Performance Max Agency New York Ecommerce: The Complete 2026 Guide

Performance Max now drives 45% of all Google Ads conversions. For New York ecommerce brands, it is no longer a campaign type to evaluate. It is the dominant Google Ads format, and how it is structured, fed, and actively managed determines whether it becomes the most efficient paid channel in the account or the most expensive source of wasted budget.

The brands getting 35% to 55% higher ROAS from Performance Max than the average are not running default settings. They are making deliberate structural decisions about asset groups, audience signals, search themes, brand exclusions, and feed quality that most single-campaign setups never address.

If your brand is searching for a Performance Max agency in New York that manages PMax as a full system rather than as automation left to run itself, this guide covers what that actually involves in 2026, including the August bidding update every ecommerce advertiser needs to act on before it takes effect.

About the Author

Shlomie Spielman is the founder of Seller Splash, a New York ecommerce performance marketing agency. After managing Performance Max alongside Standard Shopping, branded Search, and multi-platform paid media for product brands across Shopify, WooCommerce, BigCommerce, and Magento, he built Seller Splash's Google Ads practice around one principle: PMax amplifies the quality of the inputs you give it. Strong structure and strong data produce strong results. The insights in this guide come from live account management across New York ecommerce accounts, not from platform documentation.

What Performance Max Is and Why It Requires Active Management

Performance Max is a single campaign type that serves ads across all of Google's inventory simultaneously: Search, Shopping, YouTube, Display, Gmail, Discover, and Maps. Google's machine learning determines placement, audience, creative combination, and bid in real time. The advertiser provides asset groups, audience signals, a product feed, conversion goals, and a budget.

The most common misconception about PMax is that it is a set-and-forget automation tool. Accounts treated that way consistently underperform. The algorithm amplifies whatever inputs it receives. Give it a generic single asset group with no audience signals, no search themes, and a product feed full of storefront-optimized titles, and it will find cheap traffic that does not convert. Give it well-structured asset groups, a margin-aware feed, strong Customer Match signals, and 50 precisely chosen search themes per group, and it will deliver measurably better outcomes than any previous campaign type for the same budget.

A 2024 Optmyzr study across 24,702 Performance Max campaigns found that 82% of advertisers were running PMax alongside other campaign types, not as a standalone replacement. This is the 2026 consensus: PMax in isolation is not the answer. The hybrid structure with Standard Shopping is.

The Feed Is the Campaign: Why Most PMax Underperformance Starts Here

Feed-based Shopping placements account for 60% to 80% of Performance Max spend in well-managed ecommerce accounts. The product feed in Google Merchant Center is therefore not a setup task for PMax. It is the primary ongoing management lever that determines what the algorithm has to work with.

Google's April 2026 Merchant Center product data specification update raised the minimum image resolution to 500x500 pixels across all product categories and introduced a new video link attribute in the feed. Accounts with lower-resolution images now face reduced eligibility for Shopping placements. Adding video links directly into the feed enables video-format ads to run alongside Shopping placements without requiring separate creative work.

The three feed improvements that move PMax performance most:

Product title structure. Titles written for storefront aesthetics miss the search query matching that Shopping placements depend on. Optimized product titles have been shown to double or triple click-through rates by matching the queries buyers actually use. The formula: Product Type plus Key Attribute plus Material or Specification plus Color or Variant, within 150 characters with the most important terms in the first 70.

Image quality. High-resolution, properly lit product photography on clean or lifestyle backgrounds delivers 15% to 25% higher CTR compared to low-quality or manufacturer stock images. With the 2026 resolution requirement now in place, this is an eligibility threshold as much as a performance improvement.

Custom labels for margin segmentation. A single Target ROAS applied across an entire product catalog forces PMax to optimize high-margin and low-margin products identically. Custom labels tag products by margin tier, bestseller status, inventory level, or seasonal priority, enabling separate asset groups or campaigns with ROAS targets that reflect actual product economics rather than catalog averages.

Asset Groups: The Structural Foundation of Performance Max

An asset group is the organizational unit inside PMax. Each combines up to 15 headlines, 5 long headlines, 5 descriptions, 20 images, and 5 videos that Google's AI assembles into optimized combinations per placement. The quality and variety of these assets directly determine where and how often ads appear.

The optimal asset group structure for New York ecommerce accounts:

Three to seven asset groups per campaign. Fewer than three provides insufficient segmentation for the algorithm to allocate budget by theme. More than seven splits conversion volume too thinly for the machine learning to optimize each group effectively.

Segment by product category or audience intent, not by audience signal alone. A running shoes brand should have separate asset groups for men's performance, women's casual, and trail running, each with tailored headlines, images, and search themes specific to that segment. Generic imagery and copy shared across all groups produces generic algorithmic decisions.

At least one video per asset group is not optional. Google's algorithm favors campaigns with video assets, and without a provided video, Google auto-generates one from product images using its own AI. The auto-generated result is typically low quality and reduces campaign performance. As of 2026, Google's Asset Studio uses Imagen 4 and Veo 3 to generate video directly inside Google Ads, removing the production barrier for brands without video assets.

Asset rotation and refresh cadence. Creative fatigue in PMax begins within two to three weeks at significant spend. The management discipline is adding three to five new assets every six to eight weeks and replacing assets scored as "Low" in the Asset Group Performance report after they have accumulated at least 14 days of impression data.

Audience Signals: Directing the Algorithm Without Restricting It

Audience signals in Performance Max are not hard targeting restrictions. They are directional hints that tell Google's AI where to start its optimization, after which the algorithm expands beyond those signals to find additional buyers. The distinction matters: brands that treat audience signals like traditional audience targeting and keep them narrow are limiting PMax's reach unnecessarily.

The four-layer audience signal stack for ecommerce PMax campaigns:

Customer Match lists. Uploading existing customer purchase data as Customer Match provides the highest-quality audience signal available in PMax. The algorithm uses this data to find new buyers who match the behavioral and demographic profile of existing customers. Google's high-value new customer mode, rolled out to all advertisers in early 2025, allows the algorithm to bid more aggressively for new buyers who resemble the brand's highest-LTV customers rather than just any converting customer.

In-market audiences. Add in-market audience segments for categories directly relevant to the product. An athletic footwear brand adds "Athletic Footwear" and "Sports Equipment" in-market segments. A kitchen goods brand adds "Kitchen and Dining" and "Home Furniture." These tell the algorithm which actively shopping intent signals to prioritize.

Custom intent segments. Build custom segments from the specific search queries buyers use when researching products in the category. For a New York DTC brand selling leather wallets, custom segments might be built from searches for "slim wallet" and "minimalist bifold" rather than the generic "wallet." The more specific the search behavior that defines the segment, the more efficiently PMax finds buyers with genuine purchase intent.

Remarketing and recent purchaser exclusions. Including past purchasers in prospecting PMax campaigns causes the algorithm to re-serve ads to buyers who already converted, inflating ROAS with repeat purchases while reducing the actual new customer acquisition the budget is supposed to generate. The correction: exclude recent purchasers (last 30 to 60 days) from prospecting campaigns using audience exclusions in campaign settings.

Search Themes: Guiding PMax Toward the Right Queries

Search themes are query-level signals added at the asset group level that direct PMax toward the search inventory where it should compete. Google expanded the search theme limit from 25 to 50 per asset group in 2026, and using the full allocation is one of the highest-leverage free optimization actions available.

Search themes behave like broad match keywords. They guide the algorithm toward related queries without triggering exact matches. The practical implication: specific themes outperform generic ones by a wide margin. "Women's waterproof trail running jacket size medium" directs the algorithm far more precisely than "running jacket."

Google now shows a "usefulness" indicator for each search theme that reveals whether the theme is generating additional traffic or whether PMax would have found those queries anyway without the theme. Monitoring this indicator monthly and replacing low-usefulness themes with more specific alternatives is the optimization cycle that keeps PMax's search distribution aligned with actual buyer intent.

Brand Exclusions and Cannibalization Prevention

Brand exclusion is the single most financially important PMax configuration action that New York ecommerce brands consistently skip. Without brand exclusions, Performance Max bids on the brand's own name in Search, taking traffic away from a dedicated branded Search campaign at CPCs two to three times higher than necessary. This cannibalization typically runs at 10% to 25% of total PMax campaign budget.

The three-list brand exclusion approach:

Own brand list. Contains the brand name and any sub-brand or product line names. Applied to every PMax campaign in the account. Branded search traffic routes to a dedicated branded Search campaign at controlled CPCs.

Direct competitor list. Contains five to fifteen competitor brand names where bidding is not a strategic priority. PMax will not bid on queries that include these names.

Category protection list. For accounts where very generic queries produce low conversion rates, category-level exclusions at the campaign level through negative keywords prevent budget from going to queries like "free," "cheap," "discount," and "used" that attract non-purchasing traffic.

As of March 2025, campaign-level negative keywords in PMax support up to 10,000 entries per campaign. This is separate from account-level brand exclusion lists and provides query-level control beyond brand names.

The Hybrid Strategy: Performance Max Plus Standard Shopping

The 2026 consensus among practitioners managing large ecommerce portfolios is that the optimal structure is not PMax alone. It is PMax running alongside Standard Shopping with explicitly defined roles for each.

Why Standard Shopping still belongs in the account:

Standard Shopping provides the search terms report. You can see exactly which queries triggered your ads and at what cost. PMax does not offer this transparency at the same granularity. Standard Shopping also allows negative keyword management at the campaign level and provides product-level bid control that PMax distributes algorithmically.

The ad rank priority rule. Google updated campaign priority logic in late 2024, moving from automatic PMax prioritization to an ad rank model. When a user's search exactly matches an exact match keyword in a Standard Search or Shopping campaign, that campaign takes priority over PMax in the same auction. This means well-structured Standard campaigns and PMax can coexist without the internal cannibalization that plagued early PMax setups.

The recommended hybrid structure for New York ecommerce:

Standard Shopping campaigns cover top-revenue SKUs, new product launches, and high-margin products where direct bid control justifies the management overhead. These campaigns serve known-intent, in-market buyers searching for specific products or categories.

Performance Max campaigns handle full-funnel discovery across Search, YouTube, Display, Gmail, Discover, and Maps for the broader catalog. PMax excels at reaching buyers earlier in their decision process and across surfaces where Standard Shopping cannot appear.

A separate branded Search campaign captures all brand-name queries at controlled CPCs, protected from PMax by brand exclusion lists.

The August 2026 Google Bidding Update: What Ecommerce Advertisers Must Do Now

Google is rolling out a bidding behavior change on August 17, 2026, that directly affects Performance Max campaigns. Campaigns that have been consistently delivering conversions at CPAs or ROAS levels that diverge significantly from their stated targets will be pulled back toward those stated targets.

Practically: if a PMax campaign has been set to a 5x Target ROAS but has been quietly delivering 7x because budget constraints were doing the optimization work, that gap will close starting August 17. The algorithm will steer more aggressively toward the 5x stated target.

Immediate actions before August 17, 2026:

Review every PMax campaign's actual delivered ROAS or CPA against its stated target. Use the Bid Target Adjustment Tool Google is releasing from July 6, 2026, to identify and correct gaps before the algorithm forces alignment automatically.

If actual delivered ROAS has been meaningfully above the Target ROAS, consider raising the target to preserve current performance volume rather than watching spend restrict as the algorithm corrects downward.

Avoid making major bid or budget changes within three weeks of August 17 to prevent triggering a new learning phase right as the bidding mechanics shift.

Performance Max and New York's Ecommerce Auction

New York's Shopping CPCs have climbed 26% over three years as more advertisers compete on the same product and category queries. In this elevated-CPC environment, every structural efficiency in a PMax account saves more real dollars per day than the same improvement would save in a lower-competition market.

Feed title quality that improves quality score and reduces CPCs by 30% to 50% produces larger absolute savings in New York. Brand exclusions that prevent 15% of budget from going to branded queries at inflated CPCs save more in dollar terms. Asset groups so well-structured that the algorithm finds high-intent buyers efficiently rather than exploring low-quality inventory matter more when each impression costs more.

This is why PMax management in New York requires more active structural discipline than the same budget in a mid-tier market, and why the gap between well-managed and default-setting PMax campaigns is wider here than anywhere else.

What Seller Splash Clients Say About Performance Max Management

"We had one PMax campaign covering everything with a single asset group. Seller Splash rebuilt it into six asset groups by product category with custom labels for margin segmentation. ROAS improved 40% in the first eight weeks and we could see for the first time which categories were actually driving the return."

Shopify DTC brand, New York, apparel category

"The brand exclusion setup was the first thing they addressed. We were spending 18% of our PMax budget bidding on our own brand name at CPCs three times what our branded Search campaign pays. Fixed in the first week, budget redirected to prospecting."

WooCommerce store, health category, USA

"The August 2026 bidding update would have hurt us badly if Seller Splash hadn't audited our Target ROAS settings against actual delivered performance. We adjusted three campaigns before the July cutoff and avoided the restriction that would have cut our impression volume."

Shopify Plus brand, New York, home goods

"We had never used search themes before. Seller Splash filled all 50 slots per asset group with product-specific queries. The algorithm found buyers we had completely missed in the prior account structure, in new geographic areas and at times of day that never appeared in our Standard Shopping data."

DTC brand, specialty food, New York

Why Choose Seller Splash for Performance Max in New York

Seller Splash is a New York ecommerce performance marketing agency founded by Shlomie Spielman. Performance Max, Standard Shopping, branded Search, Google Ads, Meta Ads, TikTok Ads, and Microsoft Advertising are managed as one connected paid media system for Shopify, WooCommerce, BigCommerce, and Magento brands across the USA, UK, UAE, and Australia.

Standard Performance Max deliverables for every engagement:

Feed audit for margin segmentation through custom labels, product title optimization for search query matching, April 2026 image resolution compliance verification, and video link attribute setup. Asset group structure by product category with three to seven groups per campaign, each with specific search themes, tailored creative, and relevant audience signal layers. Brand exclusion lists covering own brand, direct competitors, and category protection terms. Campaign-level negative keywords removing low-intent queries. Hybrid structure with Standard Shopping covering top-revenue SKUs and new launches. Monthly asset performance review and rotation of Low-scored assets. Audience exclusions removing recent purchasers from prospecting campaigns. High-value new customer mode activation through Customer Match upload. August 2026 bidding update review completed before the July 6 Bid Target Adjustment Tool deadline.

Across all managed Google Ads accounts, Seller Splash delivers 13.8x ROAS. Across all channels including Meta, TikTok, and Walmart: 12x blended ROAS, $2.4 million in gross sales, 18,200 orders.

Full case studies at sellersplash.com/case-studies. Full service scope at sellersplash.com/services.

For related reading: the Google Ads agency New York ecommerce guide covers the full Google Ads ecosystem including campaign type sequencing. The Shopify Google Ads New York guide covers Shopify-specific PMax setup including Customer Match from Shopify data. The Google Shopping agency New York guide covers Merchant Center feed management and Shopping campaign architecture.

No long-term contracts. Month-to-month because results keep clients, not paperwork.

A free Performance Max audit from Seller Splash identifies exactly which structural issues are limiting performance and what the correct fix sequence looks like before any engagement begins.

Conclusion

Performance Max in 2026 is Google's most powerful ecommerce campaign type and the one that produces the most wasted budget when run without deliberate structure. The 35% to 55% ROAS gap between well-architected accounts and single-campaign defaults is not explained by budget size. It is explained by asset group structure, feed quality, search theme specificity, brand exclusion completeness, and the discipline of active weekly management.

For New York ecommerce brands, where Shopping CPCs are above national averages and every structural efficiency saves more real money per day, getting PMax right is more financially urgent than in most markets. The August 2026 bidding update adds a deadline dimension: accounts with Target ROAS or CPA settings misaligned with actual delivered performance need to be corrected before August 17.

If your Performance Max campaigns are spending without the ROAS their budget should support, the structural layer is almost always where the problem sits. Reach out to Seller Splash for a free Performance Max audit and the team will show you specifically what is limiting performance and what fixing it involves.

Frequently Asked Questions