How the readiness score is calculated
Deterministic rules, category weights, blocker caps, and a separate complexity rating.
Inputs
The engine reads normalized Salesforce metadata collected after OAuth: objects, fields, relationships, Flows, Apex, packages, and security configuration. Business records are not used. Details are in the security overview.
Category weights
- Data model compatibility — 35%
- Automation compatibility — 30%
- Packages and integrations — 20%
- Security model complexity — 10%
- Platform fit — 5%
Findings and deductions
Matching rules emit findings with severity and score impact. Category scores start from a baseline and deduct based on observed patterns (for example formula density, Apex footprint, or known blocker packages).
Verdict caps
Critical platform blockers can downgrade the final verdict even if the weighted score is higher. Complexity (Low → Very high) is reported separately from compatibility.
Reproducibility
The same metadata + ruleset version should produce the same score. Ruleset version is stored on the scan and shown on the report.
FAQ
- Does AI create the score?
- No. Scores come from versioned rules applied to collected metadata. AI may later explain findings, but it cannot override the score.
- Why can a high score still get a weaker verdict?
- Critical blockers such as CPQ or Experience Cloud can cap the verdict even when the weighted numeric score looks healthier.
Get your readiness score
Connect a Salesforce sandbox, scan metadata only, and receive a private compatibility report for Twenty.