high-performing assets experience fatigue. Plan creative refreshes every 8–12 weeks to maintain engagement rates.
Track your optimization impact by monitoring the same KPIs that drive ecommerce profitability — ROAS, CPA, and contribution margin per channel.
What Are the Biggest Mistakes Ecommerce Brands Make With PMax?
The three most common PMax mistakes are running one campaign for the entire store, ignoring product feed quality, and measuring success by PMax-reported ROAS without accounting for brand cannibalization. Each mistake inflates perceived performance while reducing actual incremental revenue.
Mistake 1: Single campaign, no segmentation. One PMax campaign with one asset group and one audience signal gives the algorithm no structure. It will default to the easiest conversions — usually branded searches and remarketing — while ignoring prospecting.
Mistake 2: Neglecting the product feed. PMax Shopping ads depend on feed quality. Stores with thin titles, missing attributes, and low-quality images get fewer impressions and higher CPCs. Feed optimization is PMax optimization.
Mistake 3: Claiming inflated ROAS. PMax takes credit for conversions it did not cause. If a customer searches your brand name, clicks a PMax ad, and buys — PMax claims that conversion. Without a separate brand Search campaign to capture those clicks, your PMax ROAS looks 30–50% higher than its true incremental value.
Mistake 4: No video assets. Skipping video means Google auto-generates low-quality videos from your images. These run on YouTube and Discover, wasting impressions with templated creative that does not reflect your brand.
Mistake 5: Setting tROAS too high too early. An aggressive target ROAS during the learning phase restricts the algorithm. Start lower and increase gradually as conversion data accumulates.
Comparing PMax performance to other paid channels requires consistent attribution. See our breakdown of Google Ads vs Facebook Ads for how to evaluate cross-platform ROAS without double-counting.
PMax reporting is more limited than standard campaign types but has improved significantly since launch. You can now view asset group performance, search term categories, audience insights, and placement reports — though individual search query data remains aggregated. Google's PMax reporting guide details available metrics.
Key reports available in PMax:
- Asset group reports — Performance by asset group including conversions, conversion value, and cost.
- Asset details — Individual asset ratings (Best, Good, Low) showing which headlines, images, and videos perform.
- Search term insights — Categories of search terms that triggered your ads, with conversion data per category.
- Auction insights — How your impression share compares to competitors.
- Placement reports — Where your ads appeared (YouTube channels, websites, apps). Useful for identifying low-quality placements to exclude.
Third-Party Reporting Supplements
Because PMax's native reporting has gaps, many ecommerce brands use supplemental tools:
- Google Analytics 4 — Compare PMax-attributed conversions against GA4's data-driven attribution model for a second opinion on incremental value.
- Merchant Center reports — Product-level click and impression data that PMax does not surface directly.
- Scripts and APIs — Custom Google Ads scripts can pull placement-level data and search term data at finer granularity than the UI provides.
Frequently Asked Questions
Google recommends a daily budget at least 3x your target CPA. For ecommerce stores targeting a $25 CPA, that means $75/day minimum ($2,250/month). Campaigns with less than $50/day often lack sufficient conversion volume for the algorithm to optimize effectively. Start with your best-selling product category rather than spreading thin across all products.
Yes. Google recommends running PMax alongside standard Search campaigns for branded keywords. PMax takes priority over standard Shopping campaigns for the same products, so avoid running both Shopping and PMax for the same product set. The standard approach is PMax for Shopping plus standard Search for brand terms and high-value exact-match keywords.
Google states the learning phase requires a minimum of six weeks with sufficient conversion volume (at least 30 conversions in 30 days). Some advertisers see stable performance within 3–4 weeks, while complex product catalogs may need 8–10 weeks. Avoid making significant structural changes during the learning phase, as each change resets the learning period.
PMax works for stores with as few as one product, but the campaign structure changes. With fewer products, you rely more heavily on asset group creative variety and audience signals rather than feed breadth. Stores with fewer than 20 SKUs should focus on building 3–5 strong asset groups with distinct creative angles rather than segmenting by product category.
How do you prevent PMax from wasting budget on Display and Discover?
You cannot manually exclude channels from PMax. However, you can influence channel allocation by providing strong Shopping-intent audience signals (customer match lists, in-market shopping audiences), optimizing your product feed, and setting a target ROAS that makes low-converting Display placements unprofitable for the algorithm. Review the placement report monthly and exclude specific low-quality placements (mobile games, spam apps) at the account level.
Keep Reading
