Google launched AI Max for Shopping on April 30, 2026 as a one-click upgrade for any standard Shopping campaign. The system reads your Merchant Center feed attributes — Google specifically named fabric softness, material durability, and fit — to generate ad copy on the fly, pick landing pages with Final URL Expansion, and choose ad formats with Optimal Format Selection. Your feed is now the ad copy, the landing-page picker, and the format selector. The default Google for WooCommerce plugin emits a barebones feed: title, description, price, image, GTIN. Three of Google’s named example attributes don’t exist as default fields. Don’t flip the upgrade until you’ve enriched.
What the Upgrade Actually Touches
AI Max for Shopping introduces three feature components — Text Customization, Final URL Expansion (FUE), and Optimal Format Selection — all of which read directly from the Merchant Center feed (PPC News Feed, April 30, 2026). The pitch from Google is straightforward: one click on an existing Shopping campaign turns on long-tail conversational query matching, AI-generated headlines, and dynamic landing-page selection.
What gets lost in the launch coverage is that none of those three features generate anything from nothing. Text Customization generates copy from your feed attributes. FUE picks landing pages from your URL inventory. Optimal Format Selection assembles formats from your existing images and structured data. If those inputs are thin, the outputs are thin too — only now they’re thin and auto-generated, which is harder to spot in a reporting dashboard than a stale RSA was.
Feed-based ads now account for a median 90% of Performance Max spend (XICTRON Performance Max guide, 2026). AI Max for Shopping doubles down on the same architecture: the feed is the campaign. The bidding strategy, the keyword theme, the audience signal — all of those are downstream of what Google can read on each product line.
The Default WooCommerce Feed Problem
A stock WooCommerce + Google for WooCommerce installation emits roughly these Merchant Center attributes per product: id, title, description, link, image_link, availability, price, brand, gtin, condition, shipping. That’s the floor — and for a long time it was enough. AI Max for Shopping moved the floor.
Google’s named example attributes — fabric softness, material durability, fit — aren’t in the default WooCommerce product model at all. They don’t exist as core fields. They don’t ship as default WooCommerce taxonomies. They’re not in the Google for WooCommerce plugin’s auto-mapping. Three of the attributes Google explicitly cited as inputs for the new AI features are structurally absent from a default WooCommerce store.
To get them into the feed, you need one of: WooCommerce custom fields (manually configured per product), a feed-enrichment plugin like CTX Feed or AdTribes, product taxonomies built specifically for these attributes, or a server-side pipeline that pulls from your existing product data and packages it into the Merchant Center schema before Google reads it.
You may be interested in: What Three Years of WooCommerce Data Can Tell You That Three Months Cannot
The 250-Campaign Reality Check
An independent 250-campaign study analysed AI Max performance against traditional match types and found AI Max delivered approximately 35% lower return on ad spend on average — and crucially, the variance correlated with feed and conversion-data quality (PPC Land, November 2025). Campaigns running on thin feeds with weak conversion signal underperformed badly. Campaigns running on enriched feeds with reliable CAPI-grade conversion data closed most of the gap.
Translation: AI Max isn’t broken. It’s hungry. Feed it well and it performs; feed it poorly and it spends your budget chasing queries it shouldn’t have matched in the first place.
The Shopping Graph that AI Max queries against contains over 35 billion product listings (Shine Dezign, 2026). Your product needs to be distinguishable in that index — by attribute, by descriptor, by structured field. A title that says “Men’s Cotton Shirt – Blue” is competing against ten thousand identical strings. A title plus seven enrichment attributes (fit type, fabric weight, sleeve length, collar style, care instructions, occasion, season) is a distinct entry the AI can actually match to “lightweight breathable cotton shirt for hot weather work”.
The 30-Day Enrichment Path
The right order of operations isn’t “flip the upgrade and see what happens.” It’s “audit the feed, fix the gaps, then flip.” Here’s a workable 30-day sequence.
Week 1: Audit Attribute Depth Per Category
Pull your current Merchant Center feed and count populated attributes per product, grouped by category. Any category averaging five or fewer populated attributes is failing the AI Max readiness bar. Apparel, soft goods, and home textiles are usually the worst offenders because they need texture, fit, and material descriptors that don’t map to default WooCommerce fields. Electronics tend to score better because GTIN-driven attributes auto-populate from manufacturer data.
Week 2: Map Missing Attributes to Source Fields
For every missing attribute, identify where the data already exists in your store: WooCommerce custom fields, ACF fields, product variation attributes, taxonomy terms, or copy buried in product descriptions. The work isn’t “create new data” — most stores already have this information typed somewhere. The work is routing it into the Merchant Center schema.
Week 3: Build the Enrichment Pipeline
This is where the choice between a feed-enrichment plugin and a server-side pipeline matters. A plugin sees only what one PHP process can scrape. A server-side pipeline can pull from WooCommerce hooks, custom tables, variation metadata, and even external systems like a PIM, and merge them into a canonical feed before Google reads it. The store with seven attributes per product wins the auction the store with three never enters.
Week 4: Test on a Small Campaign First
Pick one campaign — ideally one product category where you’ve enriched — and flip AI Max on it. Watch the search terms report (Google now provides this from day one for AI Max, unlike the early Performance Max playbook). If you see conversational queries showing up that match your enriched attributes, the system is working. If you see broad, off-brand queries with high spend and low conversion, FUE is taking liberties with your URLs that you need to constrain.
You may be interested in: WooCommerce to BigQuery: $5/Month. Shopify to BigQuery: $500/Month.
Here’s How You Actually Do This
Feed enrichment can’t live in the browser. The Merchant Center feed Google reads is generated server-side, from product data that already exists in your WooCommerce database, and routed to Google through a feed URL or API push.
Transmute Engine™ is a first-party Node.js server that runs on your subdomain (e.g., data.yourstore.com). The inPIPE WordPress plugin hooks into WooCommerce — every product, every variation, every custom field, every taxonomy term — and sends a structured payload to the Transmute Engine server. Transmute Engine then formats and routes that data into the canonical Merchant Center schema, with all the enrichment attributes AI Max wants to read, before Google’s crawler ever sees the feed. The same pipeline emits server-side conversion events to Google Ads Enhanced Conversions, so AI Max trains on a clean signal in both directions — feed depth on the way in, conversion accuracy on the way out.
Key Takeaways
- AI Max for Shopping launched April 30, 2026 as a one-click upgrade for any standard Shopping campaign — it reads Merchant Center feed attributes to generate ad copy, pick landing pages, and select formats.
- Feed-based ads now account for a median 90% of Performance Max spend (XICTRON 2026) — feed quality is the highest-leverage lever in AI-driven shopping campaigns.
- Google explicitly named fabric softness, material durability, and fit as attributes AI Max reads — none of which exist as default fields in the standard Google for WooCommerce plugin.
- An independent 250-campaign study found AI Max ROAS ~35% lower than traditional match types when feed and conversion-data quality were thin (PPC Land, November 2025).
- Audit attribute depth per category before flipping the upgrade. Categories averaging five or fewer populated attributes need 30 days of enrichment first.
Frequently Asked Questions
Not until you audit your feed attribute depth. AI Max for Shopping, launched April 30, 2026, reads Merchant Center feed attributes to generate ad copy and pick landing pages — if your products average five or fewer populated attributes, AI Max has nothing to work with. An independent 250-campaign study found AI Max delivered approximately 35% lower ROAS than traditional match types when feed quality was thin (PPC Land, November 2025). Enrich your feed first with attributes like fit, material, fabric weight, and care instructions, then flip the upgrade on one campaign as a test.
Google explicitly named fabric softness, material durability, and fit as example attributes AI Max for Shopping reads, alongside the standard Merchant Center schema (title, description, price, image, GTIN, brand, availability, condition, shipping). The system also uses category-specific attributes like size, color, age group, gender, pattern, and material for apparel. AI Max for Shopping introduces three components — Text Customization, Final URL Expansion, and Optimal Format Selection — and all three read directly from the feed rather than from campaign-level settings.
Final URL Expansion (FUE) lets AI Max replace the destination URL in your Shopping ad with a more relevant page from your domain based on the query. If your site structure is clean — one product per URL, no orphaned pages, no near-duplicate variants competing for the same intent — FUE generally improves landing-page relevance. If you have legacy URLs, faceted-search pages indexed alongside product pages, or thin category pages, FUE may route traffic to lower-converting destinations. You can constrain FUE behaviour through campaign-level URL inclusion and exclusion rules in Google Ads.
No. The default Google for WooCommerce plugin emits a barebones feed — title, description, price, image, GTIN, brand, availability, condition, shipping. Three of Google’s explicitly named AI Max attributes (fabric softness, material durability, fit) are not in the default WooCommerce product model. You need one of: WooCommerce custom fields configured per product, a feed-enrichment plugin like CTX Feed or AdTribes, dedicated product taxonomies for these attributes, or a server-side pipeline that pulls from existing product data and packages it into the Merchant Center schema before the feed is generated.
What Changes If You Don’t Enrich
Stores that flip the AI Max upgrade today without enriching their feed will see budget shift toward conversational long-tail queries the algorithm guesses might match their products — guesses made from title, price, and image alone. The store next door with the same products and a seven-attribute feed will win the auctions for the high-intent queries. Your store will win the broad ones at lower conversion rates. Same products, same budget, different outcomes — decided by what Google’s crawler can read on your feed.
The 30-day audit-and-enrich path doesn’t need to be elaborate. It needs to be honest. Count attributes per category. Find the gaps. Route the data that already exists in your store into the schema Google now expects. Then flip the upgrade.
Talk to Seresa about feed enrichment and server-side conversion routing: seresa.io.
