TL;DR
Meta's Opportunity score grades your account from 0 to 100 and hands you a list of recommendations to raise it. The score measures how closely your setup matches Meta's preferred patterns, not how profitable your account is. Some of its suggestions are genuinely useful, especially the ones about fixing broken basics. Others would hand Meta more control over your budget, your audiences, or your creative for no guaranteed return. Meta has reported a 12% median decrease in cost per result for advertisers who adopted its recommendations in testing, but that is a reported figure from Meta's own tests, not a promise for your account. Use the list as a checklist. Ignore the grade.
What the score actually measures
Opportunity score lives in Ads Manager, attached to your account, and it comes with recommendations: consolidate campaigns, switch on Advantage+ features, add more creative variations, broaden audiences, fix tracking gaps, and so on. Each recommendation you apply nudges the score up. Each one you ignore leaves it where it is.
Here is the part Meta does not put in big letters: the score measures adoption of Meta's recommended setup, not business results. An account can score 95 and lose money, if the recommended setup does not fit the business. An account can score 55 and print profit, if its structure is deliberate and its tracking is sound. The score has no idea what your margins are, what a good lead is worth to you, or whether that broad audience is full of people who will never buy.
That does not make it useless. It makes it a checklist with a grade attached, and grades make people do silly things. Your job is cheaper results, not a prettier score.
The recommendations worth taking
Some suggestions on the list are just good housekeeping, and ignoring them out of spite is as foolish as following everything blindly.
Fix anything about measurement first. Broken pixels, missing conversion events, and unconnected catalogues are not strategic choices, they are leaks. If Opportunity score flags a tracking gap, fix it before you debate anything else. You cannot judge any recommendation without trustworthy numbers.
Creative volume is usually worth a look too. If you are running one ad per ad set and wondering why performance is flat, the suggestion to add more genuine variations is fair.
The recommendations to question
Now the other half of the list, where you should slow down and think. Any recommendation that removes your control deserves a question: what do I lose if I accept this? Turning on full automation, handing audience selection entirely to Advantage+, or letting Meta rewrite your creative might improve the score a lot and your results not at all.
Broadening audiences is a good example. For a large account with broad appeal, it often helps. For a niche business with a specific buyer and a tight geography, broadening can mean paying to show ads to thousands of people who were never going to buy. The score rewards the change either way. Your bank account does not.
The same goes for budget suggestions that push you to spend more to "improve delivery." Spending more is not an optimisation, it is spending more. If the extra budget goes to a weak ad set because the score said so, you have just bought a larger loss. The rule of thumb: if a recommendation changes who sees your ads, what they see, or how much you spend, test it on one campaign first and measure the result yourself. If it only fixes something broken, just do it.
Why chasing 100 is a trap
Chasing a perfect score creates three problems. First, it turns your strategy into a to-do list written by the platform that sells you the ads. That is a conflict of interest wearing a helpful badge, and you should treat it as such, politely but firmly.
Second, it burns time and budget on changes with no business case. Every applied recommendation is a change to a live account, and applied all at once in pursuit of 100, they make it impossible to tell which change helped and which hurt.
Third, it teaches your team to optimise the wrong number.
Meta's own reported figure deserves the same clear-eyed treatment. Meta has reported that advertisers who adopted Opportunity score recommendations saw a 12% median decrease in cost per result in its testing. Read that sentence carefully: it is a median across Meta's test group, reported by Meta, under test conditions. It is not a guarantee, not a floor, and not a prediction for your account. It is a reason to test the recommendations seriously, not a reason to apply them all blindly.
How we use it, in practice
Once a month, open the list and sort it into three piles. Pile one: broken basics, fix now. Pile two: plausible improvements, test on one campaign with a clear before-and-after. Pile three: control giveaways and spend increases with no case behind them, ignore for now and revisit if results change.
Apply pile two one change at a time, with enough spend to read the result, and keep notes. That is slower than clicking "apply all," and it is the only way to know what actually worked.
FAQ
What is Meta's Opportunity score?
A 0 to 100 rating in Ads Manager that shows how closely your account follows Meta's recommended setup, plus a list of suggestions to raise the score. It measures adoption of recommendations, not profitability.
Which Opportunity score recommendations should I always take?
Fix measurement problems first: broken tracking, missing events, disconnected catalogues. Those are not strategy debates, they are leaks, and everything else depends on them.
Did Meta really report a 12% cost-per-result improvement?
Meta has reported a 12% median decrease in cost per result for advertisers who adopted its recommendations in testing. That is a reported median from Meta's own tests, not a promise or a guarantee for any individual account.
Should I aim for a score of 100?
No. Chasing 100 means applying every recommendation, including ones that may not suit your business, and changing so much at once that you cannot tell what helped. Aim for better results, and let the score be whatever it is.
MetaMaxd reads Meta's recommendations the way you should: useful input, never orders. If your agency reports your Opportunity score like a grade, ask them what it did to your cost per result.
