ConversionHealth · built into CBO
ConversionHealth tests whether the signals guiding paid-media decisions are present, recent, and comparable—then shows exactly how each issue changes recommendation confidence.
Measurement before optimization
Conversion-led bidding and budget recommendations inherit the quality of the events beneath them. A missing primary action, a week of spend with no recorded outcomes, or an unavailable platform connection can make an account look inefficient when the real failure is measurement. ConversionHealth separates those conditions before they become budget conclusions.
Inside ConversionHealth
ConversionHealth turns read-only platform evidence into a confidence score, a severity-ranked finding list, and a practical fix order.
Critical gaps and platform-level confidence remain visible before anyone interprets performance.
Seven focused checks
The built-in diagnostic is intentionally narrower than a full attribution audit. It concentrates on observable conditions that affect CBO’s ability to use conversion data responsibly.
Recent activity is checked against active conversion actions and current spend so a tracking break does not masquerade as a performance collapse.
Lead-generation accounts often collect value after the website event. ConversionHealth distinguishes online activity from imported or offline evidence.
Attribution choices change what each platform claims. The diagnostic surfaces different click-through windows before cross-platform CPA or ROAS is treated as comparable.
A score without fresh platform data is not presented as certainty. Missing, stale, and coming-soon states stay visible in the confidence layer.
Confidence, not false precision
Each tier has an explicit operating consequence. A lower score does not merely change a badge; it narrows or pauses the recommendations that depend on trustworthy conversion evidence.
Scorable platforms have fresh data with no score-bearing issue severe enough to limit normal budget recommendations.
CBO responseUse conversion-led recommendations with the normal human review, while reviewing any separately disclosed platform exception.
Tracking issues may weaken CPA, CPL, or ROAS interpretation on one or more connected platforms.
CBO responseKeep pacing conservative and review the affected signal before a material reallocation.
Critical signal conditions make conversion-led recommendations unreliable.
CBO responsePause CPL-based decisions for the affected platform until the evidence is restored.
No recent usable tracking result is available, or the latest result is stale.
CBO responseTreat recommendations as directional estimates and refresh the diagnostic.
Each scorable platform begins at 100. High, medium, and low score-bearing findings apply documented deductions. The account score is then calculated as a spend-weighted average across connected platforms with usable recent data. A platform with unavailable conversion data is identified separately and excluded from that weighted average.
Offline-conversion coverage, online-only lead signals, and enhanced-conversion opportunities are currently advisory inside CBO. They can materially improve measurement quality, but they do not always make budget pacing unsafe. ConversionHealth labels them separately so a useful opportunity does not read like a confirmed tracking failure.
The evidence chain
The interface keeps the raw observation, the diagnostic interpretation, and the decision boundary together. That gives an analyst enough context to challenge the finding instead of asking the team to trust an unexplained score.
The platform connection is active, but recent conversion reporting does not contain a usable primary outcome.
The finding is high severity because spend continued while the signal disappeared across the review window.
CBO stops presenting CPL-based guidance as reliable until the signal returns and a fresh check confirms it.
Prioritized findings
Severity reflects the likely effect on recommendation safety. Evidence windows and platform context help the owner distinguish a real break from low volume, normal reporting delay, or an optional improvement.
Recent spend crossed the diagnostic threshold while enabled primary actions returned no recorded conversions.
CPA and CPL recommendations are estimates until tracking is restored.
The action remains enabled but did not fire during the volume-adjusted review window.
Platform optimization may be learning from an incomplete outcome set.
A server-assisted signal may improve resilience where browser-only tracking loses events.
Useful improvement; it does not reduce the current score by itself.
Platform-aware by design
ConversionHealth does not flatten unlike systems into a single generic checklist. Each finding retains its source platform, available fields, reporting window, and connection state.
Conversion-action status, primary actions, origins, attribution windows, recent conversion volume, spend, and enhanced-conversion availability.
Account conversion actions and recent performance evidence, with platform-specific CAPI opportunity context where available.
UET conversion goals, conversion reporting, recent spend, and enhanced-conversion context from the connected account.
The confidence interface reserves a transparent coming-soon state until the read flow can provide a supported score.
Two different health questions
Read-only checks for signal continuity, outcome completeness, comparison integrity, and data freshness.
Weighted campaign scoring that uses the latest ConversionHealth result inside its measurement-confidence pillar.
Explore CampaignHealthRead-only workflow
The diagnostic never applies an external account change. Findings remain advisory until the account owner investigates and acts in the platform or implementation environment.
Authorize a supported ad account for read-only reporting access.
Collect recent conversion actions, volume, spend, settings, and availability.
Score usable platforms and sort findings by severity and budget impact.
Investigate outside CBO, then rerun the check to confirm the signal recovered.
Responsible scope
Strong diagnostic design includes visible limits. The built-in layer does not call every measurement imperfection a blocker, and it does not extend beyond the evidence returned by connected platforms.
In-house marketers can run ConversionHealth before a budget review to confirm that recent efficiency changes are supported by usable outcome data. Agencies can use the confidence tier to explain why a recommendation is normal, limited, fix-first, or estimate-only. Portfolio leaders can prioritize tracking investigations across accounts without opening every platform first.
ConversionHealth is not a tag debugger, attribution-model comparison, incrementality study, consent audit, or guarantee that every reported conversion is accurate. It cannot verify business quality that never reaches the ad platform. A healthy score means the measured platform evidence can support CBO’s budget workflow with stronger confidence—not that the full measurement stack is perfect.
Primary measurement references
Connected platform reports remain the source of record. ConversionHealth interprets the available evidence for CBO’s budget-safety workflow and keeps platform reporting limitations visible.
Straight answers
What teams need to know before using tracking confidence in a campaign budget decision.
ConversionHealth checks whether connected advertising platforms are returning usable conversion evidence for budget decisions. Current checks cover missing offline conversions, online-only lead signals, enabled actions with zero recent events, spend with no conversions, attribution-window mismatches, enhanced-conversion opportunities, and unavailable platform data.
No. The ConversionHealth layer inside CBO is read-only. It retrieves supported account data, calculates confidence, and explains findings. It never edits tracking, campaign settings, conversion goals, or budgets.
Each scorable platform starts at 100 and receives deductions for score-bearing findings based on severity. The account score is a spend-weighted average across platforms with usable conversion data. Informational opportunities do not reduce the score, and an unavailable platform is excluded from that average and shown separately.
ConversionHealth asks whether conversion evidence is trustworthy enough to use. CampaignHealth asks whether campaigns are structured, funded, measured, and performing well enough to support the next investment decision. CampaignHealth uses the latest ConversionHealth result as part of its measurement-confidence pillar.
No. ConversionHealth inside CBO is a focused confidence layer for budget safety, not a complete attribution, tag-management, incrementality, or analytics implementation audit. It prioritizes the signal conditions that can materially affect pacing and allocation decisions.