How to read your ad campaign report: which number is lying

The monthly report lands in your inbox. Impressions are up, clicks are up, even the "conversions" line has a green arrow next to it. And the till shows the same number as last month — or less. Someone is lying: either the report, or what you're actually seeing in sales.

Nobody really is. The biggest, most eye-catching numbers in a report are usually the ones that tell you the least about money. Here's which line to trust, which to skip past, and what to ask instead.

Impressions and clicks prove delivery, not sales

Every report opens with two numbers: how many people saw the ad, how many clicked it. They confirm the ad system did its job — the ad reached the intended audience, and the copy and image earned enough attention for a click. Useful to know. That's where the usefulness stops.

The problem is both numbers grow almost automatically with budget. Spend more, more people see it, more people click — whether or not the campaign itself got any better. So "impressions up 40%" proves nothing on its own; the real question is whether spend went up the same 40%. If it did, what grew was cost, not results.

What to do this week: put the "spend" column right next to impressions and clicks, and divide one by the other. If cost per click actually went down, that's real progress. If only the count went up, that's the budget talking, not the campaign.

The word "conversion" doesn't always mean a sale

The most confusing word in any ad dashboard is "conversion", because its meaning isn't set by the platform — it's set by whoever configured the campaign. Google Ads itself lists what can count as one: a website purchase, a newsletter signup, a button click, or another on-site action (Google Ads Help). So "134 conversions" on your dashboard doesn't necessarily mean 134 orders — it might mean 134 people looked at a contact page.

Nobody sets this up to mislead you. Building a campaign means answering "which action do we track", and the one that's easiest to measure technically — a page load, a tap on a form — often gets picked over the actual sale. The setting gets forgotten over time, and the report keeps calling that number a "conversion" as though it still means what it meant on day one.

How to check it yourself

• Open the "conversion actions" list in the campaign settings — each one carries its own specific definition.
• Ask your agency or team directly: how many of that number are real orders, and how many are just page views.
• If the answer isn't immediate, that's the actual signal — nobody has asked this in a while.

The number that matters: what each customer actually costs

After impressions and clicks, the first number worth reading is cost per acquisition (CPA). The definition is plain: it's the total cost of converting someone, not the budget spent as such (Unbounce glossary) — "if you're spending more to acquire a customer than they're worth, you're doing marketing wrong."

Working it out doesn't take a tool: divide spend by the number of real orders, not by the "conversions" the report names. The gap can be large — if only 12 of those 134 "conversions" were actual orders, the real CPA can run several times higher than the figure in the report.

Then weigh that number against what an average order actually earns you. If a product's margin is 30%, for example, every manat spent on ads needs to bring back roughly 3–3.5 manat in revenue to break even — a plain division, but one that almost never appears directly in a standard report, because it needs your margin data, not just the ad account's own numbers.

When two channels both take credit for the same sale

A customer sees the ad on Instagram first, searches your brand name on Google three days later, finds you and buys. Both dashboards count that sale as their own — each platform only measures what it can see, and has no idea what the other one did. Line the two reports up side by side and the combined total can end up higher than your actual number of sales.

The fix isn't checking both dashboards more often — it's pulling every source (ad accounts, site analytics, CRM) into one place and reconciling them against the real sales count. Doing that by hand every month costs hours and leaves plenty of room for error.

One month's number isn't enough to decide on

In the first few weeks of a new campaign, the ad system is still "learning" — testing who it's more efficient to show the ad to, and spend is usually unstable during that stretch. Judging "this isn't working" off a single month's result is often the wrong call, because that same month the campaign was still in its learning phase.

The better approach is to track CPA over three consecutive months, not one. If the number keeps dropping month over month, the campaign is maturing. If it holds steady or climbs, it's worth asking why before cutting the budget — has the audience shifted, has a competitor entered the market, or has seasonal demand simply dropped.

One more check needs no tool at all: ask new customers how they found you — on the phone, in the order form, or at checkout. The gap between that answer and the number in the ad dashboard is often obvious after the very first question.

A big number isn't automatically good news — the question is what it actually leads to.

A short rule for reading your next report

Next time a report lands, read it in this order: spend first, then the real order count (not whatever the agency labelled "conversions"), then the ratio between the two. Impressions, clicks and CTR are useful, but only to explain those three numbers — on their own they don't decide anything.

If pulling that together by hand every month isn't how you want to spend an afternoon, our reporting automation service merges ad, site and CRM data into one dashboard and shows the real result — spend, orders, revenue ratio — the same way every time, so you spend your time deciding instead of hunting for the number.