Google Ads Agency New York Ecommerce: How to Get Campaign Type, Structure, and Bidding Right

About the Author

Shlomie Spielman is the founder of Seller Splash, a New York ecommerce performance marketing agency. After managing Google Ads, Google Shopping, Performance Max, Meta Ads, TikTok Ads, and Amazon Sponsored campaigns for product brands across Shopify, WooCommerce, BigCommerce, and Magento, he built Seller Splash around one foundational truth: ecommerce Google Ads performance is determined by the structural layers beneath the campaigns, not the campaigns themselves. Seller Splash delivers 13.8x Google Ads ROAS across managed accounts. A New York Shopify brand achieved 9.37x ROAS within 30 days of a full account rebuild, generating $71,900 from $7,670 in spend.

When New York ecommerce brands go looking for a Google Ads agency in New York, they're usually focused on the wrong question. Most ask: which agency has the best results? The question that actually determines whether an engagement works is: does this agency understand which campaign types your specific ecommerce business needs, in what order, and why?

Google Ads is not a single tool. It's a platform with six distinct campaign types, each designed for different objectives, different audiences, and different stages of the customer journey. Running the wrong campaign type for your ecommerce store, or running the right ones in the wrong sequence, is one of the most expensive structural mistakes in paid media. And it's also one of the most common, even inside accounts spending serious money every month.

The average ecommerce Google Ads account sits at 2.87x ROAS. Accounts with properly selected campaign types, correct account structure, and margin-calibrated bidding regularly hit 5.17x. That gap isn't a function of budget. It's a function of how the account was built.

Why Google Ads for Ecommerce Is More Complex Than Most Businesses Expect

Most businesses understand PPC at a surface level. You create an ad, choose keywords, set a budget, and pay when someone clicks. That description is accurate for basic Search campaigns. For ecommerce Google Ads management, it describes maybe 20% of what actually drives performance.

Ecommerce brands running Google Ads are typically managing several campaign types simultaneously: Search campaigns for branded and non-branded high-intent queries, Google Shopping campaigns feeding from a Merchant Center product feed, Performance Max campaigns running across Google's full network, and often a Display or YouTube remarketing layer for buyers who showed intent but didn't convert. Each campaign type has different inputs, different optimization levers, and different relationships with Google's algorithm. Getting them to work together rather than compete against each other is the actual job.

The Campaign Type Decision Nobody Explains Properly

The most consequential decision in any ecommerce Google Ads account is which campaign types to run and in what order. This is also the decision most agencies either default on (everyone runs Performance Max because Google pushes it) or get wrong (launching PMax before the account has enough conversion data for the algorithm to learn from).

Here's how the campaign type decision actually breaks down for ecommerce.

Google Shopping and Standard Shopping campaigns are the foundation. They pull directly from your Merchant Center product feed and show product-specific ads to buyers actively searching for what you sell. Shopping campaigns provide search term visibility through the search terms report. They let you build conversion history on individual products and SKUs before handing them to more automated campaign types. They give you direct bid control over your most important products. For any ecommerce brand with a product catalog, Shopping campaigns are not optional and they should be running before PMax is launched.

Performance Max is Google's AI-driven campaign type that serves ads across Search, Shopping, Display, YouTube, Discover, Gmail, and Maps from a single campaign setup. When it works, it works impressively. The problem is that PMax requires conversion history and good audience signals to optimise toward the right outcomes. Launching PMax without either means the algorithm makes expensive guesses for four to six weeks before gathering enough data to perform. The accounts that get the most from PMax are the ones that built conversion history through Standard Shopping first, then introduced PMax once the algorithm had real data to learn from.

Search campaigns serve text ads to buyers typing specific queries. For ecommerce brands, Search campaigns are most valuable for branded queries, competitor conquest strategies, and high-intent category terms where Shopping doesn't fully capture buyer intent. They're less critical for product-specific searches, where Shopping ads generally outperform text ads on both click-through rate and conversion rate, but they fill real gaps in the overall account structure.

The ordering matters. A common mistake we see in New York ecommerce account audits is agencies launching Performance Max first, seeing the algorithm struggle through a long learning phase, and calling the campaign underperforming before it's had the structural foundation it needs. The correct sequence for most ecommerce accounts is Standard Shopping first, then layer Performance Max once the account has 30 to 50 conversions per month that the algorithm can reference.

Asset Group Architecture: The Performance Max Skill Most Agencies Get Wrong

Performance Max campaign performance depends heavily on how asset groups are structured inside the campaign. Most agencies create one asset group covering the entire product catalog. This produces poor results because the algorithm cannot distinguish between product categories, margin tiers, or buyer intents when all assets and products are pooled together.

How to Structure Asset Groups for Ecommerce PMax

Each asset group should contain a coherent combination of products, creative assets, and audience signals. The goal is to give Google's algorithm a tightly defined context for each segment so its targeting decisions reflect actual business priorities.

By product category: A fashion brand should have separate asset groups for footwear, outerwear, and accessories. Each gets category-specific headlines, descriptions, images, and videos. The algorithm learns which creative elements perform for which product type rather than averaging across everything.

By margin tier: High-margin products need their own asset group with aggressive performance signals. Low-margin products need a separate group with conservative guardrails. Mixing them in one asset group forces the algorithm to optimize toward a blended average that serves neither tier's economics correctly.

By audience temperature: Prospecting asset groups reach new buyers with broad discovery creative. Remarketing asset groups reach buyers who visited the site or added to cart with more specific, conversion-focused creative. Keeping these separate prevents the algorithm from allocating remarketing budget to prospecting goals and vice versa.

The 2026 data point worth noting here: accounts running 2020-era manual bidding strategies on legacy campaign types now underperform PMax accounts with correctly structured asset groups and server-side tracking by 40% to 80% across most ecommerce verticals. Asset group architecture is now the primary skill differentiator in ecommerce Google Ads management.

AI Max for Search: The September 2026 DSA Replacement

Dynamic Search Ads (DSAs), which automatically generated ads based on website content without requiring keyword lists, are being replaced by AI Max for Search campaigns by September 2026. AI Max extends Responsive Search Ad reach to additional relevant queries using Google's AI, covering the same intent territory that DSAs addressed while giving advertisers RSA-level control over headlines and descriptions.

For ecommerce brands currently running DSA campaigns, this transition requires action before September 2026. Any DSA campaign should be reviewed, and a migration plan to AI Max should be in place before Google forces the auto-upgrade. New York ecommerce accounts running DSAs without this awareness will have their campaigns auto-modified during a period where manual oversight of the transition would have produced better outcomes.

Account Structure Determines What the Algorithm Can Learn

Even within the correct campaign types, account structure determines whether Google's algorithm has enough data density to optimise effectively.

The most common structural mistake in ecommerce Google Ads accounts is over-segmentation. Creating eight Shopping campaigns to feel more in control actually produces eight campaigns each generating too few conversions to exit the learning phase efficiently. The algorithm needs data density. Five campaigns with 10 conversions each will always underperform one campaign with 50 because the learning signal is too diluted.

Useful structure for ecommerce follows business logic. High-margin products segmented from low-margin products because they can sustain different ROAS targets. New product launches isolated from the main catalog because they need conversion history before the algorithm can serve them effectively. Seasonal or promotional products with their own budget windows tied to specific time periods. Everything else in a catch-all campaign with conservative settings.

Campaign structure and product feed structure are also connected in ways most generalist agencies don't address. Custom labels in the product feed are how you tell Google's algorithm which products belong to which business logic group. Without custom labels marking margin tiers, bestseller status, and inventory levels, the algorithm makes structural decisions based on product category alone. That usually means serving whatever converts at the lowest cost, which is typically the thinnest-margin product in the catalog.

First-Party Data and Audience Signal Engineering

The quality of audience signals fed into Performance Max and Smart Bidding campaigns is now one of the primary performance differentiators in ecommerce Google Ads. As third-party cookie availability continues to decline, first-party data from the brand's own customer base has become more valuable than any platform-inferred behavioral signal.

Customer Match as a PMax Audience Signal

Customer Match allows ecommerce brands to upload hashed customer data (email addresses from Shopify or WooCommerce customer records) directly to Google Ads as a first-party audience. For Performance Max, Customer Match serves two functions:

Exclusion: Existing purchasers can be excluded from new customer acquisition campaigns to prevent spending on buyers who would repurchase organically, improving new customer ROAS.

Signal: The algorithm uses the Customer Match list to find new buyers who share behavioral and demographic characteristics with existing customers. This significantly reduces the time PMax spends in the learning phase because it starts with a high-quality reference profile rather than exploring from scratch.

Google's high-value new customer mode, available within Performance Max campaigns, allows bidding more aggressively for buyers matching the Customer Match profile. For brands where customer lifetime value justifies higher acquisition costs on the first order, this mode concentrates budget on the buyers most likely to become repeat customers.

Enhanced Conversions and Server-Side Tracking

Enhanced Conversions should be treated as a prerequisite, not an advanced option, for any ecommerce account running Smart Bidding in 2026. It uses hashed first-party data submitted at checkout to recover conversions that standard pixel tracking misses due to browser restrictions, iOS privacy changes, and ad blockers. Google's own data shows Enhanced Conversions recover 10% to 20% of conversions that would otherwise go unattributed. Accounts without it are training Smart Bidding on incomplete signals that understate actual campaign performance.

For the complete Google Shopping ads management framework including feed quality, Merchant Center setup, and campaign structure, the guide covers the full sequence that Seller Splash applies to every new ecommerce account.

Bidding Sequence: Why Order Matters More Than Target

Smart bidding in Google Ads learns from conversion signals. The bidding sequence needs to match the account's data maturity.

New campaigns start with Maximize Conversions or Manual CPC, not Target ROAS. Without conversion history, Target ROAS has no reference point. It guesses at bid amounts based on incomplete signals, which typically means either overpaying for low-intent traffic or restricting impressions so heavily the campaign can't gather the data it needs to improve.

Once a campaign has 30 to 50 conversions, move to Target ROAS with a target calibrated to the product segment's actual margin. Not an industry benchmark. Not where you'd like to be ideally. Your specific margin structure for that product group. A product with a 40% gross margin breaks even at 2.5x ROAS. Setting a Target ROAS of 4x for that product is correct. Setting a blanket 6x Target ROAS across all products because a competitor claims they hit 6x is how you starve high-margin products of impressions and budget while the algorithm restricts spend trying to hit a number that doesn't reflect the product's actual economics.

The break-even ROAS calculation is the foundation of every bidding decision. Our break-even ROAS guide walks through the exact formula for each product segment in your catalog.

The August 2026 Google Bidding Update: What Every Ecommerce Account Must Know

Google is rolling out a significant bidding behavior change on August 17, 2026 that directly affects every ecommerce account running Target ROAS or Target CPA.

The change: Google will pull campaign delivery toward the stated Target ROAS for accounts where actual delivered ROAS significantly diverges from the stated target. Campaigns that have been consistently delivering above or below their stated target will see delivery restricted toward that stated number.

For ecommerce brands, the practical implications are significant. If your account has a Target ROAS of 4x set but campaigns have been delivering at 8x (because the algorithm found efficient inventory while the stated target was set conservatively), delivery may be restricted or expanded toward 4x from August 17. Accounts that set Target ROAS from actual margin data rather than arbitrary aspirational numbers will be better positioned for this change.

The Bid Target Adjustment Tool, available from July 6, 2026, allows advertisers to identify and correct gaps between stated Target ROAS targets and actual delivered performance before the change takes effect. Any ecommerce brand working with a Google Ads agency in New York should confirm their agency has reviewed this tool for every managed account and corrected any significant divergences before August 17, 2026.

What New York's Google Ads Auction Means for Ecommerce Brands

Over 200,000 businesses operate across New York's five boroughs, and ecommerce brands compete in auction environments that are consistently more expensive than most US markets. New York's auction density pushes ecommerce category CPCs higher because the volume of advertisers bidding on the same product and category queries is greater here than almost anywhere else.

For ecommerce Google Ads accounts in New York, this elevated CPC environment makes two things more critical than they'd be in a lower-competition market.

Quality score has greater financial impact per dollar in New York. A quality score of 8 versus 5 on the same competitive keyword can reduce cost per click by 30% to 50%. In a market where your baseline CPC is already elevated, that reduction compounds significantly across the full scale of the account's spend. Quality score improves when ad copy closely matches search intent, when landing pages directly deliver on the ad's promise, and when historical click-through rates signal relevance. Most agencies treat quality score as a reporting metric. It's actually a cost management lever.

Impression share analysis is more actionable in New York than in lower-competition markets because the gap between "lost to budget" and "lost to rank" tells you specifically where to focus. Lost impression share due to rank on high-intent Shopping queries, while budget flows to broad Display placements, is a structural inversion that a campaign audit should catch on day one.

Dynamic Remarketing: Re-Engaging the 97% Who Don't Convert on the First Visit

Research consistently shows that 97% of website visitors leave without converting on their first visit. For New York ecommerce brands paying above-average CPCs to acquire that traffic, not having a structured remarketing layer is one of the most expensive structural gaps available.

How Dynamic Remarketing Works for Ecommerce

Dynamic remarketing serves product-specific ads to buyers based on their exact browsing behavior. A buyer who viewed a specific leather wallet for three minutes and then left sees an ad featuring that exact leather wallet, not a generic brand ad. A buyer who added sneakers to cart but abandoned checkout sees the sneakers with a message addressing the friction that likely caused the abandonment.

This personalization requires a Merchant Center product feed connected to the remarketing tag so Google can match browsed products to the correct ad creative dynamically. The product-level specificity consistently outperforms generic brand remarketing in both CTR and conversion rate because the ad is answering the exact question the buyer was already asking.

Segmenting Remarketing Audiences by Behavior

Effective ecommerce remarketing segments audiences by the specificity of their buying signal:

Product page viewers saw a product but did not add to cart. They need awareness-building creative showing product benefits and social proof.

Cart abandoners had declared purchase intent. They need conversion-focused creative addressing the most common friction points: shipping cost, return policy, or comparison with alternatives.

Past purchasers are the highest-lifetime-value audience. They need cross-sell and upsell creative based on their purchase history, not acquisition creative.

Running all three segments in the same remarketing campaign at the same bid treats fundamentally different buying signals as equivalent. Segmenting them allows budget to concentrate on cart abandoners, where conversion probability is highest, while maintaining lower-cost awareness campaigns for product page viewers.

Landing Page Testing as a Google Ads Performance Lever

The conversion chain in every Google Ads campaign has three links: the search query, the ad, and the landing page. Agencies that optimize the first two while leaving the third to the client are handing off the final conversion decision at the highest-cost moment in the buyer journey.

For ecommerce brands, landing page quality affects Google Ads performance through Quality Score directly. Google evaluates landing page relevance as an input to Ad Rank. A landing page that loads within two seconds on mobile, matches the specific product or category shown in the ad, displays the add-to-cart action without scrolling, and shows real social proof (review counts, bestseller badges) improves Quality Score which reduces CPC simultaneously with improving conversion rate.

The most impactful landing page change for most ecommerce Shopping campaigns is ensuring each asset group or product group links to the most specific relevant collection page, not the homepage. A Performance Max asset group for leather wallets should link to the leather wallet collection. A branded Search campaign should link to a brand story or hero product page. Every ad-to-landing-page mismatch both costs money and loses conversions.

Why Seller Splash Is the Google Ads Agency Built for New York Ecommerce

There's a version of this section on every agency website. This one is going to tell you specifically what's different about how Seller Splash runs Google Ads for ecommerce brands in New York, because the difference is operational and it's the part that actually determines ROAS.

Every new account engagement starts with a campaign type audit before any new campaigns are created or existing ones are changed. The audit establishes what campaign types are running, whether the sequencing is correct, what the account's conversion data density looks like per campaign, and whether Smart Bidding has been introduced before the data foundation existed to support it. In most new accounts from agencies who launched campaigns quickly without this foundation, the answer to that last question is yes. PMax was launched first, or Target ROAS was set before the account had enough conversions to reference, or both.

Feed management is part of every Google Ads engagement at Seller Splash because it has to be. For Shopping campaigns and Performance Max, the product feed is the most consequential variable in the account. Product title quality determines which search queries your Shopping ads appear for. GTIN accuracy determines whether Google can verify product details and improve competitive placement. Custom labels are how margin tiers, bestseller status, and promotional products get communicated to the algorithm so campaign structure can reflect business logic rather than product category defaults.

The Google Shopping ads management guide and the approach to Performance Max both start from feed quality, move to campaign structure, then to conversion tracking, and only then to bidding strategy. That sequence is not an arbitrary preference. It reflects the order in which each layer depends on the previous one. The 7 metrics that actually improve ROAS guide covers what to measure once campaigns are running correctly.

Seller Splash has delivered 13x ROAS for ecommerce clients by treating campaign type selection, feed quality, and conversion tracking as the structural foundation rather than afterthoughts. For New York ecommerce brands ready to find out whether their Google Ads account has the right foundation, a free account review from Seller Splash is where that conversation starts. The team identifies specifically which structural issues are limiting performance and what the correct sequence of fixes looks like.

You can review full results in the Seller Splash case studies and understand the complete service scope on the services page.

What Seller Splash Clients Say About Google Ads Management in New York

"Our Performance Max was one asset group covering everything. Seller Splash rebuilt it into separate asset groups by product category with dedicated creative and audience signals for each. Conversion rate on Shopping traffic improved significantly within six weeks and ROAS moved from 3.1x to 7.4x on the same total budget."

Shopify DTC brand, New York, apparel

"The campaign type audit in week one found that our Target ROAS had been set before we had 30 monthly conversions in any campaign. Smart Bidding was making expensive guesses. Seller Splash reset bidding to Maximize Conversions, let the campaigns accumulate data over four weeks, then introduced Target ROAS calibrated to actual product margins. ROAS improved 40% from there."

WooCommerce brand, New York, home goods

"Nobody had ever told us about the August 2026 bidding update. Our Target ROAS was set at 4x but we were actually delivering at 9x. Seller Splash used the Bid Target Adjustment Tool in July to correct the gap before August 17. We avoided the delivery disruption that several brands in our category experienced."

Shopify Plus brand, New York, specialty food

Conclusion

Choosing a Google Ads agency for your New York ecommerce brand comes down to one foundational question: does this agency understand campaign type selection, account structure, and bidding sequence well enough to build the right foundation before optimising on top of it? Most don't. Most launch Performance Max first because Google recommends it, set Target ROAS before the account has conversion data, and treat the product feed as a client responsibility rather than a campaign lever.

Seller Splash is built around the discipline of getting the foundation right first. If your Google Ads account is running but not compounding, or if you've never confirmed whether your campaign type mix and bidding sequence match your actual ecommerce economics, reach out for a free account review. The team will tell you directly what's wrong and what fixing it involves.

Frequently Asked Questions