Your lead scoring rules are guessing. Your closed-won deals know the truth—and you're not listening to them.
Lead scoring drift is silent revenue loss. Marketing hands off leads that look good on paper. Sales calls them garbage. RevOps has no way to prove who's right. Here's what happens: You built scoring rules six months ago based on assumptions. A prospect filled out a form, visited pricing twice, downloaded a case study—boom, 85 points, MQL. But your closed-won data shows that prospects who actually bought did something different. They visited your ROI calculator. They spent 8+ minutes on the integration page. They came back three times in one week. You're scoring for activity. You should be scoring for intent. This workflow forces a monthly reckoning. Pull your last 30-60 closed-won deals. Extract the actual attributes: which pages did they visit? How many touches before first meeting? What company size? What industry? What job title? Which content pieces did they consume? Then compare that fingerprint to your current scoring rules. The gap is your calibration problem. The output: a revised scoring model that reflects reality, not assumptions. Sales gets hotter leads. Marketing stops chasing phantom signals. RevOps has a repeatable process that runs every 30 days. This isn't a one-time audit. Buying signals shift. Your rules have to shift with them. The workflow includes the approval gates—sales leader validates the signal, marketing confirms the data source, RevOps locks the new rules into your CRM. If you skip the approval step, you'll recalibrate into a worse place.
Trigger: Monthly, on the first Monday. Or: whenever sales leadership reports that MQL quality has dropped below your baseline (e.g., 'only 25% of MQLs are taking meetings').
Query CRM for deals closed in the past 90–180 days. Segment by deal size, sales cycle length, and source. Document data completeness: flag records with missing lead source, engagement history, or scoring attributes. Target minimum 50 deals; 100+ preferred for statistical confidence.
For each closed-won deal, extract: company size, industry, job title of buyer, engagement frequency (emails opened, pages visited, demo attended), lead source, time-to-first-touch, and any custom attributes in your scoring model. Create a frequency table: which attributes appear in 70%+ of wins? Which appear in <20%?
Map each winning attribute to your current scoring rule. Identify: (a) high-value attributes underweighted in current model, (b) low-value attributes overweighted, (c) attributes not in model but present in 50%+ of wins, (d) attributes in model but absent from wins. Document discrepancies with examples.
Present attribute analysis to VP Sales and top closers. Ask: Do these patterns match your experience? Which attributes felt most predictive of deal velocity and size? Capture qualitative feedback on attributes that felt like noise. Resolve disagreements between data and field intuition.
Adjust point values based on win-rate frequency and sales input. Example: if 85% of wins include a demo, increase demo attendance from 5 to 15 points. If job title 'VP Engineering' appears in 60% of wins but scores only 3 points, increase to 10. Keep total possible score stable (e.g., 100 points) to avoid threshold creep. Document rationale for each change.
Apply new scoring rules to the closed-won cohort and a sample of closed-lost deals. Calculate: (a) % of closed-won deals now scoring above your MQL threshold, (b) % of closed-lost deals now scoring below threshold, (c) distribution shift. Target: 85%+ of wins above threshold, <30% of losses above threshold. If results are poor, iterate step 5.
Define score ranges: e.g., 0–30 = nurture, 31–60 = MQL (marketing-qualified), 61–100 = SQL (sales-qualified). Base thresholds on backfill results and sales capacity. Document the business logic: 'We convert MQL to SQL at 61 points because historical data shows 70% of deals scoring 61+ close within 6 months.'
Deploy recalibrated scoring to a subset of incoming leads (e.g., one region, one industry vertical, or 20% of volume). Run for 2–4 weeks. Track: lead volume, MQL-to-SQL conversion rate, average deal size, and sales feedback. Compare against control group using previous rules.
Analyze pilot data: Did MQL quality improve (higher conversion rate, larger deals)? Did lead volume drop significantly? Did sales report better lead fit? If conversion rate improved 10%+ and volume held steady, proceed to full rollout. If results are mixed, identify which new rules underperformed and refine.
Deploy new scoring rules to all incoming leads in your CRM and marketing automation platform. Update any downstream automations (lead routing, nurture workflows, sales alerts) to reflect new thresholds. Notify sales and marketing teams of changes and new lead quality expectations.
Create a single-source-of-truth document: scoring rules, point allocations, thresholds, rationale, and last calibration date. Assign ownership (typically RevOps lead). Define review cadence (quarterly recommended). Include decision log: what changed, why, and by whom.
Set up monthly dashboards tracking: MQL-to-SQL conversion rate, average deal size by lead source, time-to-close by initial score, and sales feedback sentiment. If conversion rate drops >15% or deal size declines >10%, trigger an unscheduled recalibration. Schedule next full calibration in 90 days.
1. Export your last 60 closed-won deals from your CRM. Include: lead source, company size, industry, job title, pages visited, time on site, form submissions, email opens, sales cycle length, deal size. 2. Create a simple scoring audit sheet (Google Sheets or Airtable). Columns: Deal ID | Lead Source | Pages Visited | Touches Before First Meeting | Days from First Touch to Close | Deal Size | Current Score at MQL | Actual Conversion Signal. 3. For each closed-won deal, manually review the activity history. Mark which attributes were present: pricing page visit (Y/N), ROI calculator use (Y/N), integration docs (Y/N), 3+ touches in 7 days (Y/N), etc. 4. Segment the closed-won deals by conversion pattern. Group A: quick converters (7-14 days). Group B: long cycles (30+ days). Group C: high-touch (5+ sales touches). Note which scoring attributes were common in each group. 5. Compare to your current scoring rules. Where are the gaps? Are you scoring for things that don't convert? Are you missing signals that do? 6. Draft new rules: e.g., 'ROI calculator use = +30 points (was: pricing page = +15 points)' or 'Integration docs visit + 2+ touches in 7 days = auto-qualify'. 7. Run the new rules against your last 30 MQLs. Would they have scored higher or lower? Would you have qualified more or fewer? Share this simulation with sales leadership.