What the results column counts
Results is not a fixed metric. It is a placeholder that fills with whatever outcome the ad set was told to optimise for. Set an ad set to chase leads and a result is a lead; set it to chase link clicks and a result is a link click; set it to chase purchases and a result is a purchase. The result type column, sitting beside it, names what is being counted in each row.
This is why the column can be read confidently within one ad set and dangerously across several. Two rows showing the same figure may be counting entirely different events, and adding them together produces a number that means nothing at all.
Why results matter
Used properly it is the fastest way to see whether an ad set is doing the specific job it was given. It also keeps the account honest about intent: if the result type is not the outcome the business cares about, the campaign is optimising for the wrong thing, and that is visible at a glance rather than buried in settings.
It carries a second, quieter signal. Meta’s delivery system learns from the result event, so whatever sits in that column is what the system is actively hunting for. A campaign optimising for a cheap intermediate action will get better and better at producing that action and no better at producing revenue.
Common mistakes with results
Summing the column across a whole account is the most common. A total made of leads, video views and add-to-carts is arithmetic without meaning, yet it appears in reports regularly because the column carries the same label in every row.
The second is forgetting that results are attributed, not observed. What appears depends on the attribution setting in force, so the same week can show different totals after a settings change, and a result may be credited to a day earlier than the sale actually happened. It is also Meta’s own count, taken from the pixel or the Conversions API, and it will not line up exactly with your CRM.
The third is leaving the result type unstated in a report. Anyone reading a bare results figure without knowing the type cannot judge whether it is good, bad or irrelevant.
How to act on it
Always show the result type next to the number, in the dashboard and in the report. Compare results only between ad sets that share an optimisation event, and compare across objectives using money — cost per result — while remembering that even that is only fully comparable when the underlying event matches.
Choose the result event as close to revenue as the account can sustain. Optimising for purchases or qualified leads beats optimising for clicks, provided there is enough weekly volume for delivery to learn from. If there is not, move one step earlier in the funnel deliberately and record why, rather than defaulting to whatever event happens to be cheapest. And check that the campaign objective agrees with the result you have chosen, because a mismatch there quietly undoes everything else.
