RFRevFleet
RevFleet scoring methodology

How RevFleet scores a HubSpot portal

RevFleet evaluates 53 deterministic checks across eight operating areas. The score is a repeatable health signal, not a black box or a generic AI opinion.

The calculation

Severity first, weighted average second.

Each rule evaluates a measured HubSpot signal and returns one state. Rule weights produce a module score; module weights produce the global score from 0 to 100.

Good = 100
The measured signal meets its rule threshold.
Warning = 50
The signal needs review or additional context.
Critical = 0
The signal crosses the rule's critical threshold.

If an entire module cannot be collected completely, its checks remain visible as evidence gaps and are excluded from the score. A rule-level evaluation error also remains visible as “Could not evaluate” and is excluded rather than being treated as negative portal evidence.

Audit coverage

Eight modules, 53 deterministic checks.

Global score = weighted available-module average
CRM Core
Data Quality

Record quality, associations, properties, and deal hygiene.

14 checks25% weight
Marketing
Portal Health

Email health, forms, lists, and workflow safety.

11 checks15% weight
Sales
Portal Health

Pipelines, activity, forecasting, and sales process coverage.

6 checks20% weight
Service
Portal Health

Tickets, ownership, SLA signals, knowledge base, and feedback.

5 checks10% weight
CMS
Portal Health

Publishing, redirect, and template operation signals.

3 checks5% weight
Operations
Operations / Access

Integrations, data-quality rules, and automation governance.

6 checks10% weight
Settings
Operations / Access

Users, teams, admin risk, legal basis, and brand setup.

5 checks10% weight
AI Readiness
AI Readiness

Data foundation, business context, and governance guardrails.

3 checks5% weight
What the score means

A consistent operating baseline.

  • Compare the same portal against the same thresholds over time.
  • Locate the module and measured condition behind a critical or warning result.
  • Start a consultant or RevOps review with repeatable evidence instead of a blank checklist.
What it does not claim

Useful, not magical.

  • This is not a market percentile or a certification of a portal.
  • Thresholds may need calibration when a portal has multiple countries, brands, teams, or operating models.
  • AI does not calculate this score. BYOK AI can explain evidence after deterministic scoring, never replace it.
Privacy and control

Read-only audit by default.

No writes are made to HubSpot during a standard RevFleet scan. The Chrome extension keeps the Private App token browser-side and sends aggregate audit evidence to RevFleet. Public scorecards are optional, expiring, and aggregate-only.