A working ecommerce PPC strategy runs in a fixed order: work out your break even return before you spend anything, fix the product feed, protect branded search cheaply, win the high intent product searches, then scale with Shopping and Performance Max, add remarketing once you have traffic worth chasing, and review results by product margin tier rather than by campaign average. Most stores skip straight to step five, which is why their numbers look fine at the top and bad at the bottom.
The reason order matters is that each step makes the next one cheaper. Knowing your break even number tells you what a good result even is. A clean feed decides what you get matched to. Getting those two right before scaling means the scaling actually compounds instead of amplifying a mistake.
Here is the sequence, what to do at each stage, and how to know you are ready to move on.
What does the sequence look like?
Seven stages. You can run some in parallel once you are established, but if you are starting or rebuilding, run them in this order.
| Stage | What to do | You are ready for the next stage when |
|---|---|---|
| 1. Know your break even ROAS | Calculate the return you need just to cover product cost, shipping, fees and overhead. | You can state the number out loud, and you know it per margin tier, not just overall. |
| 2. Fix the feed | Rewrite titles, set product types, add GTINs, fix images, sync price and stock daily. | Merchant Center shows no critical errors and titles read like what buyers search. |
| 3. Protect branded search | Own branded terms in a small dedicated Search campaign with tight match types. | You hold a high share of your own brand searches at a low cost per click. |
| 4. Win high intent product searches | Search campaigns on exact product, model and category terms with buying intent. | These campaigns hit or beat your break even return consistently. |
| 5. Scale with Shopping and Performance Max | Run the feed hard across Shopping and PMax, split by margin tier. | Non branded return holds up when you strip brand traffic out. |
| 6. Add remarketing | Retarget cart abandoners and product viewers with real offers, capped frequency. | You have enough qualified traffic that lists rebuild within a short window. |
| 7. Review by margin tier | Report profit per product group, not blended revenue per campaign. | You can name which SKUs make money on ads and which never will. |
What this means for you: if you cannot tick the readiness column on a stage, moving to the next one will not fix it. It will just make the problem more expensive.
Why start with break even ROAS?
Because without it, you have no way to judge any number you see. A 4x return sounds good. On a product with a 20% margin it loses you money.
Break even return on ad spend is the point where the gross profit from a sale exactly covers what you paid to get it. Anything above it is profit. Anything below it is subsidy. Our guides on what ROAS is and break even ROAS walk through the math.
Calculate it per margin tier, not once for the whole store. A store selling both a 60% margin accessory and a 12% margin appliance has two completely different floors. Treating them as one number guarantees you overpay for one and underspend on the other.
Why is the feed the biggest lever?
In Shopping and Performance Max you do not pick keywords. Google reads your product feed and decides which searches your products match. Google’s own documentation on Shopping ads spells out that product data lives in Merchant Center and the campaign pulls from it. That makes the feed your targeting, not a technical formality.
A product with a vague title gets matched badly or not at all. No bid fixes that, because you were never in the auction. Fixing the title puts the product into auctions it was locked out of, which is a much bigger change than any bid adjustment.
The fields doing the most work are title, product type, GTIN, image, price and availability. We cover how to write titles that get matched, and the common mistakes on each field, in our Google Shopping ads guide.
How should you handle branded search?
Own it, keep it small, and keep it separate. Branded searches are cheap and convert well because the person already picked you. A small dedicated Search campaign with exact and phrase match will usually hold that traffic for very little money.
The mistake is letting branded traffic sit inside your scaling campaigns. It inflates their reported return, so you think a campaign is working when it is really collecting sales you had already earned. Use brand exclusions on Performance Max, keep brand in its own campaign, and judge everything else on non branded results only.
How do you win the high intent searches?
High intent means the search names a specific product, a model number, or a category with a buying word attached. Someone typing a model number is much closer to paying than someone typing a category name.
Build Search campaigns on those terms, and let the ad and landing page match the search exactly. If someone searches for a model number, send them to that product page, not the category page or the homepage.
Keep negatives tight from day one. Research terms, “free”, “cheap”, “used”, “repair” and “parts” quietly drain ecommerce budgets. Pull the search terms report weekly at first, then monthly.
When should you scale with Shopping and Performance Max?
After stages one to four are solid, not before. Both campaign types lean on Smart Bidding, which sets bids automatically using auction time signals, and automation amplifies whatever you give it. Give it a clean feed, working tracking and real conversion history, and it finds demand you would never have built campaigns for. Give it a broken feed and thin data, and it spends fast on nothing.
Split campaigns by margin tier when you scale. One campaign holding everything means your best sellers and your dead stock share a budget, and the winners soak it up while nothing else gets tested.
Our Performance Max guide covers the situations where it works, the ones where it burns money, and the control levers you still have.
When is remarketing worth adding?
Only once you have enough qualified traffic that your audience lists refill quickly. Remarketing to a tiny list is expensive per person and does very little.
When you do add it, be specific. Cart abandoners are a different audience from category browsers, and they deserve a different message. Cap frequency so you are not showing the same product to the same person twenty times, which annoys people and wastes money.
Remarketing multiplies existing demand. It does not create it. If your top of funnel is weak, remarketing will not save the account.
What are the biggest ecommerce PPC mistakes?
Four come up again and again, and they are all reporting problems rather than bidding problems.
- Optimizing on blended revenue instead of per product profit. A campaign can grow revenue while shrinking profit if the growth came from your thinnest margin items. Revenue is not the goal.
- Letting best sellers subsidize dead SKUs. Inside one campaign, a few winners can carry a hundred losers and the average still looks acceptable. Split by margin tier and the losers become visible.
- Ignoring stock levels. If your feed updates weekly, you are paying for clicks on products you cannot ship. Sync stock daily at minimum, more often during sales.
- Judging performance in too short a window. If your buying cycle is three weeks, a seven day report tells you almost nothing. Match your reporting window to how long people actually take to decide.
A fifth is broken conversion tracking. Automated bidding chases whatever you told it a conversion is, so if tracking double counts or fires on the wrong page, every decision after that rests on a wrong number.
How should you review results?
By margin tier, against the break even number for that tier, over a window that matches your buying cycle. Not by campaign average, and not by last week’s revenue.
A rhythm that works: weekly checks on feed health, budget pacing and search terms, monthly checks on return by margin tier, and quarterly decisions on which products should stop being advertised at all. That last one is the conversation nobody wants to have. Some products will never make money on paid traffic, and deciding that on purpose beats funding it by accident for a year.
Common questions
What is a good ROAS for ecommerce?
There is no single good number, because it depends entirely on your margins. A 3x return is excellent on a high margin product and a loss on a low margin one. Calculate your break even return first, then set your target above it.
Should ecommerce stores use Google Ads or Meta ads?
Most established stores use both, for different jobs. Google catches people already searching for the product. Meta creates demand from people who were not looking. Google traffic is usually warmer, and usually costs more per click.
How much should an ecommerce store spend on PPC?
Work backwards from margin, not forwards from a percentage rule. Decide how much gross profit a sale gives you, what return you need to clear your costs, and how many sales you can actually fulfill. The budget falls out of that.
How long before ecommerce PPC starts working?
Expect the first month to underperform while conversion data builds and automated bidding learns. Give it long enough to cover a few full buying cycles before you judge it, and fix tracking and feed problems before you start counting.
Do I need a separate campaign for each product?
No, and it usually creates more work than it is worth. Group by margin tier or product priority instead, so each group can carry a target that matches its economics.
Why is my ROAS good but my profit flat?
Usually one of three things. Branded traffic is inflating the reported return, your growth came from thin margin products, or your tracking is counting conversions that are not really incremental.
Want a second opinion on your setup?
If your ads are running but you cannot say which products actually make money, that is a reporting problem worth solving before you change a single bid. Our ecommerce advertising team can look at your feed, your campaign structure and your margin reporting and tell you where the leaks are. If you are still deciding whether to hire help at all, our guide to choosing an ecommerce PPC agency covers what to ask.
No pitch and no commitment. Get in touch and we will walk you through what we find.
