TL;DR
- Meta retired the 1 to 10 relevance score on 30 April 2019 and replaced it with quality ranking, engagement rate ranking and conversion rate ranking.
- Each ranking compares your ad with ads competing for the same audience and reports a bucket: above average, average, or below average (bottom 35%, 20% or 10%).
- The columns are on the Ads tab under Customise columns, and need about 500 impressions before Meta shows a value.
- Opportunity Score (0 to 100, introduced 2024) rates how closely your setup follows Meta's recommendations; relevance diagnostics report how delivered ads compared with competitors.
Meta ads relevance diagnostics are three ad-level metrics in Ads Manager, quality ranking, engagement rate ranking and conversion rate ranking, that compare your ad against other ads competing for the same audience. They replaced the single 1 to 10 relevance score on 30 April 2019, and each one reports a bucket (above average, average, or one of three below-average tiers) rather than a number, so the job is to read the three together and fix the weakest one.
If you came here searching for the Facebook relevance score, this is the page that explains what it became. The sections below cover what each ranking measures, where the columns live in Ads Manager, why an ad shows "below average", how the diagnostics differ from the newer Opportunity Score, and what changes move each ranking, limited to the mechanics Meta documents rather than benchmarks nobody can verify.
What replaced the Facebook relevance score
Until 2019, every Facebook ad with enough impressions got a relevance score from 1 to 10. It was one number that blended several signals, which made it easy to glance at and hard to act on: a 4 told you something was wrong but not whether the creative, the audience or the landing page was the problem. Meta retired the score on 30 April 2019 and replaced it with the three ad relevance diagnostics, each of which isolates one dimension. The replacement was announced in March 2019 along with a wider clean-up of Ads Manager metrics.
The design change matters more than the name change. The old score was absolute; the new rankings are relative. Each one compares your ad with the other ads that competed for the same audience during the reporting period, so the same ad can move between buckets when the audience or the competition changes, with no edit on your side.
Quality ranking, engagement rate ranking and conversion rate ranking explained
Meta defines the three diagnostics as follows.
- Quality ranking: how your ad's perceived quality compared with ads competing for the same audience. Quality is measured from feedback people give on ads (such as hiding or reporting them) and from Meta's assessment of low-quality attributes in the ad. Meta also lists attributes it treats as low quality: engagement bait, withholding information to get a click, sensationalised language, and claims the landing page does not back up.
- Engagement rate ranking: how your ad's expected engagement rate compared with ads competing for the same audience. Engagement covers interactions such as clicks, reactions, comments and shares; the prediction is Meta's, built from the signals it sees while the ad serves.
- Conversion rate ranking: how your ad's expected conversion rate compared with ads that have the same optimisation goal and compete for the same audience. The "same optimisation goal" clause is important: a purchase-optimised ad is compared with other purchase-optimised ads, never with lead or traffic ads.
All three are reported at the ad level, and an ad has to reach about 500 impressions before Meta shows a value; below that, the column reads as not available. Meta also notes that quality ranking is only available for the last 35 days and is not available for dynamic creative. Meta also states that the diagnostics are not themselves a ranking factor inside the auction. The auction uses its own estimates of action rates and ad quality; the diagnostics are a reporting view that tells you how those estimates compared with competitors. Improving the underlying ad improves both.
Where to find ad relevance diagnostics in Ads Manager
The three columns live on the Ads tab (not the campaign or ad set tabs, because the values are per ad). To show them: open Ads Manager, select the Ads tab, open the Columns menu, choose Customise columns, and search for "ranking". Tick quality ranking, engagement rate ranking and conversion rate ranking, then save the preset so the view persists. The values can also be exported in reports with the same column names.
Two reading habits avoid most confusion. First, set the date range to a period in which the ad actually ran at volume; the rankings describe that window, and a short range on a low-budget ad will often show no value at all. Second, compare ads within one ad set first, because they faced the same audience; comparing an ad in a broad audience with an ad in a retargeting audience tells you little, since the competition differed.
Why an ad shows "below average"
The buckets are percentiles. Meta reports above average for ads above the 55th percentile of competing ads, average for roughly the 35th to 55th percentile, and three below-average tiers: bottom 35%, bottom 20% and bottom 10%. A "below average (bottom 20% of ads)" label therefore means that on that one dimension, 80% of the ads competing for the same people were predicted to do better.
Because the measure is relative, a below-average label does not by itself mean the ad is failing against your own targets. Meta's own guidance says that if an ad is meeting your advertising goals, a below-average ranking is not necessarily a problem, and that moving an ad from below average to average has a bigger effect than moving it from average to above average. In practice that means a below-average quality ranking on an ad that converts at an acceptable cost is a watch item, while a bottom-10% label on an ad whose costs are rising is an action item.
The table below is a practical reading of the combinations, following the direction of Meta's own guidance on which lever each ranking points to.
| Pattern across the three rankings | What it usually points to | First change to test |
|---|---|---|
| Quality below average, engagement and conversion average or above | The ad is working but reads as low quality: bait-style copy, missing information or claims the page does not back up | Rewrite the copy to state the offer plainly and check the landing page matches the claim |
| Engagement below average, quality and conversion average or above | The creative is honest and the page converts, but it does not stop the scroll for this audience | New creative formats or hooks for the same audience; test video against static |
| Conversion below average, quality and engagement average or above | People engage but do not complete the optimised action: landing page friction, offer mismatch or the wrong optimisation event | Fix page speed, forms and offer; confirm the pixel or Conversions API event is the right one |
| All three below average | The audience is probably wrong for this ad, or the ad is wrong for this audience | Test a different audience and new creative, then re-read the rankings after 500 impressions |
How relevance diagnostics differ from Opportunity Score
Opportunity Score, introduced by Meta in 2024, is a score from 0 to 100 for the ad account as a whole, with recommendations attached to individual campaigns, ad sets and ads. It measures how closely your setup follows Meta's recommendations (for example Advantage+ settings, placements, budget or creative suggestions), and applying a recommendation raises the score. It is forward-looking advice about configuration.
Relevance diagnostics are backward-looking reporting about delivery. They do not change when you accept a recommendation; they change when the ad's predicted quality, engagement or conversion rate changes relative to competitors. A campaign can have a high Opportunity Score and ads with bottom-20% conversion rate rankings at the same time, because following Meta's setup suggestions does not fix a slow landing page. The two are complementary: Opportunity Score tells you whether the account is set up the way Meta's systems prefer, and the diagnostics tell you how individual ads performed against the market once they ran.
One more difference: Opportunity Score is Meta's judgement of your choices, so it moves whenever you change settings. The diagnostics are Meta's judgement of your ads against other advertisers' ads, which is why they are the better tool for deciding which creative to cut and which page to fix.
How to improve each ranking
Improving quality ranking
Quality ranking responds to negative feedback and to low-quality attributes in the ad, so the fixes are about honesty and friction. Remove engagement bait ("tag a friend", "comment yes"), remove copy that withholds the price or the catch to force a click, tone down sensational wording, and make sure the landing page delivers what the ad promised within the first screen. Check the ad's comments for complaints, because people who report an ad usually said why first. If the ad uses a low-quality attribute Meta lists, a rewrite will usually move the ranking faster than any targeting change.
Improving engagement rate ranking
Engagement rate ranking is about whether this audience interacts with this creative. Test a different hook in the first three seconds of video or the first line of copy, try a format the audience sees less of (a static image in a feed full of video, or the reverse), and make sure the creative is built for the placement rather than cropped into it. Keep the audience constant while you test, or you lose the comparison. Engagement bait raises engagement at the cost of quality ranking, so it is a trade that backfires.
Improving conversion rate ranking
Conversion rate ranking compares predicted conversions for the optimised event, so start with the event. Confirm the pixel or Conversions API fires the event you are optimising for, that the ad set is optimising for the event that matches the ad's promise, and that the landing page loads fast on mobile with the form or checkout above the fold. Offer mismatch is the quiet cause: an ad promising a discount that the page does not show converts poorly no matter how good the creative is. If the ad set is optimising for a rare event with little data, Meta's prediction is noisy, and the ranking can swing without any real change.
What this means for Thai marketers
Thai feeds are dense with video and with LINE-style conversational offers, so engagement rate ranking is often the first of the three to slip for a static ad, while quality ranking is the one that bites accounts that lean on urgency language ("ด่วน", "เหลือวันสุดท้าย") and prices hidden behind "inbox for price". Meta reads both patterns as low quality attributes. Conversion rate ranking in Thailand is frequently a tracking problem rather than a page problem, because many SMEs convert in LINE or in Messenger and never fire a website event, so the optimised event does not describe the real conversion. Mapping the real conversion path before judging the rankings saves a lot of creative rewrites. If you want the diagnostics read alongside the rest of the account, a Facebook Ads management service that builds the reporting view first and tests creative second usually gets more out of the same budget.
FAQ: Meta ads relevance diagnostics
Does the Facebook relevance score still exist?
No, Meta removed the 1 to 10 relevance score on 30 April 2019 and replaced it with quality ranking, engagement rate ranking and conversion rate ranking. Any tool still showing a relevance score is showing a stale or estimated value, not a Meta metric.
Why are my ad relevance diagnostics blank?
An ad needs around 500 impressions in the selected date range before Meta reports the rankings, so new ads, low-budget ads and short date ranges show no value. Widen the date range or wait for delivery; the columns also stay empty for objectives where a given ranking does not apply, such as conversion rate ranking on an ad set that is not optimising for conversions.
Do relevance diagnostics affect how much I pay?
Not directly: Meta says the diagnostics are not themselves a factor in the auction, which uses its own estimates of ad quality and action rates. The underlying signals they report do influence delivery and cost, so an ad that improves its real quality or conversion rate will usually see both the ranking and the cost move.
Should I pause every ad that is below average?
No, Meta's guidance is that a below-average ranking is not necessarily a problem if the ad is meeting your goals, and that lifting an ad from below average to average matters more than pushing average to above average. Use cost per result as the decision metric and the rankings as the diagnosis of what to fix.
Is Opportunity Score the new relevance score?
No, Opportunity Score rates how closely your setup follows Meta's recommendations, while relevance diagnostics report how your delivered ads compared with competitors. They answer different questions: Opportunity Score is one account-level score, while the diagnostics are reported per ad.
Reading quality ranking, engagement rate ranking and conversion rate ranking together is the fastest way to tell whether an underperforming ad needs new creative, a new audience or a better landing page. If you would like the diagnostics built into a standing creative-testing process across Instagram and Facebook, with retargeting layered on top, our social ads team in Bangkok can set up the reporting and run the tests.
