You’re throwing money at Facebook. Google’s eating your lunch. TikTok? Let’s not even talk about it.
If you’re running paid ads but you’re not sure they’re doing anything beyond making your notifications ping, this one’s for you. Running ads isn’t the same as running them well, and the only way to tell the difference is to measure the right things.
Measuring ad performance goes well beyond counting clicks. You have to zoom out, zoom in, and follow the money. Here are the metrics that matter, what they actually tell you, and what to do when they look wrong.
The short version
Track the numbers that pay the bills
- A metric without a goal is noise. The same number can be a win or a disaster depending on what you were trying to do.
- ROAS and CPA carry the money questions. Everything else is diagnostics telling you where the leak is.
- Cheap clicks aren’t the goal. A 50 cent click that never converts is still 50 cents gone.
- Low CPA with a 0% close rate is failure. Lead quality decides whether the number means anything.
- If you’re not testing, you’re guessing. Change one thing at a time or the data tells you nothing.
The metrics that matter, and what they actually tell you
New to ad analytics, it’s easy to drown. Impressions, reach, click-through rate, ROAS, CPA, CPM. It reads like someone tipped a Scrabble bag onto your screen. Here are the ones worth your attention.
ROAS, return on ad spend
What it means: how much revenue you earn for every dollar you spend on ads.
Why it matters: it’s the fastest read on whether your ads are profitable.
How to use it: a ROAS of 4 means you earned $4 for every $1 spent. What counts as good depends entirely on your margin, and the honest answer to “what’s a good ROAS” works through that maths. One trap worth naming: high ROAS with no volume isn’t a win. It usually means you’re underspending on something that works.
CPA, cost per acquisition
What it means: what it costs you to get a customer, lead or conversion.
Why it matters: it tells you how efficient your funnel really is.
How to use it: compare it against customer lifetime value. Spending $150 to win a customer worth $100 is a problem no amount of scaling will fix.
CTR, click-through rate
What it means: the percentage of people who saw your ad and clicked.
Why it matters: it’s the clearest signal that your creative, copy and targeting are landing.
How to use it: around 1% is the rough benchmark on most platforms. Below that, revisit your ad creative or your audience before you touch anything else.
CPC, cost per click
What it means: what you actually pay for each click.
Why it matters: a low CPC is only good news if those clicks convert.
How to use it: stop chasing cheap clicks. A 50 cent click that goes nowhere is still 50 cents you’ll never see again. Read CPC next to conversion rate or it means nothing.
Impressions and reach
What it means: impressions are how many times your ad was shown. Reach is how many people saw it.
Why it matters: they’re the right measures for awareness campaigns and the wrong ones for everything else.
How to use it: use them to understand how much of your audience you’re reaching. High impressions with low engagement is a messaging problem, not a budget problem.
Conversion rate
What it means: the percentage of clicks that became sales, leads or signups.
Why it matters: it shows whether your ad and your landing page are telling the same story.
How to use it: a low conversion rate points at targeting, messaging or the page itself. Start with why ads get clicks but no sales, then work on conversion rate optimisation properly.
Vanity metrics vs decision metrics
Some numbers make you feel good. Others tell you what to do next. Knowing which is which saves a lot of wasted meetings.
| Your goal | Feels important | Actually decides it |
|---|---|---|
| Sales | Clicks, impressions, likes | ROAS, CPA against lifetime value |
| Lead generation | Number of leads | Cost per lead, and close rate on those leads |
| Brand awareness | Follower count | Reach, video views, retargeting pool growth |
| Efficiency | Cheapest CPC | Conversion rate at that CPC |
| Whole business | Per-platform ROAS | MER, total revenue over total spend |
That last row is the one most businesses never look at. Platform ROAS only counts what a platform can see, while MER measures your whole marketing machine against every dollar you spent. Since the iOS 14 tracking changes, the gap between those two numbers has been where a lot of good campaigns quietly got killed.
What success looks like depends on the goal
This is where most people come unstuck. They stare at numbers without tying them to what the campaign was for. Metrics without context are steps on a treadmill: plenty of movement, no distance covered.
If the goal is sales
ROAS and CPA are your north star. Two questions matter: are you spending less to acquire a customer than they’re worth, and can you scale without your margins folding?
Paid search often shines here because it catches people with intent already formed. For ecommerce, track add-to-carts, checkouts and purchases, then use retargeting to close the loop.
If the goal is lead generation
Watch cost per lead, lead quality, and what happens to those leads after they arrive. Good leads justify the spend. Junk leads burn budget and your sales team’s patience at the same time.
Use qualifying questions on your forms and connect your CRM so you can see which leads actually closed. A low CPA with a zero percent close rate is still a failure, just a cheaper one.
If the goal is brand awareness
Look at impressions, reach, video views and engagement, and accept that these campaigns don’t convert on the spot. You’re warming an audience for later, which is exactly what makes retargeting work when it’s their turn.
The most common thing I see in an audit is a business optimising hard toward a number nobody chose on purpose. Usually it’s cost per click, because it’s the one sitting at the top of the dashboard. Meanwhile the close rate on those leads is dreadful and nobody’s looked at it in months. Pick your one number before you open the platform, or the platform picks it for you.
The tools worth using
You don’t need to be a data scientist, but you do need a stack that connects spend to outcomes.
For tracking: Meta Ads Manager, Google Ads, Google Analytics 4, UTM parameters, and Looker Studio to pull it into one view.
For attribution and behaviour: Triple Whale or Northbeam, Hyros, and Hotjar or Clarity to watch what people actually do once they land.
Whatever you use, the reporting has to answer business questions rather than flatter the platform. How agencies should report results covers what a report is meant to look like, and if you want the short version of what to track, it’s four questions every business owner should be able to answer.
Testing beats guessing
If you’re not testing, you’re guessing with a budget attached.
A structured A/B testing approach kills what isn’t working quickly, so you can put the money behind what is. Test your headlines, your ad copy, your calls to action, image against video, and your landing page layouts.
When something wins, don’t stop. Marketing is small improvements compounding, and the compounding only happens if you keep going. One rule though: isolate your variables. Change five things at once and you’ve learned precisely nothing about which one mattered.
Free strategy call
Get your ad data saying something useful.
Thirty minutes on your account with someone who reads these dashboards daily. You’ll leave knowing which number to optimise and which ones to ignore.
What to do with all this
Measuring ad performance isn’t about hunting the lowest CPC or the highest CTR. It’s about tying what the ads do back to what the business needs.
Align your metrics with your goal, track them with tools that tell the truth, and build a loop where every month teaches you something. That’s how the spending stops feeling like a gamble.
Still not sure whether your ads are earning their keep? That’s what our paid advertising team does, and the ROAS calculator is a free place to start on the numbers.
Measuring paid ad performance FAQs
How do I know if my paid ads are working?
Compare performance against the goal you set before launching. For sales, that’s ROAS and CPA against customer lifetime value. For leads, it’s cost per lead and the close rate on those leads. For awareness, it’s reach and engagement. A number is only good or bad relative to what you were trying to achieve.
What is a good ROAS?
It depends on your margin, not on a benchmark someone posted online. Your break-even ROAS is roughly 1 divided by your gross margin, which is the floor you have to clear before anything counts as profit. A ROAS of 3 can be healthy for one business and a slow loss for another.
What’s the difference between ROAS and MER?
ROAS measures revenue an ad platform credits to its own campaigns. MER measures total business revenue against total ad spend across every channel. Use ROAS to optimise a campaign and MER to judge whether your paid spend is moving the whole business.
Which metrics are vanity metrics?
Any number that goes up without changing a decision. Impressions, likes, follower counts and cheap clicks all qualify when they’re reported on their own. They become useful only when paired with what happened next: conversion rate, cost per acquisition, or revenue.
How often should I check my ad metrics?
Weekly for campaign management, monthly for business decisions. Checking daily tempts you into reacting to noise, and most platforms need time to gather enough data before a change means anything. Set a rhythm and stick to it.