Ad metrics: reading CPM, CTR, CPC, CPA and ROAS

In the video, I compare advertising objectives. To interpret them, decide which action matters: a visit, an enquiry, a registration or a purchase. Then check that you can observe it and understand what the platform counts.

If you want qualified enquiries, counting only clicks misses the main part of the journey. Review the landing page, form and subsequent response too. The administrator's labels and options may differ from the recording.

Impressions count how often an ad is shown, not unique people. CPM expresses cost per thousand impressions. CTR relates clicks to impressions, but check which click type the selected column counts. CPC expresses the average cost of those clicks.

This adaptation clarifies the distinctions to avoid mixing reach, impressions and visits. One person may see an ad several times. Before comparing campaigns, check that the metrics and time periods are equivalent.

In the class, I use spending and conversion examples. If you spend €5 and record one attributed action, the observed cost for that action is €5. To interpret it, you need to know whether it was a registration, purchase or another event.

A lower cost per click does not guarantee a lower cost per client. The proportion of visits completing the next step may change. Review both the quantity and quality of enquiries and the sales actually confirmed.

ROAS relates revenue attributed to ads to advertising spend. It does not include every business cost by itself. To assess the result, also consider delivery costs and which revenue is confirmed.

Attribution has context: the time window, measured events and possible differences between tools. Do not present the administrator's figure as a complete explanation of events. Refunds or duplicate records require adjustments to the interpretation.

The recording includes creative examples and ways to distribute budgets. The useful conclusion is to observe what works for a particular offer and audience without assuming video always beats images or one setup suits everyone.

Define a question before changing the campaign. Keep a record of the message, destination, period and observed result. A small test may identify a comprehension or journey problem; it does not turn an estimate into a sales promise.