The 5 metrics that actually matter in your ad account

You open Ads Manager on a Tuesday morning with your coffee, and it presents you with roughly forty columns of numbers. Reach. Impressions. CPM. Post engagements. Three-second video plays. ThruPlays. Estimated ad recall lift, which sounds important and is not.

Somewhere in there is the answer to the only question you actually have — is this working? — and the dashboard has done a magnificent job of hiding it. So you scroll sideways, decide you'll look properly on the weekend, and close the tab.

Here's the short version: five numbers tell you nearly everything. The rest are diagnostics for people managing six-figure budgets, or vanity metrics dressed up as insight. Learn the five and a weekly check takes about four minutes.

The only question a dashboard should answer

Before any metric means anything, you need to have told the platform what a "result" is. This is where most small accounts quietly fall over. If your pixel isn't tracking enquiries, form submissions or purchases, then every number in the account is measuring attention rather than outcomes — and attention is cheap.

So: conversion tracking set up first, then metrics. If that part's shaky, the complete guide to Google Ads walks through conversion tracking properly, and the same principle applies on Meta.

1. Cost per result

The headline number. What you paid for each of the things you actually wanted — a lead, a booking, a sale. Not a click. Not a like. The thing that puts money in the till.

The trap is comparing yours to someone else's. A $9 lead sounds brilliant until it turns out those leads never answer the phone, and a $70 lead sounds appalling until you remember your average job is worth $4,000. The only benchmark that matters is your own maths: what a customer is worth, times how often you close, tells you what you can afford to pay.

Do that sum once and write the number down. It turns "is $52 a lead good?" from a feeling into an answer you can check.

2. Click-through rate

The percentage of people who saw the ad and clicked it. This is your creative's report card — it tells you whether the picture, the headline and the offer are landing with the people you're showing them to.

When cost per result climbs, click-through rate usually tells you why. A falling CTR means the ad has stopped being interesting. A healthy CTR with an expensive cost per result means the ad is doing its job and something after the click isn't — which brings us to the number that isn't in your ad account at all.

If your CTR is the problem, it's almost always the first line and the offer rather than the budget. We pulled that apart in how to write a Facebook ad that actually gets clicks.

3. Frequency

How many times the average person has seen your ad. It's the one metric that quietly ruins small local campaigns, because Adelaide is not a big audience. Target "Adelaide, 30–55, interested in home renovation" and you might be talking to 80,000 people — which sounds like plenty until you're spending daily and the same 80,000 keep getting served the same photo.

Once frequency creeps past about three or four on a cold audience, performance usually starts sagging: people have seen it, they've decided, and you're paying to annoy them. The fix is new creative, a wider audience, or a lower daily spend — not a bigger budget on the same tired ad.

Retargeting is the exception. Those audiences are small and warm by design, so a higher frequency is normal there — just cap it so you're not following people around for a fortnight. More on that in retargeting explained.

4. Conversion rate on the page

This one isn't in Ads Manager, and it's the reason a lot of campaigns get blamed for a problem they didn't cause. Of the people who clicked, how many did the thing?

If 200 people clicked and two enquired, your ad worked and your page didn't. No amount of audience tinkering fixes that. Look at where you're sending people: does the page say the same thing the ad promised, is there one obvious next step, and does it load quickly on a phone on mobile data in a carpark? That's usually the whole answer.

5. Return on ad spend — or its honest cousin

If you sell online, return on ad spend is straightforward: revenue divided by spend. Spend $500, make $2,000, that's a 4x. Watch it over months rather than days, because a single big order can make a bad week look like a triumph.

If you're a service business, the platform can't see your revenue, so build the honest version yourself. Track enquiries out of the ad account and into a spreadsheet: how many became quotes, how many became jobs, what those jobs were worth. It takes five minutes a week and it's the only number that settles the argument about whether ads are working.

The ones to stop looking at

Not useless, exactly — just not decision-making numbers for a small budget.

Metric	Why it's distracting	Look at instead
Reach & impressions	Measures how much you spent, not what you got	Cost per result
Likes & engagement	Cheap to buy, rarely correlated with enquiries	Click-through rate
CPM (cost per 1,000 views)	Goes up and down for reasons outside your control	Cost per result
Video plays	A three-second view is not interest	Conversion rate
Yesterday's numbers	Too small a sample to mean anything	The last 7–14 days

How often to actually look

Weekly. Maybe twice a week if you're itchy. Daily checking on a small budget is how good campaigns get killed at three days old — you see a quiet Tuesday, panic, change the targeting, and reset everything the platform had started to learn.

Give any change at least a week and a decent number of results before you judge it. On a modest local budget that can mean a fortnight, which feels slow and is still faster than guessing. We ran the numbers on what a small budget realistically buys in Meta ads on a $500/month budget.

Five numbers, once a week, four minutes. That's a functioning reporting habit, and it beats forty columns you never read.