Ecommerce PPC agency judged on profit, not ROAS
Paid media for online stores: Google Shopping, Performance Max, Meta and retail media. Blended ROAS averages your best and worst products into one number. We manage spend by product margin, so budget goes to what actually makes money.
5.0
4.8
5.0What should an ecommerce PPC agency actually do?
An ecommerce PPC agency manages paid campaigns across Google Shopping, Performance Max, Meta and retail media for online stores. The work that matters most is product feed quality, campaign structure that separates products by margin, and measurement tied to profit per product rather than blended return on ad spend. It is our PPC service, shaped for ecommerce marketing.
Where ecommerce ad budgets go to waste
The problems we find most often when we audit ecommerce ad accounts, the usual cause, and what we change.
| Problem | What you see | Usual cause | What we change |
|---|---|---|---|
| Feed problems | Products not showing, or showing for the wrong searches | Weak titles, missing attributes, disapprovals | Titles, attributes and variants fixed in the feed |
| One big campaign | ROAS looks fine, profit does not | Every product in one Performance Max campaign | Products split by margin and role |
| Brand taking credit | Brand search shows the best return | Paying for shoppers who would come anyway | Brand excluded from PMax, pause tests on brand |
| Returns ignored | Revenue that later gets refunded | Bidding on gross revenue | Returns and cancellations fed back as adjustments |
| Wrong targets | Budget follows high-revenue, low-margin products | One ROAS target for everything | Targets set by product margin |
| Prospecting cut | Search slowly weakens after social is cut | Prospecting judged on last click | New-customer and assisted revenue measured |
Ecommerce PPC services, from feed to profit
Four areas of work. Feed first, structure second, bidding last. Most accounts we inherit are the other way round.
Feed management
Titles, attributes, images, variants and disapprovals fixed, so products show for the right searches.
Structure by margin
Products grouped by profit and role, so budget can push what earns and cap what does not.
Campaign management
Shopping, Performance Max, Search, Meta and retail media planned as one budget.
Profit measurement
Product costs, returns and fees brought into reporting and bidding, with regular incrementality tests.
Better landing pages lift every campaign. See ecommerce CRO.
Where ecommerce ad money actually goes
Ecommerce paid media is several surfaces with different jobs. Judging them all on one number is where most waste starts.
Shopping and Performance Max run on your product data. Titles, attributes, images and stock decide which searches you appear for. Paid social creates demand rather than capturing it, so judging it on last-click return makes it look worse than it is.
Retail media is often managed by a different person in a different spreadsheet, even though it competes for the same customer and the same margin. We plan all of it as one budget. Organic demand is covered by ecommerce SEO, and paid social in more depth on the ecommerce paid social page.
| Surface | What it does | How it should be judged |
|---|---|---|
| Google Shopping and PMax | Captures existing product demand | Profit per product group |
| Google Search (brand) | Defends traffic you might get for free | Incrementality: test before assuming it works |
| Google Search (non-brand) | Captures category and problem searches | New-customer cost, not blended ROAS |
| Meta (prospecting) | Creates demand that converts later | Assisted revenue and new-customer rate |
| Meta (retargeting) | Speeds up a decision already made | Incremental lift, not attributed return |
| Retail media | Wins placement where shoppers buy | Margin after retailer fees |
Brand search usually looks best and is often least incremental. Prospecting usually looks worst and is often doing the most work.
How our ecommerce PPC process works
Five steps. Restructuring an account resets learning, so results often dip for a few weeks before they improve. We say so up front.
- Weeks 1-2
Margin and feed audit
We load product costs, shipping, returns and fees, and review the feed against them. Until the account knows which products make money, bidding cannot send budget to them.
- Weeks 2-3
Fix the feed
Titles rewritten to match how people search, attributes filled, variants corrected, disapprovals cleared.
- Weeks 3-5
Restructure by margin
Products grouped into campaigns by profit rather than category, with brand kept separate.
- Weeks 4-8
Fix measurement
Server-side tracking, consent handling and returns fed back, so bidding learns from net revenue.
- Ongoing
Test and scale
Feed and creative tests run continuously, with brand and geo holdout tests each quarter to check what spend really adds.
What we test, and what it must not break
Each test has a main metric and a guardrail. Higher revenue is not a win if it comes from low-margin or high-return products.
Feed titles
What we test: Search terms, attributes and order in titles
- Primary metric
- Impressions on target searches
- Guardrail metric
- Click-through rate must hold
Product groups
What we test: Separate campaigns by margin
- Primary metric
- Profit after ad spend
- Guardrail metric
- Total revenue must hold
Bidding targets
What we test: Targets set by product margin
- Primary metric
- Profit after ad spend
- Guardrail metric
- New-customer share must not fall
Brand search
What we test: Reduced brand bids in test regions
- Primary metric
- Incremental revenue
- Guardrail metric
- Total brand orders must hold
Creative
What we test: Product, lifestyle and user-made content
- Primary metric
- New-customer orders
- Guardrail metric
- Return rate must not rise
Offers
What we test: Free shipping thresholds, bundles
- Primary metric
- Average order value
- Guardrail metric
- Margin per order must not fall
Two products, same ROAS, opposite results
The argument for this page, in arithmetic you can check. Both products return 4.0 on ad spend. One funds the business; one loses money on every order.
| Product A | Product B | |
|---|---|---|
| Revenue from ads | 10,000 | 10,000 |
| Ad spend | 2,500 | 2,500 |
| Blended ROAS | 4.0 | 4.0 |
| Cost of goods | 3,000 | 6,500 |
| Shipping and fulfillment | 800 | 1,400 |
| Returns | 400 (4%) | 1,800 (18%) |
| Payment and platform fees | 300 | 300 |
| Profit after ad spend | +3,000 | -2,500 |
Illustrative figures chosen to show the mechanism, not client data.
Who this works for, and who it does not
A good fit if your store
- Knows or can build product-level costs
- Sells through its own store, with or without marketplaces
- Spends enough for the account to learn
- Can share order and returns data
- Wants results measured in profit
- Can accept a short dip during restructuring
No product cost data
Everything here depends on knowing costs, shipping and returns by product. Without them, we are just another agency chasing ROAS.
Spend too low to learn
Below a certain budget, the account cannot gather enough data to test, and our fee becomes a large share of spend.
You need this quarter rescued
Restructuring makes results worse before they improve. If you need a number in six weeks, leave the structure alone.
Margins too thin
Some products cannot support paid acquisition at any efficiency. That is a pricing problem, not a PPC one.
How much does ecommerce PPC cost?
Three stages, and you can stop after any of them. Management is quoted from the audit, based on catalog, channels and markets.
Audit
Best forYou suspect the account has stalled and want to know why.
- Profit by product group
- Feed quality and disapproval review
- Structure and wasted spend review
- Tracking and attribution check
Managed program
Best forOne brand, one or two markets, a catalog you can describe in a sitting.
- Everything in the audit, refreshed quarterly
- Google and Meta management
- Ongoing feed work
- Profit reporting by product group
Embedded
Best forSeveral markets or brands, retail media in the mix, or paid is your main channel.
- Everything in the managed program
- Retail media and marketplace ads
- Budget planning across markets
- Quarterly incrementality tests
What pushes scope up
- Large or fast-changing catalogs
- Several markets, currencies or languages
- Retail media alongside Google and Meta
- No product cost data yet
What brings it down
- One market and a stable catalog
- Product costs available from your platform or ERP
- In-house creative for tests
- One ad account, not several overlapping ones
Request your ecommerce PPC audit
Tell us where to look. A senior strategist will reply within one business day to scope an audit covering:
- Profit by product group, not just ROAS
- Feed quality and disapprovals
- Campaign structure and brand handling
- Tracking, returns and attribution
- Where budget goes versus where profit is
- A ranked list of fixes
Ecommerce PPC, answered
What is ecommerce PPC?
Ecommerce PPC is paid advertising for online stores, mainly Google Shopping, Performance Max, Google Search, Meta and retail media. The most important parts are product feed quality, campaign structure by margin and measurement tied to profit per product.
Why report profit instead of ROAS?
Because two products with the same ROAS can have opposite effects on profit once cost of goods, shipping and returns are counted. The worked example on this page shows the arithmetic. ROAS is useful inside a campaign, but as a target it sends budget to revenue, not profit.
How much of ecommerce PPC is about the product feed?
A lot. Shopping and Performance Max match on your product data, not a keyword list. Titles, attributes, images and stock decide which searches you show for, so a feed problem looks exactly like a bidding problem.
Should we run brand search campaigns?
Test it rather than assume it. Brand search usually shows the best return because it captures people already looking for you, many of whom would arrive anyway. A pause test on part of your traffic shows what it really adds.
How do you manage Performance Max?
By controlling what goes in. We split products into campaigns and asset groups by margin, give it accurate conversion values instead of raw revenue, and exclude brand traffic so it cannot take credit for demand you already had.
How do you handle returns in the numbers?
Where the platforms allow, we feed returns and cancellations back as adjustments, so bidding learns from net revenue. In categories with high return rates, an account bidding on gross revenue will overspend on the products most likely to come back.
Will performance drop when you restructure?
Often, for a few weeks. Rebuilding campaigns resets the platforms’ learning. We flag this in the proposal, and if you cannot absorb a short dip right now, we will recommend waiting.
How much does ecommerce PPC management cost?
It depends on catalog size, channels and markets. Ad spend is separate and paid to the platforms. We start with a fixed-scope audit and quote management from it.