← Back to blog
Topic: Meta AdsCategory: Account Optimization8 min read2026-10-21

Should you follow every Meta Ads opportunity score recommendation?

Learn what the Meta Ads opportunity score measures, why a low score is not the same as poor performance, and how to test recommendations without disrupting profitable campaigns.

Performance marketer comparing two abstract optimization recommendation panels before choosing an action.

Quick answer

Treat Meta's opportunity score as a review queue, not a target. Check each recommendation against your objective, constraints, and qualified business results before applying, testing, or dismissing it.

Quick answer: use the score as a review queue, not a goal

No. A Meta Ads opportunity score can surface useful best-practice recommendations, but reaching 100 should not outrank profitable, qualified business outcomes. Review every recommendation, identify the setting it would change, and decide whether to apply it, test it in a controlled way, or keep the current setup for a documented reason.

Meta describes opportunity score as a 0-to-100 score paired with recommendations intended to improve campaign performance. Meta also reports positive aggregate test results from advertisers who adopted recommendations, but that does not prove every recommendation will improve every account. Your decision still needs account-specific evidence.

  • User symptom: Ads Manager shows a low opportunity score or a list of recommendations that conflict with the current strategy.
  • Control level: account-wide recommendations and the campaign, ad set, ad, audience, placement, budget, goal, or creative setting each recommendation would change.
  • Time grain: a point-in-time review followed by a mature comparison period after any accepted change.
  • Likely mechanism: the score reflects unadopted Meta recommendations, not a direct grade of revenue, profit, lead quality, or incrementality.
  • Decision produced: apply a low-risk correction, test a plausible recommendation, or dismiss it with evidence.

What the opportunity score does and does not tell you

The opportunity score summarizes whether an ad account is following Meta's current recommendations. The associated suggestions can cover objectives and performance goals, audience and account structure, creative and placements, budget, and other delivery settings. A higher score means fewer outstanding recommendations; it does not by itself establish that the account is commercially healthy.

Keep three statements separate. Verified platform fact: Meta offers recommendations and assigns them an opportunity score. Diagnostic hypothesis: one ignored recommendation may explain a delivery or efficiency problem. Practitioner decision: whether that recommendation fits this account's economics, measurement quality, creative constraints, and testing plan.

  • A score of 100 means the currently scored recommendations have been addressed; it is not a promise of strong CPA, ROAS, margin, or lead quality.
  • A low score can coexist with good results when the account intentionally avoids a recommendation that conflicts with a valid constraint.
  • A high score can coexist with poor results when the offer, measurement, creative, destination, or downstream economics are weak.
  • Recommendation points should not be translated into an invented percentage improvement for your account.

Triage recommendations by the change they would make

Do not accept or reject the list as one package. Record each recommendation, the object it affects, the exact before-and-after setting, the claimed benefit, and the business guardrail it could disturb. This turns a score-chasing exercise into a change-control process.

Start with clear configuration defects that have low strategic ambiguity, then move to recommendations that change auction participation or measurement. A suggestion to repair an incomplete asset is not the same decision as broadening an audience, changing a performance goal, consolidating ad sets, enabling more placements, or allowing an automatic adjustment.

  • Measurement: could the change alter the selected event, attribution context, event quality, or how success is counted?
  • Delivery: could it expand audiences or placements, consolidate structure, change budget allocation, or alter bidding behavior?
  • Creative: could it introduce automated variations, crop assets, add enhancements, or change the approved presentation?
  • Business constraints: could it weaken geography, compliance, inventory, margin, lead-quality, or brand requirements?
  • Reversibility: can you restore the prior state with IDs, settings, and timestamps preserved?

Build a baseline before accepting a recommendation

Freeze the reporting context before making a change: account and object IDs, completed date range, account time zone, attribution setting, optimization event, columns, filters, breakdown state, spend, impressions, outbound clicks, landing page views, selected results, and the qualified outcome that matters to the business. Save the recommendation wording and score at the same time.

Use a mature baseline that represents the current setup rather than the first hours of a campaign. For ecommerce, connect platform purchases and value to orders, refunds, margin, and new-customer status where available. For lead generation, connect platform leads to unique valid records, contactability, qualification, opportunity creation, and sales.

  • Preserve immutable campaign, ad set, ad, and creative IDs.
  • Record the recommendation, point value, affected setting, and observation time.
  • Define one primary outcome and the guardrail metrics that must not deteriorate.
  • Check recent edits so an existing change is not mistaken for the recommendation's effect.
  • Separate Meta-reported results from downstream business outcomes.

Apply, test, or dismiss each recommendation

Apply a recommendation directly when it corrects a verified defect, respects the campaign's constraints, and has a clear rollback. Test it when the mechanism is plausible but the account-specific impact is uncertain. Dismiss or defer it when it conflicts with a documented business rule, changes too many variables, or cannot be evaluated with trustworthy measurement.

When testing, change one responsible layer, preserve a comparable control where practical, and wait for enough mature outcomes to judge the intended result. Do not call a recommendation successful because the opportunity score increased. Call it successful only when the primary outcome improves without unacceptable damage to the guardrails.

  • Apply: verified configuration repair, low ambiguity, low downside, and clear validation path.
  • Test: plausible delivery improvement with uncertain account-specific effect.
  • Dismiss: conflicts with a hard operational, geographic, compliance, inventory, or brand constraint.
  • Defer: measurement is broken, the cohort is immature, or another active change would contaminate the result.
  • Rollback: qualified economics worsen or the change creates an unintended delivery, creative, or measurement effect.

Do not let automatic adjustments bypass change control

Some recommendation workflows can make it easy to apply several changes quickly. Convenience is not evidence. Before enabling any automatic adjustment, document which settings may change, who owns the decision, what alert or review cadence will expose the change, and which outcomes trigger rollback.

Accounts with strict geography, regulated offers, controlled creative approvals, limited inventory, margin-sensitive product mixes, or qualified-lead requirements need especially explicit guardrails. The goal is not to resist automation; it is to make automated changes observable and accountable.

How this differs from nearby Meta Ads diagnostics

This is an account-recommendation decision path, not a general performance audit. A sudden CPA increase starts from an observed cost change. Campaign competition starts from duplicated auction participation. Learning Limited starts from an ad set's optimization state. Relevance diagnostics compare individual ads. The opportunity-score problem starts with Meta's recommendation queue and asks whether each proposed change deserves adoption.

The closest library guides are the Meta Ads audit checklist, the inconsistent-results diagnostic, the campaign-overlap guide, and the Learning Limited guide. They may supply evidence for a recommendation, but none should be replaced by blindly optimizing the opportunity score.

How an AdSpecIt-style audit helps evaluate the recommendations

An AdSpecIt-style audit can inspect campaign structure, objectives, optimization events, budgets, bid strategy, audiences, placements, creative mix, delivery, and performance patterns to test whether the mechanism behind a recommendation is visible in the account data. It can prioritize recommendations that address a verified constraint and flag those that may conflict with profitable patterns or business guardrails.

The opportunity score and exact recommendation wording still need to be captured from the current Ads Manager interface. The useful combined output is a decision log: recommendation, affected object, supporting evidence, expected mechanism, test design, guardrails, owner, outcome, and rollback decision.

Sources and further reading

Research checked on 2026-10-21. Meta can change recommendation types, scoring, and interface placement, so verify the current wording and settings in your own ad account before applying a change.

  • Meta Business Help Center — About opportunity score in Meta Ads Manager: https://www.facebook.com/business/help/804913634782260
  • Meta for Business — Supercharge your campaign performance with new Advantage+ AI features and experimentally proven recommendations: https://www.facebook.com/business/news/supercharge-performance-advantage-opportunity-score
  • Jon Loomer Digital — Opportunity Score and Campaign Score Explained: https://www.jonloomer.com/opportunity-score-campaign-score/

Keep going with a few more answers on Meta Ads audits, reporting, and performance issues.

Want AdSpecIt to audit your own account?

Connect your Meta Ads account, get a free score, and see the issues most likely to be hurting campaign performance before you spend more.

Get your free audit