Google Ads management for ecommerce brands, run to your margin

Google Ads management for an ecommerce brand means running Shopping, Performance Max and search as one system built on a clean product feed, then separating branded demand you already own from new demand Google actually created. The job is judged on new customers and contribution margin, not the ROAS Google reports.

Why Google reports more revenue than Google created

Google counts the conversions its tag records, and a large share of those come from people who searched your brand name after seeing an Instagram ad, a creator video or an email. Performance Max makes this harder to see, because it blends Shopping, search, YouTube, Display, Discover, Gmail and Maps inventory into one campaign and, unless told otherwise, is free to serve on your branded searches. The result can be a campaign that looks brilliant inside Google Ads and adds very little the business would not have had anyway.

None of this makes Google a bad channel. For most ecommerce brands it is the channel closest to purchase intent. The problem is judging it on its own report while costs rise. Triple Whale’s 2026 benchmark across more than 21,000 brands puts median Google Ads ROAS at 3.27, down 3.54% year over year, with median CPA up 9.96% to $28.14 (Triple Whale, August 2026). If the account cannot tell branded revenue from new revenue, it cannot tell whether that extra cost bought anything.

Signal	What Google reports	What we check it against
Branded search	Conversions from people typing your name	The share of branded traffic you would keep for free, and who is bidding on your terms
Performance Max	Blended conversion value across all of Google’s inventory	Channel and placement reports, brand exclusions, new customer share
Shopping	ROAS by product group	Contribution margin by product, stock status and return rate
YouTube and Demand Gen	View through and engaged view conversions	Blended new customer acquisition over the same weeks

What Google Ads management covers here

Five workstreams, owned by one person and run as one plan with the rest of your paid media.

Product feed and Merchant Center

Shopping and Performance Max are only as good as the feed behind them. Titles written for how people search rather than how the warehouse names things, complete attributes (color, size, material, gender, age group), GTINs where they exist, clean imagery, custom labels for margin tier, stock depth and seasonality, and supplemental feeds for anything the store platform cannot export. Disapproved and limited products get fixed weekly, not discovered at the end of a quarter.

Shopping and Performance Max structure

Campaigns are split by what actually differs in the economics: margin tier, hero products versus long tail, new launches versus proven sellers. Brand exclusions keep Performance Max off your own name so it has to find new demand. Asset groups are built around real audiences and real creative, briefed through the performance creative testing system, rather than images Google generates on its own. Where Standard Shopping gives more control over a product set, it stays.

Search: brand, non-brand and competitors

Brand defense is sized, not assumed: what share of branded clicks you would keep for free, who bids on your terms (competitors, unauthorized resellers, affiliates) and what defending them is worth in margin. Non-brand and category terms get budget in proportion to the new customers they return. Competitor terms are tested with a kill threshold agreed before the first dollar.

YouTube and Demand Gen

Video earns budget when there is a way to measure it: a geo or time based holdout, or at minimum a read on blended new customer acquisition in the weeks it runs. It does not earn budget because it reports cheap views.

Stock, pricing and promotions

Budget should never pay for a click to a product you cannot ship. Jason runs an automation that checks the products in live ads against stock levels, a workflow he walked through publicly in Triple Whale’s Moby Deep Dives session, so sold out and low stock variants drop out of spend before the money goes. Promotions, price changes and launches are loaded into the account calendar before they happen, not after.

How budget decisions get made on Google

1. Separate brand from non-brand before anything else. Every report splits the two, because they answer different questions. Branded search measures demand you already created. Non-brand measures demand Google found for you.
2. Set a margin floor per product tier. Contribution margin per order after cost of goods, shipping, processing, discounts and returns, translated into the return each tier needs to clear. High margin products can carry a lower ROAS target than low margin ones, and the account is built so they can.
3. Reconcile weekly. Google’s reported revenue against store revenue and against new customer counts in your analytics or Triple Whale. When they drift apart, the gap gets explained before budget moves.
4. Move money on written criteria. Scale and kill thresholds agreed in advance, so a campaign is not paused for one slow Tuesday or scaled for one lucky weekend.
5. Judge Google inside the whole media plan. Google’s budget is set against Meta, TikTok and everything else on blended results. The channel earns its share. It does not defend it.

The same margin logic runs across every channel we manage. The full version, including how blended ROAS, MER and contribution margin relate, is on the paid media management page.

Is this a Google Ads agency?

Not in the usual sense, and the difference is worth being clear about. Most people searching for a Google Ads agency want an outcome: more profitable sales from Google without babysitting the account. Plaid Testing delivers that outcome through a different structure. Jason Lu is a fractional CMO who executes. He builds the plan, stays in the account, and puts specialists from his own team on the builds, the feed work and the creative, under his direction.

That matters on Google in particular, because the biggest decisions are not made inside Google Ads. Whether to defend your brand name, how much Performance Max should get relative to Meta, which products deserve budget given stock and margin: those are business decisions. In a standard account management setup they get made by whoever runs the account, measured on the account’s own report. Here they get made by someone accountable for the whole paid media number and for the margin underneath it.

Jason is a Meta Business Partner, presented “Meet Moby 2” on Triple Whale’s Customer Education Series and has spoken on panel at The Whalies. You work with him directly, from the audit onward.

Two ways to work together

Some brands want the channel run for them. Others have an in-house buyer who needs senior direction rather than replacement. Both work, and the choice usually comes down to who you already have.

	Done for you	Done with you
Who runs the account	Plaid Testing, with Jason owning strategy and his team on execution	Your in-house team, with Jason directing strategy and reviewing the account
Feed and Merchant Center	Built and maintained by us	Audited by us, fixed by your team against a written list
Reporting	Daily snapshot, weekly readout, monthly strategy session	Weekly review of your team’s numbers, monthly strategy session
Best fit	No dedicated paid search specialist in house	A capable buyer who needs a senior operator above them

Either way you own the Google Ads account, the Merchant Center, the tag and the data. Nothing is held back if you leave.

What the first 30 days look like

Week one is measurement, the same foundation described under tracking and attribution: confirming the Google tag and enhanced conversions fire once per order, that conversion actions count purchases rather than page views, and that new and returning customers can be separated. Week two is the feed: a full Merchant Center audit, disapprovals fixed, titles and attributes rewritten for the top revenue products, and custom labels for margin and stock. Weeks three and four are structure: brand and non-brand split cleanly, Performance Max brand exclusions in place, campaigns rebuilt around margin tiers, and the reporting cadence live.

By day 30 you should know what share of your Google revenue is branded, what non-brand is actually returning, and which products deserve more budget. Growth comes after that, because scaling a campaign that mostly buys your own brand name just buys the same customers at a higher price.

What this looks like in a real account

For a womens fashion brand, Google went from a supporting channel to a core one. From January to August 2026, against the same months of 2025, Google spend grew 399% to $311K while Facebook spend grew 56% to $829K, which took Google from roughly a tenth of the two platforms’ combined spend to more than a quarter. Over the same period, new customer acquisition cost across the account fell 21%, from $86 to $67, and new customer ROAS rose 58%, from 1.38x to 2.17x.

What this does not claim: those new customer figures are blended across every paid channel, not credited to Google alone, and no incrementality test isolated Google’s contribution. The point is narrower and more useful. Google’s budget grew roughly fivefold inside a plan judged on new customer cost, and that cost went down, not up. The full breakdown is in the womens fashion case study. Every figure traces to Triple Whale exports held on file, and brands are described by category because the numbers belong to them.

Who this fits, and who it does not

It fits ecommerce brands investing $50K or more a month across paid media, with Google as a meaningful share of it, most often in apparel, fashion and accessories, or wellness and supplements. It fits catalogs with real margin differences between products, brands whose Performance Max results look too good to be true, and founders who want to know what share of their Google revenue they would have had anyway.

It does not fit brands spending under roughly $50K a month across all paid media, where the honest advice is a simple, well fed Shopping or Performance Max setup and the rest of the money into offer and creative. It does not fit B2B lead generation, local services, or anyone who wants their current ROAS defended rather than examined.

Get your Google numbers looked at first

The first step is a free 30 minute growth audit, not a proposal. Bring access or exports and we will look at your brand versus non-brand split, your feed health and your Performance Max settings, then leave you with three specific fixes you can implement with or without us. Book your free growth audit. If Meta is the bigger line in your budget, Meta ads management covers how that channel is run, and the whole picture sits on the paid media page.

Common questions

How much should an ecommerce brand spend on Google Ads?

There is no universal share. It depends on how much search demand exists for your products and category, and how much of your current Google revenue is branded. As a working rule, non-brand Shopping and search earn budget on new customer cost against your margin floor, and brand spend is sized to what you would lose without it. The audit gives you that split on your own numbers.

Should we bid on our own brand name?

Usually some, rarely as much as the account currently does. Bid where competitors, resellers or affiliates are taking clicks you would otherwise get, and test pulling back where nobody is. The answer is a margin calculation rather than a best practice, and it gets revisited as the competitive picture changes.

Is Performance Max worth using for ecommerce?

Yes, with guardrails: brand exclusions so it cannot live on your own name, a clean feed, product segmentation by margin, and reporting that separates new customers. Without those, it tends to report strong results made largely of demand you already owned.

Do you only manage Google Ads?

No. Google is managed as one part of a paid media plan that usually includes Meta and sometimes TikTok or Snapchat, and budget moves between channels on blended results. If you want Google managed in isolation and judged only on Google’s own ROAS, this is probably not the right fit.