Ad relevance diagnostics are three rankings in Ads Manager that show how your ad compares with other ads competing for the same people. They replaced Facebook's single relevance score in 2019. They are diagnostic tools: they help explain results, they aren't results themselves, and Meta says they are not used directly in the auction.
The three rankings
| Ranking | What it compares | A low ranking suggests |
|---|---|---|
| Quality ranking | Your ad's perceived quality, using feedback and post-impression signals | The ad looks low quality, clickbait-like or gets hidden |
| Engagement rate ranking | Your ad's expected engagement rate | The creative isn't stopping people or getting reactions |
| Conversion rate ranking | Your ad's expected conversion rate for your optimisation goal | The post-click journey or offer is weak for this audience |
Each ranking shows Above average, Average, or a Below average band (bottom 35%, bottom 20% or bottom 10% of ads). Meta only shows the rankings once an ad has more than 500 impressions.
How to use them
Meta's own guidance is to read the three together:
- Low quality, others average or above: improve the creative itself. Avoid engagement bait and withheld information.
- Low engagement, others fine: the ad isn't interesting to this audience. Try a new hook or angle.
- Low conversion rate, others fine: people like the ad but don't convert. Check the landing page, price, offer and whether the ad attracts the right people.
- All three low: the ad and the audience don't fit. Rethink both.
What diagnostics are not
- Not the auction score. Meta's ad auction uses bid, estimated action rates and ad quality. The diagnostics reflect similar ideas but are calculated separately for reporting.
- Not a goal. An ad with average rankings and a strong cost per purchase is a good ad. Don't pause it to chase above-average labels.
- Not precise. They are broad bands compared with an unknown set of competitors, and they shift as the audience and competition change.
Worked example
Example: an ad with rising cost per purchase.
| Ranking | Week 1 | Week 3 |
|---|---|---|
| Quality | Average | Average |
| Engagement rate | Above average | Below average (bottom 35%) |
| Conversion rate | Average | Average |
Engagement fell while quality and conversion held. Combined with rising frequency, this points to ad fatigue: the audience has seen it too often. A refreshed hook is the likely fix, not a new landing page.
Common mistakes
- Judging an ad on diagnostics with few impressions.
- Comparing rankings across very different audiences or objectives.
- Ignoring them completely. When cost per purchase moves, they help show whether the cause is the creative, the audience fit or the page.
For other signs an ad has stopped working, read why Facebook ads stopped working, and check CTR and conversion rate alongside.
Frequently asked questions
What are ad relevance diagnostics on Facebook?
They are three Ads Manager rankings, quality, engagement rate and conversion rate, that compare your ad with ads competing for the same audience. They help explain why an ad performs as it does and appear once the ad passes 500 impressions.
Does Facebook still have a relevance score?
No. Meta replaced the single 1 to 10 relevance score with the three ad relevance diagnostics in 2019. Old guides that mention the relevance score describe a metric that no longer exists.
Do ad relevance diagnostics affect my costs?
Meta says the diagnostics themselves are not used in the auction. They reflect similar factors, such as expected engagement and ad quality, which do affect auction results, so poor rankings often go with higher costs.
Should I pause ads with below average rankings?
Not automatically. Judge ads on cost per purchase and ROAS first. Use the rankings to diagnose why a weak ad is weak, for example whether to fix the creative, the audience or the landing page.
