Google Made Pickup Cost Mandatory in 32 Markets to Feed AI Max

Google announced on April 28, 2026 that pickup_cost and minimum_order_value will become mandatory feed attributes in the United Kingdom, Switzerland, and all 30 EEA countries on September 30, 2026, for any merchant whose in_store products are enabled for pickup. Products that fail to comply may be disapproved. The compliance window is roughly 150 days, and the audit work has to happen on both the feed and the website at the same time.

The PPC trade press wrote it up as a Merchant Center spec change. PPC Land covered the announcement as part of the broader 2026 product data specification update, and Search Engine Roundtable folded it into the same April 14 changelog. Both takes are accurate. Both are also missing the actual reason this update exists.

The two attributes, and what they actually carry

The official announcement, posted to the Merchant Center help center on April 28, 2026, lays out two new requirements:

pickup_cost takes two sub-attributes. pickup_cost_flat_rate is the fee a shopper pays to reserve and collect the product in store. pickup_cost_free_threshold is the basket value at which that fee disappears. If you charge nothing, you submit zero. If you charge a fiver below £30 and free above, you submit both. The structure is closer to your shipping settings than your price field, which means most retailers will need an engineering ticket, not a feed-tweak.

minimum_order_value goes one level deeper. It carries country, service, surface, and price as sub-attributes, so a single SKU can declare different thresholds for different countries and different fulfillment services. The field has been documented as optional since 2024. September 30 is when optional turns into a disapproval condition.

The trigger is conditional. Google’s wording: “if pickup costs or minimum order values are applied on your website and in_store products are enabled for pickup, you will have to provide the associated costs.” Read carefully. The trigger isn’t whether you want to declare a fee. It’s whether your website charges one. If your site charges £4 for pickup at low basket sizes and your feed says nothing, your products are now mismatched against your own checkout.

This is not really an admin update for Shopping

Google does not roll mandatory feed fields across 32 markets in five months for hygiene. There is a product reason, and the reason is on Google’s own front page.

AI Max for Shopping launched as Google’s pitch to retailers that the Merchant Center feed is now the substrate for AI-driven matching, not a structured catalog. Their framing: AI Max reads your Merchant Center feed to understand product context, then expands reach against a wider matching surface. What AI Max needs to compete with Amazon in agentic commerce is not a sharper product title. It’s a richer fulfillment model that can answer the question shoppers actually ask: “can I get this today, where, and what does it cost me to do that.”

Local pickup is the one thing Amazon’s European logistics network doesn’t answer easily. GeekWire’s read on Google’s Universal Commerce Protocol noted that Amazon was conspicuously left out of the early UCP rollout. Etsy, Wayfair, and a Target/Walmart/Shopify cohort are inside. The pitch to retailers is: route shoppers to your store before they end up inside the Amazon app. That pitch only works if Google can rank, surface, and price local-pickup inventory accurately, which it cannot do today because two of the relevant signals are optional.

This connects to a bigger pattern. Google is auto-upgrading DSA campaigns to AI Max in September too. Same month, same direction. The feed gets richer, the campaign type gets handed to AI, and the merchants who didn’t prepare the underlying data get worse outputs.

The trap is the conditional, not the deadline

If you read the announcement quickly, you might conclude that the rule only applies if you charge a pickup fee, so most merchants are fine. That’s where the disapprovals will come from.

The trigger is what your website does, not what you intend. If a single SKU below a free-shipping threshold incurs a pickup fee at checkout in any of the 32 markets, and your feed doesn’t carry that, you’re non-compliant. If a £29 basket triggers a £35 minimum at any participating store, and the feed declares no minimum_order_value, you’re non-compliant. ALM Corp’s field guide flagged the obvious risk early: low-priced items are most likely to trip a hidden minimum, and they’re also the SKUs that disproportionately drive volume in Shopping.

Google’s history with this kind of mandate is consistent. Disapproval doesn’t arrive as a notice. It arrives as a quiet drop in impressions. By the time you correlate the impression decline to a feed audit, the holiday quarter has done some damage. From what I’ve seen, the merchants who lose the most aren’t the ones who ignore the rule. They’re the ones who assumed the rule didn’t apply.

A pre-September audit you can run this week

Five things, in order. Most of this is feed-side work, but the awkward part is that two of them require sitting down with whoever owns checkout.

One. Pull a sample of in-store-pickup-enabled SKUs across each of the 32 markets and step through the actual checkout. Note where pickup fees, free-pickup thresholds, or minimum order values appear. If anything appears that’s not in the feed, that SKU is at risk on October 1.

Two. Audit the existing shipping and minimum_order_value attributes for every market. The April 14 spec update bundled several related changes to shipping attributes, and merchants who update pickup_cost in isolation will probably miss adjacent breaks. Treat this as one project, not two.

Three. Confirm the country-service-surface-price tuple structure for minimum_order_value matches what your engineering team can actually emit. The field accepts multiple thresholds per SKU. If your feed generator only handles one, you have a build cost, not a config change.

Four. Test against a pickup_cost of zero where appropriate. If you offer free pickup with no minimum, that’s declarative, and submitting 0 is the right answer. The field expects something, not nothing.

Five. Set a reminder for September 15. Get the feed live two weeks before the deadline so you have time to catch disapprovals while the support team is still treating it as a pre-launch courtesy rather than an enforcement action. Submitting on September 30 means competing for support attention with everyone else who waited.

What I’d watch through October

Two things, both quiet. Watch your Merchant Center diagnostics for any spike in “limited” or “disapproved” status on in-store SKUs the first week of October. Watch your local-inventory-ad impression share against a baseline week from August. The tell will not be a notice. It will be the gap.

And the broader thing. Google’s 2026 commerce roadmap is built around the assumption that retailers will let AI route inventory, route ads, and route fulfillment. Mandatory pickup signals are one of the cheaper steps in that build. There will be more. The merchants who treat each one as paperwork will keep losing visibility to the ones who treat the feed as the product.

I think the rule itself is reasonable. The version of this story that ages badly is the one where most retailers find out their feed has been quietly wrong since late September because nobody on the marketing side talked to anyone on the checkout side.

Notice Me Senpai Editorial