The Facebook ads metrics that matter are the ones tied to business results and to the signals Meta’s delivery system uses: cost per purchase or lead, return on ad spend, conversion event quality, frequency, click-through rate on links, and for video, the share of viewers who watch past the first few seconds. Vanity metrics such as reach and post likes describe exposure, not outcomes. Defensible results also need incrementality checks, because attributed numbers include sales that would have happened anyway.
Ads Manager shows more than a hundred columns. Picking the right dozen and reviewing them on a schedule is most of the job.
Which metrics should be in your default column set?
Create a saved column preset in Ads Manager under Columns > Customize Columns. Include the result your campaign optimizes for, cost per result, amount spent, purchase or lead value, ROAS, CPM, link click-through rate (CTR link), cost per link click, frequency, and for video, ThruPlays and 3-second video plays.
Add landing page views next to link clicks. A large gap between the two means people click but the page does not load, often due to slow mobile pages. That is a site problem that no ad change fixes.
- Results and cost per result for the optimization event.
- Purchase value and ROAS, if you sell online.
- CPM and frequency for delivery.
- Link CTR and cost per link click for creative interest.
- Landing page views to spot load problems.
- 3-second plays divided by impressions as a hook rate for video.
Why does conversion data quality come first?
Meta optimizes toward the events you send. If the Pixel misses purchases, or the Conversions API is not set up, the system learns from a partial picture. Events Manager shows Event Match Quality scores for server events and warnings about deduplication; check them before trusting any cost-per-purchase figure.
Since Apple’s App Tracking Transparency changes, browser-only tracking undercounts. Sending events through the Conversions API, directly or through a partner integration such as Shopify’s, recovers some of them. Deduplicate with an event ID so the same purchase is not counted twice.
- Pixel and Conversions API both active.
- Deduplication with event_id for events sent both ways.
- Event Match Quality reviewed in Events Manager.
- Aggregated Event Measurement priorities set for your domain where relevant.
How do you read ROAS and cost per purchase honestly?
Ads Manager reports attributed results inside an attribution setting, commonly 7-day click and 1-day view. View-through results credit ads seen but not clicked, which can inflate results for retargeting. Compare results under 7-day click only to see how much depends on views.
Then compare with your store or CRM. If Ads Manager reports more purchases than your store recorded from all sources in the same period, attribution overlaps with other channels. Blended figures, such as total revenue divided by total ad spend across channels, give a sanity check.
A simple weekly comparison
For example, an online store might track three numbers side by side each week: Meta-attributed ROAS, ROAS under 7-day click only, and blended marketing efficiency (all revenue divided by all ad spend). When Meta ROAS rises but blended efficiency does not move, the gain is probably attribution, not new sales.
What does frequency tell you?
Frequency is the average number of times each person saw your ads in the period. Rising frequency with rising cost per result usually means the audience is saturated and the creative is tiring. There is no universal threshold; it depends on audience size and purchase cycle.
Watch the trend rather than a fixed number. If frequency climbs week over week while CTR falls, add new creative or widen the audience. Broad targeting with Advantage+ audience settings can reduce saturation by letting the system find new people.
How do you judge creative with metrics?
Look at creative in three steps. Hook: did people stop scrolling (3-second plays divided by impressions for video, or CTR for images)? Hold: did they keep watching (ThruPlays or average watch time)? Action: did they click and convert (link CTR, cost per result)?
A video with a strong hook but weak action suggests the message after the opening does not match the offer. A weak hook means the first seconds need work. Breaking it down this way tells the creative team what to change rather than just “make a better ad”.
- Hook rate: 3-second video plays / impressions.
- Hold rate: ThruPlays / 3-second plays.
- Action: link CTR and cost per result.
How do you prove ads cause sales?
Meta offers Conversion Lift studies for eligible advertisers, which hold out a randomized group from seeing ads and measure the difference in conversions. Ask your Meta representative or check Experiments in Ads Manager for availability. A simpler option is a geographic holdout: pause ads in a set of regions for a few weeks and compare sales trends with regions that kept running.
Lift results are usually lower than attributed results. That is expected. The lift number is the one to use when deciding total budget, while attributed numbers help compare ads within the account.
Which metrics can you safely ignore most of the time?
Reach, impressions and post engagement are useful for awareness campaigns but distract in conversion campaigns. Relevance diagnostics (quality, engagement and conversion rate rankings) are comparisons against ads competing for the same audience; treat them as hints, not targets.
CPM on its own also misleads. A cheap CPM with poor conversion often means low-intent placements. Judge CPM together with cost per result.
Setting thresholds for your own account
Instead of borrowing benchmarks from forums, set thresholds from your own history. Take the last three months, find the median cost per result and link CTR for active ads, and mark anything far worse as a candidate to pause after enough spend. A common rule is to wait until an ad has spent at least two or three times your target cost per result before judging it; earlier decisions often kill ads during normal early variation.
Finally, keep one shared reporting sheet for the team. When everyone reads the same columns with the same attribution setting, discussions focus on decisions instead of on whose number is right.
Revisit the column preset each quarter, since Meta renames and adds metrics regularly.
Where to start this week
Build and save a column preset with the metrics listed above and apply it to every campaign. Open Events Manager, check that the Conversions API is active and deduplicated, and note your Event Match Quality.
Then compare last month’s Meta-attributed purchases with your store’s orders and with 7-day click only results. Write the three numbers down, and repeat the comparison every week so you can see which results hold up.
Keep reading: Facebook Ads · Social Media Marketing