Most ecommerce ad problems are not dramatic. Nobody wakes up to a campaign that lost everything overnight. What happens instead is quieter: cost per acquisition drifts up over ten days, an ad account pauses on a Friday because a card expired, or a campaign keeps spending behind a product that sold out on Tuesday. By the time the weekly report lands, the money is gone.
This guide covers five warning signs worth watching every week, why each one is easy to miss, and what to check when it shows up. Where it helps, we use SouqMetrics, an AI-powered ecommerce intelligence system built for brand teams, operators, growth marketers and agencies, as the example of how automated monitoring handles the same signals.
Why these problems get caught late
A lack of data is rarely the root cause. An online business today has more of it than ever. The trouble is that the pieces live in different places: sales in the store, spend in two or three ad platforms, behavior in analytics, margins in a spreadsheet.
Reading them together means jumping between tabs and stitching numbers by hand. Most teams do that once a week, at best. Anything that changes between two reports is invisible until the next one. SouqMetrics describes its own goal in exactly those terms: not giving operators more data, but reducing the time between something changing and somebody realizing it matters.
Sign 1: CPA moves above your target
Cost per acquisition is the price you pay for each new order or customer from ads. Every business has a number above which a sale stops being worth it. Write that number down. A target that lives only in someone's head is not a target.
What to check: compare CPA by campaign, not just the account average. One campaign can drag the whole account above target while the others look fine. Then check whether the rise comes from a higher cost per click or a lower conversion rate: the first points at the auction and your creative, the second at the landing page, pricing or stock.
A system like SouqMetrics is designed to flag the moment CPA crosses the target you set, rather than waiting for you to notice it in a report.
Sign 2: ROAS slips below an acceptable floor
Return on ad spend tells you how much revenue comes back for every unit of money spent. It is the most quoted ecommerce metric and one of the easiest to misread, because a single large order can hide a bad week.
What to check: set a floor, then look at ROAS over a rolling window rather than day by day. If the rolling number is falling, look at what changed in the same period: a new audience, a creative swap, a price change, a competitor promotion, or a shift in which products are selling. ROAS rarely drops on its own; something upstream moved first.
Sign 3: an ad account stops delivering
This one costs the most for the least reason. A payment fails, a card expires or a policy flag pauses the account, and delivery drops to zero. Nothing looks wrong in a revenue chart until you notice the chart has been flat for three days.
What to check: look at daily spend per platform, not just total spend. A platform sitting at zero when it normally spends every day is the signal. SouqMetrics lists this case specifically among the things its intelligence engine surfaces: an advertising account that stops delivering because of an account or payment issue.
Sign 4: spend keeps running behind an out-of-stock product
Ads and inventory are usually managed by different people in different tools. So a product sells out in the store while the campaign promoting it keeps spending, sending paid traffic to a page that cannot convert.
What to check: match your top-spending campaigns against current stock at least twice a week. If a product is out of stock or close to it, pause or redirect the campaign. Monitoring that connects store data with ad data, which is the core of what SouqMetrics does, is built to catch this join that no single platform shows on its own.
Sign 5: sales look healthy but depend on returning customers
This is the hardest one to see, because the top-line number looks good. Orders are coming in and ROAS holds up. But when you split the orders, a growing share comes from people who already bought before. The ads are getting credit for customers you would have kept anyway, and new-customer acquisition is quietly stalling.
What to check: track new versus returning customers as a share of revenue, campaign by campaign. Watch the trend, not the absolute split. A campaign whose returning share climbs month after month is a campaign whose true acquisition cost is rising. SouqMetrics reads new versus returning customers, customer value and repurchase behavior as part of the same signal set, and flags campaigns that appear to generate sales but lean increasingly on returning buyers.
A weekly routine in four checks
Start with thresholds: decide your CPA target and ROAS floor, and write both into whatever tool you use. Next, check spend per platform per day and confirm every account that should be live is live. Then match top campaigns against stock. Finally, look at the new versus returning split for your biggest spenders.
Done by hand, this means four or five tools and a spreadsheet. Done with a system that connects the store, the ad platforms, attribution and customer behavior, it becomes a short list of alerts with a suggested next step. Either way, the goal is the same: find out today what would otherwise wait for next week's report.
Explore how SouqMetrics monitors these signals across your store and ad accounts at souqmetrics.co.
Souqmetrics
