ROAS is a vanity metric.
There, I said it. Again.
Most agencies will show you a 10x ROAS and start mentally polishing the case study. Maybe they will add a fire emoji to the Slack update. Maybe someone will say “performance is strong” in a QBR while everyone nods politely.
But if your COGS, shipping, returns, discounts and ad spend mean you are losing money on every order, that ROAS is not performance.
It is just a make-believe number.
It is easy to look like a hero in a Google Ads dashboard. It is much harder to be a hero on a P&L statement.
That is the bit ecommerce PPC needs to get more honest about. A Google Ads account can hit its ROAS target while quietly pushing low-margin products, overspending on brand terms you would probably have won anyway, ignoring rising new customer acquisition costs and letting one or two hero SKUs cover for a lot of rubbish underneath.
The account looks fine at the top level. The business, not so much.
JudeLuxe was recently named Best Small PPC Agency at the European Search Awards 2026, but this is not an award victory lap. Nobody needs another “we are delighted to announce” article. This is about the operating model more ecommerce PPC teams need before they scale spend.
Because high ROAS can still mean bad business.
And “good enough” is usually where the rot starts.
If we stopped when things were good enough, we would not have toast.
Some maniac baked bread and thought:
“You know what this needs? More fire.”
Toast was born.
Unnecessary. Overcooked. Glorious. And now? Non-negotiable.
That is the point. Bread was fine. Toast was better.
Most e-commerce PPC accounts are bread. They work. They report. They produce revenue. They look acceptable in a dashboard.
But if you want profit, control and scale, acceptable is not enough. You need to apply a bit more fire.
The problem with blended ROAS
Blended ROAS is not useless. It is useful in the same way a bathroom scale is useful.
It gives you a number. It does not tell you whether you are healthy.
For ecommerce teams, blended ROAS can be a quick account health check. It tells you how much revenue came back for the media spend that went out. That is handy. The problem starts when it becomes the main decision-making metric.
The majority of ecommerce businesses do not sell one product with one margin, one return rate, one fulfilment cost and one level of stock. They sell a catalogue. That catalogue usually contains margin winners, margin drains, clearance products, seasonal lines, bestsellers, awkward sizes, slow movers, bundles, upsell products and items that look great in platform revenue but terrible once the finance team gets involved.
When all of that gets flattened into one ROAS target, the important stuff disappears.
A 6x ROAS on a low-margin product may be worse for the business than a 3x ROAS on a high-margin product. A campaign that looks inefficient may be doing the hard work of new customer acquisition. A campaign that looks fantastic may just be harvesting brand demand that was already coming your way.
This is why ecommerce PPC teams should treat blended ROAS as a dashboard metric, not a steering wheel.
It can tell you something needs a closer look. It should not tell you what to scale.
The five inputs ecommerce PPC teams need before scaling
Before you increase budget, rebuild campaigns or start changing targets, you need a clearer view of what the business actually wants to sell.
Not what Google finds easiest to sell.
Not what makes the dashboard look shiny.
What the business actually wants to sell.
For most ecommerce accounts, five inputs matter more than another round of surface-level “optimisation”.
| Input | Why it matters |
| Contribution margin | ROAS is not profit. Margin tells you what each sale is actually worth after product economics are considered. |
| Stock position | There is no glory in aggressively pushing a product that is about to sell out, has a broken size curve or cannot be fulfilled efficiently. |
| Feed quality | Google can only optimise from the product data it receives. Weak titles, poor categorisation and missing attributes limit performance before the auction even starts. |
| SKU-level performance | Average performance hides which products deserve more investment, less investment or a completely different role in the account. |
| Wasted spend | Search terms, product groups and campaigns that consume budget without commercial value need to be controlled before scale. |
If you do not have those inputs, you are not really running profit-led PPC.
You are running Google Ads and hoping the blended number behaves.
Start with product segmentation, not campaign structure
A lot of PPC teams try to fix commercial problems with account structure.
Performance drops, so they rebuild campaigns. Performance Max looks messy, so they split asset groups. Shopping spend looks inefficient, so they change targets. Someone says “we need more control” and suddenly everyone is knee-deep in naming conventions.
Sometimes structure is the issue. Often, it is not.
Often, the real issue is that the account has not been given a useful view of the catalogue.
A profit-led ecommerce PPC account should start with product segmentation. Which products are high-margin? Which are stock constrained? Which are strategic for acquisition? Which are seasonal? Which are clearance-led? Which should not be getting meaningful spend at all?
Once products are grouped by commercial role, PPC decisions get much cleaner.
You can set different targets for different product groups. You can protect spend for strategic categories. You can stop poor-margin products from looking attractive just because they produce a nice revenue multiplier.
This does not need to be over-engineered. In many accounts, a simple product scoring model is enough to change the conversation.
Score products based on margin, stock, conversion rate, revenue, return rate and strategic value. Then use that score to inform campaign priority, budget allocation and bidding expectations.
That is not glamorous. It will not get as many LinkedIn likes as a screenshot of a 14x ROAS account.
But it is much more useful.
Fix the feed before blaming the algorithm
Google Ads automation gets blamed for a lot of things that start much earlier.
If the feed is weak, the account is already working with bad inputs.
For ecommerce PPC, feed quality is not an admin task. It is a performance lever. Product titles, descriptions, categories, custom labels, attributes, availability and pricing all influence how products are matched, understood and prioritised.
If high-priority products are not labelled correctly, if poor-margin products are treated the same as profitable products, or if seasonal products are not separated from evergreen lines, the bidding system has no real commercial context.
It may optimise efficiently against the wrong goal.
That is the uncomfortable truth. Automation is not magic. It is very good at following signals. If the signals are commercially lazy, the output will be commercially lazy too.
Before changing campaign structure, ask a simpler question:
Does the feed tell Google what the business actually cares about?
If the answer is no, your next optimisation task probably is not inside the Google Ads interface.
Use wasted spend analysis as a scaling gate
Scaling an account before understanding wasted spend is like pouring water into a bucket before checking for holes.
More budget does not fix leakage. It funds it.
A proper wasted spend review should look at search terms, products, categories, campaigns, devices, locations and time periods. The aim is not to cut anything that looks inefficient in isolation. That is how you accidentally kill useful demand.
The aim is to understand where spend has no clear commercial role.
A product may have a poor direct return but support new customer acquisition. Another product may have a strong ROAS but weak margin and high returns. A search term may look expensive but sit close to a valuable purchase path.
Context matters.
The key is to separate useful inefficiency from waste.
Useful inefficiency supports a known business goal. Waste just eats budget and asks for seconds.
How to explain profit-led PPC to founders and CMOs
One reason blended ROAS survives is that it is easy to present.
One number. One slide. Everyone can understand it before the coffee goes cold.
Profit-led PPC requires a more detailed conversation, but it does not need to become complicated. The simplest way to explain the shift is this:
ROAS tells us what revenue came back. Profit-led PPC asks whether that revenue was worth buying.
That sentence changes the discussion.
Instead of arguing about whether the account hit a target, the team can talk about which products should receive spend, which categories deserve protection, where Google needs better inputs and what growth actually means for the business.
Reporting needs to change too. A useful ecommerce PPC report should not only include spend, revenue and ROAS. It should show margin-aware performance, product group performance, wasted spend, feed issues, stock constraints and the actions being taken as a result.
The goal is not to drown stakeholders in data.
The goal is to show the commercial logic behind the PPC decisions.
A simple audit checklist
If you want to move an ecommerce PPC account away from blended ROAS worship, start with a practical audit.
Ask these questions before scaling spend:
- Which products generate the most contribution, not just the most revenue?
- Which products spend heavily but produce weak commercial value?
- Which high-margin products are underfunded?
- Which products have stock, fulfilment or return-rate issues that should affect bidding?
- Are custom labels being used to reflect margin, priority, seasonality and stock position?
- Do product titles and attributes match how customers actually search?
- Which search terms or product groups are consuming budget without a clear role?
- Are different product groups being held to different commercial expectations?
- Can the founder, CMO or ecommerce lead see why budget is being moved?
If you cannot answer those questions, the account probably is not ready for aggressive scale.
It might be ready for more diagnosis. It might be ready for better feed work. It might be ready for cleaner product segmentation.
But “increase budget and hope” is not a strategy.
The takeaway
Blended ROAS is not the enemy. Lazy interpretation is.
ROAS can tell you part of the story, but it cannot tell you whether your paid search activity is creating the right kind of growth.
For modern ecommerce PPC, the best accounts will be the ones with the clearest commercial inputs. Margin. Stock. Feed quality. SKU-level performance. Wasted spend. Contribution. New customer economics.
Not just a big multiplier in a Google Ads dashboard.
Because the goal is not to win the screenshot.
The goal is to make the business more profitable.
