Lead scoring template
Fit and intent as separate axes, with the MQL threshold set from your own conversion data.
- HubSpot
- Salesforce
- GA4
- Google Sheets
Rebuild our lead scoring. Sales says half of our MQLs are junk.
- Analyzed closed-won accounts187 wins vs 2,400 losses; 6 fit attributes predict conversion
- Tested current point values3 of 11 behaviours have no relationship to conversion
- Rebuilt on two axesFit from firmographics, intent from 8 weighted actions with decay
- Calibrated the threshold62 MQLs/week at the new cutoff vs 140 today
Three scored behaviours — newsletter opens, blog visits, and social clicks — show no relationship to conversion and are inflating the score for people who are only reading. On two axes the MQL count drops from 140 to 62 per week, and the 78 leads that fall out are high-intent low-fit, which is a nurture problem rather than a sales one.
A lead scoring template that adds firmographic points to behavioural points produces a number nobody can act on. This one scores fit and intent separately, calibrates the MQL threshold against which leads actually converted for you, and defines decay so a webinar attendee from March does not look hot in August.
What is a lead scoring template?
A model that rates fit — how much a lead looks like your best customers — and intent — what they have done recently — on two separate axes, then defines which combinations route to sales, which go to nurture, and which are disqualified.
Build the fit axis
- Derive attributes from your closed-won accounts
- Weight by how strongly each predicts conversion
- Define hard disqualifiers separately from low scores
Build the intent axis
- Score actions by proximity to a buying decision
- Set decay rates per action type
- Cap repeated low-value actions so they cannot accumulate
Calibrate and route
- Set the MQL threshold against historical conversion
- Match volume above the threshold to sales capacity
- Route each fit-and-intent quadrant explicitly
Why separate fit and intent scoring?
Because they demand different actions. High fit with low intent is a nurture target; low fit with high intent is often a student or a competitor. Summing them into one score hides exactly the distinction sales needs.
How do you set the MQL threshold?
From your own conversion data: find the score above which historical leads actually converted, then set the threshold to the volume sales can genuinely work. A threshold chosen so the current pipeline looks healthy is not a threshold, it is a report.
Which lead scoring criteria should decay?
All behavioural ones. Intent is a statement about the present, so a pricing page visit should be worth much less after three weeks. Fit attributes do not decay, though they should be refreshed when the account data changes.
How the lead scoring template works across your stack
The agent reads closed-won and closed-lost records from HubSpot or Salesforce to derive fit attributes empirically, and pulls behavioural events from your CRM and analytics to weight intent. The model exports as field-level scoring rules plus routing logic, and shares its MQL definition with the sales handoff template so both sides use one threshold.
- HubSpot
- Salesforce
- GA4
- Google Sheets
Who uses this lead scoring template
- RevOps
- Replace inherited point values with weights derived from conversions.
- Demand gen managers
- Route high-fit low-intent leads to nurture instead of burning them on sales.
- Sales leaders
- Get an MQL volume that matches the capacity of the team.
How to run this lead scoring template in Metaflow
Connect your CRM
Closed-won and closed-lost history is what makes the weights empirical.
Review the derived fit attributes
The agent shows which attributes predicted conversion and how strongly.
Set intent weights and decay
Actions scored by proximity to a decision, with decay per action type.
Calibrate the threshold and route
Match MQL volume to sales capacity, then define routing per quadrant.
What you provide
- CRM closed-won and closed-lost history
- Behavioural event data
- Sales capacity for MQLs per week
What you get back
- Fit attributes with weights and disqualifiers
- Intent actions with weights and decay rates
- Calibrated MQL threshold
- Routing rules per quadrant
Why use this lead scoring template?
Fit and intent stay on separate axes, so routing is actionable
Weights are derived from your closed-won data, not guessed
MQL threshold is calibrated to sales capacity
Decay and caps prevent stale or farmed scores
Lead scoring template FAQs
What is a good lead scoring model?
One that separates fit from intent, derives its weights from conversion history, and produces MQL volume the sales team can actually work. If the model cannot be traced to closed-won data, its point values are decoration.
What should the MQL threshold be?
The score above which your historical leads converted at a rate sales considers worth pursuing, bounded by capacity. There is no universal number, and copying one from another company is how you get a queue nobody works.
Should lead scores decay?
Behavioural scores should, because intent is about the present. Fit scores should not decay, but they should be refreshed when firmographic data changes.
How often should you recalibrate lead scoring?
Quarterly, and immediately after any change in ICP, pricing, or channel mix. A model tuned to last year’s traffic quietly misroutes this year’s.
Similar email sequence templates
Key takeaways
- Two axes, not one summed score
- Derive weights from closed-won data and bound MQLs by capacity
- Decay behavioural scores; refresh fit attributes