Organisations that implement structured lead scoring see 77% higher lead generation ROI than those that do not score leads. Yet most B2B teams still route every inbound lead directly to sales — burning rep time on prospects who will never buy whilst high-intent buyers wait days for follow-up.
The solution is a scoring model that separates signal from noise. This guide walks you through building a custom lead scoring rubric with specific point values for demographic fit, behavioural intent, and disqualification signals — plus a free Excel template you can deploy in your CRM today.
Whether you are using HubSpot, Salesforce, or Marketo, the framework below gives your sales team a prioritised pipeline based on data, not gut feel. Organisations using this approach report MQL-to-SQL conversion rates above 30%, compared to the 13% industry average.
Effective lead scoring operates on two independent dimensions. Demographic scoring evaluates whether a prospect matches your ideal customer profile — job title, company size, industry, geography, and budget authority. Behavioural scoring measures what a prospect actually does — pages visited, content downloaded, emails opened, demos requested.
The distinction matters because each dimension answers a different question. Demographic scoring asks: "Should we sell to this person?" Behavioural scoring asks: "Should we sell to this person right now?" You need both. A VP of Marketing at a 200-person SaaS company (high demographic fit) who has not visited your site in six months is not ready for sales outreach. Conversely, a student downloading every whitepaper you publish has zero purchase authority.
The two-dimensional approach also makes diagnostics straightforward. If high-scored leads are not converting, you can isolate the problem: is it demographic accuracy (wrong companies reaching sales) or behavioural weighting (engagement signals not predicting intent)?
The following framework is based on patterns from organisations achieving 30%+ MQL-to-SQL conversion rates. Adapt the point values to match your sales team's actual qualification criteria.
Job title (decision-maker): +30 points — VP, Director, or Head of department scores 30. Manager-level scores 20. Individual contributors score 10.
Company size: +30 points — Companies with 50 to 500 employees score 30. Adjacent ranges (20-49 or 501-1000) score 15.
Industry alignment: +25 points — Primary target industry scores 25, adjacent industries score 15.
Demo or pricing request: +40 points — Strongest buying signal. Should always trigger immediate sales routing.
Pricing page visit: +30 points — Active evaluation. Prospect is comparing costs and considering purchase.
Case study download: +25 points — Building internal business case. Strong mid-funnel signal.
Content download: +15 points — Educational interest. Early-stage awareness, not purchase intent.
Competitor employee: -50 points — Researching, not buying. Remove from sales pipeline immediately.
Personal email in B2B: -25 points — Gmail or Yahoo in B2B context signals low authority.
Email unsubscribe: -25 points — Active disengagement. Prospect has rejected your communications.
Many organisations implement only positive scoring whilst ignoring factors that should disqualify a lead. Include negative signals to prevent score inflation.
Download our free Excel template with pre-built scoring rubrics for demographic, behavioural, and negative scoring factors. Includes threshold definitions, SLA recommendations, quarterly review KPIs, and GDPR compliance notes.
Download Free Lead Scoring Template (Excel)
Want AI to score and route your leads automatically? See how Marketing Mary's AI agent integrates with your CRM.
Your scoring rubric is only useful if your CRM can act on it. Here is how the three major platforms handle lead scoring.
Native scoring via the HubSpot Score property. Pro plan includes rule-based scoring; Enterprise adds AI-powered predictive scoring. See our HubSpot AI integration guide for setup details.
Einstein Lead Scoring provides predictive capabilities trained on your historical deal data — requires minimum 1,000 historical leads with won/lost outcomes. Strong for enterprise but complex for SMEs.
Built-in scoring via Smart Campaigns. Separate demographic and behavioural score fields by default — the only major platform with native two-dimensional scoring.
Regardless of platform, the implementation sequence is the same: define scoring criteria with sales, build rules, set threshold-based routing (75+ points triggers immediate sales alert), and establish a quarterly review cadence. If you are running marketing workflow automation, integrate scoring thresholds directly into lead routing workflows.
Warning: Don't Block Hand-Raisers. Any prospect who requests a demo, asks for pricing, or calls sales should bypass scoring thresholds and route immediately. Scoring informs prioritisation — it should never create false gates that filter out genuine buyers.
A scoring model built today becomes unreliable within three to six months. The organisations achieving 39% MQL-to-SQL conversion rates maintain performance through disciplined quarterly recalibration.
The quarterly review: compare highest-scored leads against closed-won deals. If fewer than 25% converted, adjust thresholds. Examine which factors correlated with wins. Check negative scoring. Validate score decay settings — halve behavioural points at six months, zero beyond twelve.
Four KPIs govern model health: MQL-to-SQL conversion rate by score tier, sales acceptance rate, average deal size by score tier, and pipeline velocity by score tier. Track monthly and recalibrate when any metric drifts more than 15% from baseline.
Lead scoring fails when marketing builds the model in isolation. The most common failure is misalignment between marketing's scoring criteria and sales' actual qualification standards.
Start by asking your sales team: "Of the last 20 deals we closed, what did those prospects have in common?" Build your initial model from those patterns, set thresholds jointly, and commit to a 90-day review.
Then operationalise the handoff. High-scored leads (75+ points) receive sales contact within one hour. Mid-scored leads (50-74 points) enter an accelerated nurture sequence. Below 50 points stays in marketing nurture. Configure your CRM to route leads automatically based on score via marketing workflow automation.
Start with 5-7 core criteria that predict 80% of your conversions. Focus on job title, company size, industry alignment, and 3-4 high-intent behavioural signals. Add complexity only when data shows additional criteria improve prediction accuracy.
Rule-based scoring uses manually defined criteria and point values. Predictive scoring uses machine learning to analyse historical conversion data. Rule-based works for organisations with fewer than 500 leads. Predictive requires 500-1,000+ leads with outcomes.
Quarterly at minimum. Compare highest-scored leads against closed-won deals. Also recalibrate when launching new products or entering new markets. Most models drift within 3-6 months.
Yes, but start with rule-based scoring. Interview your salespeople about which characteristics they associate with successful deals. Even a simple 5-criteria model dramatically outperforms no scoring. Graduate to predictive models once you have 500+ leads with outcome data.
GDPR requires explicit consent before tracking behavioural signals. Design your model to function on demographic data alone for non-consenting prospects, with behavioural signals as enhancement for those who opt in.
Marketing Mary's AI Co-Pilot automates lead scoring, content creation, and campaign workflows — so your team focuses on strategy, not spreadsheets.
Join 500+ Marketing Leaders on the Waitlist | Book a personalised demo
Sources: Landbase 2025, Apollo.io 2025, Verse.ai 2025, Prospeo 2026, HubSpot 2025
Clwyd Probert — Founder, Marketing Mary. Clwyd builds AI-powered marketing tools that help SME teams escape the content treadmill.
Clwyd Probert: May 5, 2026 11:00:00 AM
How to Run a MarTech Stack Audit: Framework, Scoring Template, and ROI Calculator A MarTech stack audit systematically evaluates every marketing...
Clwyd Probert: May 5, 2026 10:00:00 AM
The best buyer persona tool in 2026 depends entirely on how you plan to use your personas — and most comparison articles won't tell you that. We...
Clwyd Probert: May 4, 2026 11:00:00 AM
Key Takeaway Building a B2B martech stack that works means starting with an architecture decision, not a tool decision. The average B2B team runs...
1 min read
Clwyd Probert : May 5, 2026 10:00:00 AM
The best buyer persona tool in 2026 depends entirely on how you plan to use your personas — and most comparison articles won't tell you that. We...
1 min read
Clwyd Probert : May 5, 2026 11:00:00 AM
How to Run a MarTech Stack Audit: Framework, Scoring Template, and ROI Calculator A MarTech stack audit systematically evaluates every marketing...
1 min read
Clwyd Probert : Mar 23, 2026 12:00:00 PM
The Complete Guide to Marketing Workflow Automation in 2026 Marketing teams are bleeding time. Between tool-switching, manual data transfers, and...