LeanScale Scoring Pro

Audit your lead-scoring model against LeanScale's canonical methodology. Sales-led (MQL) and product-led (PQL) covered.

Two skills bundled:

• lead-scoring-audit-sales-led — audit Fit Score + Engagement Score + thresholds against MQL→SQL conversion. Surfaces weight imbalances, missing negative scoring, time-decay gaps, threshold drift.
• lead-scoring-audit-product-led — audit PQL scoring (product usage signals + activation milestones + firmographic fit) against trial-to-paid conversion. Surfaces fit-dominated scoring, activation gaps, sync lag, threshold miscalibration.

What you get

Each skill takes your CRM/MAP scoring config + a sample of leads with conversion outcomes, then produces:

• 7-dimension (sales-led) or 8-dimension (PLG) scorecard — graded 0–10 each
• Score distribution chart — count + conversion rate per band
• Predictive lift validation — closed-won median score vs closed-lost median (target ≥1.4×)
• Specific weight adjustments — exact criteria to add, retire, or reweight
• Threshold recommendations — backed by the customer's own data

Why this is different

• Real methodology, not generic. Built on LeanScale playbooks #37 (sales-led) and #38 (product-led) — the same frameworks LeanScale uses across 100+ B2B SaaS engagements.
• No autonomous changes. The skills audit and recommend; humans configure.
• Reliability gate. ≥90% task completion + no-fabrication rules. We are explicitly not 11x.ai.
• Calibrated to your data. Score distributions and lift calculations come from the customer's actual leads, not benchmarks.

Install

/plugin install github:LeanScaleTeam/scoring-pro

Usage

Sales-led MQL audit

> Run a lead-scoring audit on hubspot-leads.csv. Use the scoring config from scoring-config.json. 
  We're a B2B SaaS sales-led company with MQL threshold currently at 50.

Product-led PQL audit

> Run a PQL audit on trial-users.csv. Product analytics export is at mixpanel-export.json. 
  Activation event is "first integration completed". Current PQL threshold is 60.

What the skills do NOT do

• ❌ Redesign the model from scratch (Sprint engagement; project #37 or #38)
• ❌ Configure changes in the CRM/MAP
• ❌ Build predictive ML models — rules-based scoring audit only
• ❌ Audit the lifecycle stages themselves (use leanscale/lifecycle-pro)

Sister plugins

• leanscale/lifecycle-pro — Lead / Sales / Customer Lifecycle audits
• leanscale/migration-pro — CRM migration readiness (HS↔SF)

Pricing

Free through the Claude marketplace. The skills produce reports and recommend embedded engagements where structural gaps exceed what hygiene can fix.

If your audit reveals the model needs a full redesign (8 of 10 audits do), the recommendation surfaces a LeanScale Sprint — $50K, 6-week paid pilot that installs the scoring methodology end-to-end.

License

MIT — see LICENSE

---

Built by LeanScale — the agentic GTM agency.