Ads Manager can show dozens of columns, and most of them will not help you make a decision. A handful of numbers tell you nearly everything: how much it costs to be seen, whether people care, and what a result costs. Read together, they also tell you where a campaign is going wrong.
CPM: the cost of being seen
CPM is the cost of a thousand impressions. It reflects how expensive it is to reach your audience, which depends on competition, the time of year, the audience you chose and how Meta rates your ad. A rising CPM is not automatically bad, but it means you are paying more for the same attention.
CTR: whether people care
Click-through rate is the share of people who saw the ad and clicked. Look at the link click-through rate, which counts clicks to your destination, not every tap on the ad. A low rate usually points at the creative or the offer: the image did not stop people, or the message did not interest them.
CPC and frequency
Cost per click follows from the two numbers above: cheap impressions and a high click rate give cheap clicks. Frequency is the average number of times each person has seen your ad. When frequency climbs and click rate falls, your audience is tiring of the ad and it is time for fresh creative.
Cost per result: the number that matters
Cost per result, often called CPA, is what you pay for the action you asked Meta to optimise for, such as a lead or a purchase. This is the figure to judge a campaign by. An ad with expensive clicks can still win if those clicks convert. An ad with cheap clicks can lose if nobody acts.
Reading them together
The numbers form a chain. High CPM with a good click rate suggests a competitive audience. A good click rate with a poor cost per result suggests the landing page or offer is letting you down. A falling click rate with rising frequency suggests creative fatigue. Follow the chain to the weak link and fix that.
Metrics are there to point you at decisions, not to be admired. We track cost per result down to the individual creative, because that is the level at which scaling decisions are made.