Jump to a section
You launch an ad, refresh the dashboard an hour later, and the numbers look alarming. The cost per click is higher than you hoped, there are no sales yet, and the spend is climbing. So you panic. You turn the ad off, or you start changing five things at once, or you convince yourself paid traffic just does not work for you. This is how most beginners lose money, and it is almost never the ad's fault. It is a reading problem.
Ad platforms hand you a wall of metrics and imply all of them matter equally and immediately. They do not. Some numbers are early signals you can act on, some are lagging results that mean nothing until you have enough data, and a couple are mostly there to make the dashboard look busy. This guide is about reading the numbers well enough to know which thing is actually broken, the creative, the audience, or the landing page, so you fix that instead of blowing up a campaign that was one tweak away from working.
Where does the money actually come from?
The money does not come from a good CTR or a low CPM. Those are steps. It comes from the full chain holding together: the ad gets a click, the click becomes a lead or a sale, and that sale is worth more than what you paid for the click. Every metric is just measuring one link in that chain.
Ad shown to people (CPM = cost to show it)
|
v
Some of them click (CTR = how many, CPC = cost per click) <-- creative + audience live here
|
v
They land on your page
|
v
Some of them convert (landing page does its job here)
|
v
That conversion is worth money (CPA = cost per, ROAS = return) <-- the whole thing pays or it doesn't
|
v
Profit (only if the sale is worth more than the click cost)
Reading the metrics correctly means finding which link is breaking. If the CTR is fine but nobody buys, the ad is doing its job and the page or offer is not. If the CTR is terrible, the page never even gets a fair test, because too few people arrive. Kill the wrong link's ad and you throw away the half that was working. If you want the full picture of how paid spend turns into revenue, how paid advertising makes money lays out the model this all fits inside.
How it actually works
Take the six metrics one at a time, in the order they actually matter.
CTR, click-through rate. Of the people who saw the ad, how many clicked. This is your fastest, most honest signal, because it needs the least data to mean something. A low CTR says the ad is not connecting: wrong audience, weak hook, or an offer nobody wants. A healthy CTR says the top of your funnel is working and you can trust what happens after the click. This is the number to read first.
CPC, cost per click. What you pay for each click. It is mostly a consequence of CTR and CPM, not an independent lever. When platforms reward a high CTR with cheaper clicks, a bad CPC is usually a CTR problem wearing a costume. Fix the CTR and the CPC tends to follow.
CPM, cost per thousand impressions. What it costs to be shown. This is set by the auction, the competition, and the time of year far more than by anything you did. A high CPM is not a personal failing. It is a market condition. Do not agonize over it; understand it. How advertising auctions work explains why this number moves on its own.
CPA, cost per acquisition. What you paid to get one conversion, whether that is a lead or a sale. This is a money metric, and it is only trustworthy once you have real conversions behind it. One sale off two clicks does not tell you your CPA. It tells you almost nothing.
ROAS, return on ad spend. Revenue divided by ad spend. This is the number that actually decides whether the campaign lives. The catch is that ROAS is meaningless without a break-even point to compare it to. A 2x ROAS is great for one business and a slow death for another. Know your break-even before you judge it, which break-even ROAS explained walks through directly.
Frequency. The average number of times each person has seen the ad. Low is fine. As it climbs, the same people keep seeing the same thing, CTR drops, and CPMs effectively rise. High frequency on a small audience is a sign the creative is getting stale or the audience is too narrow, not that ads have stopped working.
None of these mean anything without the plumbing to measure conversions. If you cannot see which clicks turned into sales, you are flying blind on the two metrics that matter most. Get conversion tracking working before you spend real money.
A clearly hypothetical example
These numbers are invented to show how to read a funnel, not a promise of results. Yours will look different.
Say you run a campaign for a $60 product and spend $200. You get 10,000 impressions, 100 clicks, and 2 sales.
A beginner looks at that and sees "only 2 sales, this is a disaster," turns it off, and moves on. Read the funnel instead.
Impressions: 10,000
Clicks: 100 -> CTR = 1.0% (decent, people want the offer)
Cost: $200 -> CPC = $2.00 (fine for this CTR)
Sales: 2 -> CPA = $100 (you paid $100 to sell a $60 thing)
Revenue: $120 -> ROAS = 0.6x (losing money)
The instinct is to blame the ad. But the CTR is healthy at 1 percent, which means the creative and audience are doing their job: 100 people wanted this enough to click. The break is lower down. Two sales out of 100 clicks is a 2 percent conversion rate on the page, and against a $60 product with a $2 click, that is not enough to break even. The ad is fine. The landing page or the offer is where the money is leaking.
So you would not touch the creative. You would work on the page: the headline, the match between what the ad promised and what the page delivers, the checkout friction. If the same ad sent to a better page converted at 5 percent instead of 2, that is 5 sales, $300 revenue, a 1.5x ROAS, and a campaign that suddenly works. Same ad. Same spend. Different diagnosis. That is the entire point of reading the numbers instead of reacting to them. If the page is where yours is leaking, landing pages for paid traffic is the fix.
What you need (required vs optional)
Required:
- Conversion tracking that actually fires, so you can tie clicks to sales. Without it, CPA and ROAS are guesses.
- Your break-even ROAS, calculated before you launch. You cannot judge a return without knowing the number that means "profitable."
- Enough patience to let the campaign gather data before you read the money metrics. A handful of clicks is not a sample.
Optional but helpful:
- A simple spreadsheet where you log the funnel numbers daily, so you see trends instead of reacting to single hours.
- A couple of ad variations running at once, so a bad CTR points at the creative rather than leaving you guessing.
- A benchmark for your industry's typical CTR and conversion rate, just to know whether a number is genuinely bad or just normal.
What it costs
Reading metrics costs nothing but discipline. The real cost is the money you spend gathering enough data to read them honestly.
That is the part beginners resent. You have to spend some amount before CPA and ROAS mean anything, and that spend can happen while the numbers still look bad. This is not waste; it is the price of a real sample. The waste is spending that money, collecting the data, and then ignoring what it says because you panicked at hour two. Decide up front how much you are willing to spend to get a readable result, treat that as tuition, and do not pull the plug before the data is in. How to set an ad budget covers how to size that number so a test cannot quietly bankrupt you.
How long it takes
Long enough to get statistically real numbers, which depends on your price and conversion rate, not the clock.
If a conversion happens once every 50 clicks, you need a few hundred clicks before CPA means anything. At a slow spend that could be a week. The click metrics, CTR and CPC, stabilize far faster, often within the first few hundred impressions, which is exactly why they are the ones to act on early. The mistake is applying the timeline of a fast metric to a slow one: judging your ROAS after an afternoon the way you would judge your CTR. Give each metric the sample it needs. Do not attach a fixed number of days to it; attach it to volume.
What beginners usually get wrong
The biggest mistake is judging money metrics too early. Two clicks and no sale is not a failed campaign, it is not a campaign yet. People kill ads at a sample size where the numbers are pure noise, and they never find out the ad would have worked. This is one of the main ways beginners burn money on ads, by paying for data and then refusing to wait for it.
The second mistake is changing everything at once. CPA looks bad, so they swap the creative, the audience, and the budget in one move. Now the campaign performs differently and they have no idea which change did it. Change one variable at a time or you learn nothing.
The third mistake is obsessing over the metrics you cannot control. CPM is set by the auction. Fretting over it while ignoring a broken landing page is spending worry where it does nothing.
The fourth mistake is reading a good CTR as success. A high CTR with no sales does not mean the campaign works. It means the ad works and something after the click does not. A great CTR feeding a broken page is an efficient way to lose money.
How I would start
If I were staring at a dashboard that looked bad, here is the order I would work in.
- Confirm tracking actually fires. Run a test conversion and make sure it shows up. If the data is wrong, every decision after this is wrong.
- Write down my break-even ROAS before looking at anything else, so I am judging against a real target instead of a feeling.
- Read the CTR first. If it is weak, the problem is the creative or the audience, and no downstream fix matters until people click.
- If the CTR is healthy but sales are not coming, ignore the ad and look at the page and the offer. That is where the leak is.
- Give the money metrics a real sample before trusting them. Note the numbers daily, but do not act on CPA or ROAS until there are enough conversions to mean something.
- Change one thing, then wait. Read the effect on the specific metric that thing should move.
- Once the funnel is profitable and stable, and only then, think about scaling, which when to scale an ad campaign covers.
What I would not do
I would not turn an ad off after an hour because the spend outpaced the sales; that is the campaign gathering the data I am paying for. I would not change five things at once and call the result a test. I would not lose sleep over CPM, which the auction sets for me. I would not treat a good CTR as proof the campaign makes money, or a bad early ROAS as proof it never will. And I would not run a single dollar of paid traffic without conversion tracking working, because everything on this page depends on being able to see which clicks turned into money.
The bottom line
Ad dashboards are designed to make every number feel urgent. Most of them are not. Read the click side first, because CTR tells you fast whether the creative and audience connect. Read the money side last, because CPA and ROAS only mean something after a real sample, and only against a break-even number you set in advance. When something looks broken, walk the funnel top to bottom and find the first link that fails, then fix that one thing. The difference between people who make paid traffic work and people who quit is rarely the ads. It is whether they read the numbers calmly enough to fix the right thing before killing the campaign that was about to pay.
Want to know what actually works?
We break down money-making methods, tools and programs without the ridiculous promises.