Google Shopping’s AI-Powered PMax Overhaul Is Reshaping DTC Acquisition Costs
Google's Performance Max overhaul, rolled out broadly in May 2026, is forcing DTC brands to rethink bidding strategy, creative inputs, and attribution as CPAs shift dramatically across categories.
By Ryan Wilson ·
·
7 min read
For much of the past two years, Performance Max campaigns were a source of frustration for DTC marketers — a black box that delivered inconsistent results and gave operators little insight into where their budgets were actually going. But Google’s May 2026 overhaul of PMax, which embedded Gemini-powered creative generation, granular asset group reporting, and a new “brand exclusion” bidding layer, is forcing brands to re-evaluate the channel from scratch. The early results are mixed, and the industry is watching closely.
Across the Shopify merchant community, the PMax changes are already showing up in the numbers. Media buyers report CPAs swinging by 20% to 45% in either direction within the first 30 days of campaign migration, depending heavily on how well brands structured their asset groups and how much first-party data they fed into Google’s Customer Match pools before launch.
📊 Marketing & Growth · By The Numbers
📈
20%
Growth
🎯
45%
Impact
💰
61%
Revenue
⚡
9x
Efficiency
What Exactly Changed in Google’s May 2026 PMax Update?
The May rollout was the most significant structural change to Performance Max since its 2021 introduction. Google embedded its Gemini 1.8 model directly into asset generation, allowing PMax to autonomously draft headlines, descriptions, and image variants from a brand’s product feed and landing page content. More importantly, Google introduced channel-level reporting — for the first time, advertisers can now see budget allocation broken down by Shopping, YouTube, Display, Search, Discover, and Gmail within a single campaign.
The brand safety layer, which Google previewed at its Marketing Live event in April, lets brands exclude specific competitor keywords and brand terms from triggering their own Shopping placements — a feature that aggregators and multi-brand operators had been requesting since 2023.
“The channel-level reporting alone is worth the migration pain. We were flying blind for two years. Now I can see that 61% of our PMax spend was going to Display with a 9x CPA relative to Shopping. That’s not a campaign — that’s a money leak.” — Cody Iverson, co-founder of Vestr, a performance agency managing $40M+ in annual Google Ads spend for DTC apparel brands
💡 Article Summary
Key Insights
1
What Exactly Changed in Google’s May 2026 PMax Update?
2
How Are DTC Brands Restructuring Their Google Shopping Strategy?
3
What Does the PMax Overhaul Mean for Standalone Google Shopping Campaigns?
4
How Is the PMax Shift Affecting Customer Acquisition Costs Across Categories?
5
What Tools and Integrations Are Operators Using to Manage the Transition?
Source: Ecommerce Times
How Are DTC Brands Restructuring Their Google Shopping Strategy?
The brands seeing the strongest early results share a common playbook: they are treating PMax less like a set-it-and-forget-it campaign and more like a structured media buy with deliberate creative and data inputs. Industry practitioners are converging on several tactical shifts:
Asset group segmentation by margin tier: Rather than one PMax campaign per product catalog, leading operators are breaking asset groups by gross margin — typically 60%+ GM products in one group, lower-margin SKUs in another, each with separate ROAS targets.
Customer Match audiences loaded pre-launch: Brands that uploaded 90-day purchaser lists and 180-day browse lists to Customer Match before activating PMax reported 18–30% lower CPAs in the first two weeks versus cold-start campaigns, according to agency benchmarks shared at the Commerce Growth Summit in Chicago in May.
Suppressing cannibal spend on branded queries: The new brand exclusion layer is being used aggressively. Several agencies report reallocating 12–18% of recaptured branded query budget into dedicated Standard Shopping campaigns where bid control is more precise.
Gemini-generated creative as a testing input, not a final asset: Most experienced media buyers are using Gemini’s auto-generated headlines as a starting point, then overriding with manually crafted copy once asset performance data surfaces after 14–21 days.
Offline conversion imports via Shopify’s Google & YouTube app: Brands with high in-store or phone order volume are importing offline conversions at the order level using Shopify’s native Google channel integration, improving signal quality for Gemini’s bidding model.
Direct-to-consumer pet brand Bark Avenue, which sells across its own Shopify store and Amazon, reported that restructuring its PMax asset groups by product margin and uploading a 6-month Customer Match file dropped its blended Google Shopping CPA from $38 to $24 within 45 days. The brand’s director of growth, Mara Chen, credited the channel-level reporting for surfacing that 38% of its PMax budget had been funding Display placements with an effective CPA above $90.
“We had no idea we were burning $14,000 a month on Display within PMax. The moment we could see it, we fixed it. Our overall Google ROAS jumped from 2.8 to 4.1 in six weeks — same budget, completely different outcome.” — Mara Chen, Director of Growth, Bark Avenue
What Does the PMax Overhaul Mean for Standalone Google Shopping Campaigns?
The May update has reignited an old debate: should DTC brands run Standard Shopping campaigns alongside PMax, or let PMax absorb the full catalog? The consensus in 2025 was to migrate everything to PMax. In mid-2026, practitioners are pulling back from that position.
The current emerging best practice, as outlined by Tinuiti’s Google Shopping team in a June 2026 playbook shared with agency partners, is a hybrid architecture: Standard Shopping campaigns for top-performing SKUs where brands want precise bid control and proven creative, and PMax handling new product launches, broad catalog coverage, and YouTube/Display prospecting where Gemini’s automation has a data advantage.
Tinuiti’s VP of Marketplace Strategy, Andy Ellwood, noted in a LinkedIn post last week that brands running this hybrid structure were seeing 8–15% lower blended CPAs than those running PMax-only across comparable verticals including home goods, apparel, and sporting goods.
How Is the PMax Shift Affecting Customer Acquisition Costs Across Categories?
The impact on CAC is uneven. Categories with rich first-party data — subscription brands with large email lists, Shopify stores with multi-year purchase histories — are seeing CPAs compress as Google’s model benefits from stronger Customer Match signal. Categories with thin data histories, including newer dropshipping operations and marketplace-only sellers who are now building out DTC storefronts, are seeing CPAs rise as the Gemini model takes longer to optimize without historical conversion data.
Benchmarks compiled by Triple Whale from a cohort of 1,200 Shopify brands using Google Ads as of June 2026 show the following median CPA changes since the May PMax rollout:
Health & Wellness (non-supplement): CPA down 14% for brands with 10,000+ Customer Match records
Apparel & Footwear: CPA down 9% for brands using Gemini creative with manual override; up 11% for brands relying fully on auto-generated assets
Home & Kitchen: CPA flat to down 6%, with significant variance based on feed quality
Consumer Electronics: CPA up 18% on average, driven by increased auction competition from Amazon Ads’ expanded Google Shopping presence
Beauty & Personal Care: CPA down 22% for brands with subscription data feeding Customer Match
The electronics spike is being closely watched. Amazon has significantly expanded its Google Shopping investment in Q2 2026, bidding aggressively on branded and category keywords in electronics, and its scaled conversion data gives it an auction advantage over most independent DTC brands in that vertical.
What Tools and Integrations Are Operators Using to Manage the Transition?
The operational overhead of managing the new PMax structure has accelerated adoption of several third-party tools. Feedonomics, which manages product feed optimization for over 14,000 brands, reported a 40% increase in clients activating its supplemental feed feature in May and June 2026, as operators worked to improve data quality ahead of PMax migration. Richer product titles, accurate availability attributes, and margin data passed via custom labels are the primary levers brands are using to improve Gemini’s asset selection and bidding decisions.
On the analytics side, Northbeam and Triple Whale both released PMax-specific attribution dashboards in June that break out channel-level performance using first-party pixel data — addressing the longstanding complaint that Google’s own reporting overstated PMax contribution by including view-through conversions that other channels were driving.
“Google’s native PMax reporting will always favor Google. The only way to get an honest read on channel contribution is to run your own pixel and attribution model in parallel. Brands that skip that step are optimizing against bad data.” — Rabah Rahil, Chief Evangelist, Triple Whale
Agencies managing larger accounts are also leaning on Google’s API to automate asset group creation at scale — particularly relevant for catalog-heavy operators with 500+ SKUs who cannot manually build asset groups for each product tier.
What Should Merchants Do in the Next 60 Days?
For Shopify and DTC operators who have not yet acted on the PMax changes, the operational window to course-correct before Q4 planning is narrowing. Practitioners recommend a prioritized action sequence:
Audit current PMax asset groups for channel-level spend concentration using the new breakdown reports — campaigns over-indexing on Display or Discover with CPAs above 3x your Shopping benchmark should be restructured immediately.
Upload or refresh Customer Match audiences using purchase data from the last 180 days before expanding PMax budgets. Brands on Klaviyo can use its Google Audience Sync integration to automate this refresh on a rolling 30-day basis.
Implement custom labels in your product feed to segment by gross margin tier, allowing separate ROAS targets for high- and low-margin SKUs within PMax.
Test the new brand exclusion feature to prevent PMax from cannibalizing branded search queries, then reallocate that recaptured budget into a dedicated brand campaign with manual CPC bidding.
Install an independent attribution tool (Northbeam, Triple Whale, or Rockerbox) if you are relying solely on Google’s conversion reporting to judge PMax performance.
The broader picture for the second half of 2026 is one where Google Shopping’s performance ceiling has risen for brands with strong data assets, but the skill gap between sophisticated operators and average merchants is widening. PMax’s Gemini backbone rewards brands that invest in feed quality, first-party data infrastructure, and thoughtful creative inputs — and punishes those who treat it as an autonomous spend vehicle. With Q4 budgets being finalized in the next six to eight weeks, the brands that restructure now will have a meaningful cost-per-acquisition advantage heading into the most competitive shopping period of the year.
