Lead scoring calculator

Lead scoring: the single highest-leverage activity most B2B teams skip

Lead scoring is a crude ML model that most marketing teams should run in a spreadsheet before they try to build anything sophisticated. The goal: rank incoming leads so SDRs call the top 20% first. The dirty secret: most B2B orgs with a "lead scoring model" in HubSpot or Salesforce have one that was built once, never validated, and is ignored by the SDR team — who intuitively pattern-match their own criteria anyway. The result: leads sit in the CRM for 12–48 hours while the SDR eats lunch and calls whoever happens to be on their list.

According to InsideSales and Harvard Business Review research, contacting a qualified lead in the first 5 minutes increases conversion probability by 9x vs. contacting after 30 minutes, and by 100x vs. contacting after 24 hours. Speed-to-lead × accurate prioritization is a 10–30% pipeline multiplier for almost every B2B org. This calculator gives you the scoring framework; the hard work is the SLA and enforcement.

The fit + intent + engagement framework

Every functional lead-scoring model has three orthogonal dimensions:

1. Fit (firmographic match). Does this company look like your ICP? Industry, size, geography, tech stack, role of the contact. Static attributes — you can score on sign-up.
2. Intent (buying signals). Are they researching a solution right now? G2/Capterra reviews viewed, competitor comparison pages visited, Bombora/6sense third-party intent data, search activity. Time-sensitive — scores decay if not acted on in weeks.
3. Engagement (relationship depth). How deeply have they engaged with your content? Email opens/clicks, demo requests, pricing page visits, webinar attendance. Accumulative.

A lead with high fit + high intent + low engagement is your SDR's top priority. High fit + low intent + low engagement is a nurture candidate. Low fit with any combination usually gets suppressed. Most teams that complain about "low-quality leads" are mixing up fit problems (marketing targeting failure) with engagement problems (sales follow-up failure).

A starter scoring model you can ship this week

Before you build anything complex, build this. Total score = fit + intent + engagement (each capped at 100 points).

• Fit (0–100): Title match 30, industry match 25, company size match 25, geography match 20. Minus 50 if title is student/intern/competitor.
• Intent (0–100): Pricing page visit 30, comparison/alternative page 25, demo request 30, G2/Capterra research (via intent data) 15.
• Engagement (0–100): Email open 3 (capped at 15), email click 5 (capped at 25), webinar attendance 15, content download 10, 2+ return visits 15, direct chat 20.

Band the totals: 200+ = A (contact within 5 minutes), 130–199 = B (contact same day), 70–129 = C (nurture, contact within week), <70 = D (nurture only). Validate against actual pipeline contribution after 90 days and adjust weights.

Data enrichment: the unlock nobody talks about

Your lead scoring is only as good as your data. A form that asks only "email" gives you almost nothing to score on. Enrichment services (Clearbit, Apollo, ZoomInfo, Lusha, 6sense) append firmographic data (company size, industry, revenue, tech stack) to raw email addresses with 60–85% match rates. Budget $600–4,000/month for enrichment on a typical SMB–midmarket B2B team. It pays for itself almost immediately because enriched scoring routes leads correctly instead of SDRs wasting time on intern signups.

Intent data: the newer input most teams underuse

Third-party intent platforms (Bombora, 6sense, Demandbase, G2 Buyer Intent, Clearbit Reveal) monitor buying-signal activity across the web and surface accounts that are researching your category — even if they haven't visited your site. A typical B2B team integrating intent data sees 20–40% more qualified opportunities surfaced in a 90-day pilot. The cost ($1,500–8,000/month depending on scope) is usually justified at any ACV above $20K.

Use intent as a score boost, not a primary trigger. An account showing high "CRM software" buying intent but no fit match is still not your customer. Combine intent with fit filters.

The SDR feedback loop

Every week, have SDRs flag leads that scored high but turned out low-quality (and vice versa). These disagreements are where your model learns. Without this feedback loop, your score drifts over quarters as the market and your ICP evolve, and SDRs lose trust in the model. Review model performance quarterly, adjust weights, retrain if needed.

Automated vs. manual scoring

Rule-based scoring (this calculator's approach) is what 80% of B2B teams should use. It's explainable, auditable, and improvable. AI/ML predictive scoring (HubSpot Predictive Lead Scoring, Salesforce Einstein, MadKudu) works great once you have 10,000+ closed-won deals with clean attribution — until then, the model has too little signal and produces magical-looking outputs that don't actually improve conversion. Start with rules, graduate to ML when data volume justifies.

Three lead-scoring archetypes with full pipeline math

Archetype 1: SMB B2B SaaS ($6k ACV, $80k MRR target)

1,400 MQLs/month via content, paid, and webinar. Fit-intent-engagement scoring in HubSpot Marketing Pro ($800/month) with Apollo.io enrichment ($99/user/month × 3 seats). Scoring bands: A (200+) = 140/month (10% of MQLs), B (130-199) = 420/month (30%), C (70-129) = 560/month, D = 280/month. SDR coverage: 2 SDRs handle A tier within 5 minutes (webhook to Twilio + Aircall), B tier within 24 hours, C tier receives automated nurture. Outcome: A-tier converts MQL to SQL at 42%, B-tier at 18%, C-tier at 4%. Monthly SQLs: 59 from A + 76 from B + 22 from C = 157 SQLs. At 34% close rate × $6k ACV = 53 new customers × $6k = $318k new ARR/month. Without scoring, same SDRs working the MQL list round-robin would hit 95 SQLs, 32 customers = $192k ARR — the scoring model adds $1.5M/year in ARR from the same top-of-funnel volume.

Archetype 2: Enterprise B2B ($180k ACV, 14-month sales cycle)

220 MQLs/month, but only 28 of them are in-ICP after fit filtering. Scoring layers: fit (firmographic match via ZoomInfo at $18k/year), intent (6sense at $85k/year tracking category research + competitor visits), engagement (HubSpot Enterprise at $3,600/month). Any MQL scoring above 180 with a 6sense "Decision" stage intent rating gets routed to a named AE within 15 minutes. Annual outcome from 336 MQLs (in-ICP, year): 54 opportunities, 14 closed-won × $180k = $2.52M new ARR. 6sense + ZoomInfo + HubSpot stack costs $147k/year — payback under 3 weeks on net-new closed revenue.

Archetype 3: PLG SaaS with in-product scoring ($29/month starter)

15,000 trial signups/month. Traditional MQL scoring doesn't fit — instead, product-qualified-lead (PQL) scoring based on activation + depth signals. Custom score in Mixpanel + Reverse-ETL via Hightouch ($350/month) sync to HubSpot. Signals: workspace invites 3+ users (+30), connected to a data source (+25), created 5+ assets in week 1 (+20), hit feature X (+35). PQLs scoring 75+ get a PLS (product-led sales) touch from a 3-person sales-assist team. Outcome: of 15k trials, ~900 PQLs/month, 18% convert to paid within 30 days = 162 paid × $29 = $4,700 MRR/month incremental beyond self-serve. At 12-month LTV of $340 contribution each, $55k annual LTV lift on top of self-serve baseline.

Lead-scoring stack reference, April 2026

Decision framework: when to upgrade scoring sophistication

Start with rule-based scoring in a spreadsheet or HubSpot native tool. Upgrade to an enrichment layer (Apollo, ZoomInfo, Clearbit) when you hit 500+ MQLs/month — below that the $5k-$18k/year investment does not pay for itself on routing accuracy alone. Upgrade to intent data (6sense, Bombora, G2) when ACV crosses $20k and sales cycle exceeds 60 days — intent accelerates the front half of long cycles. Upgrade to predictive ML (MadKudu, HubSpot Predictive, Salesforce Einstein) only after you have 10k+ closed-won deals with clean field completeness — otherwise the model hallucinates signal. If you skip these stages, you end up paying for tooling your SDR team ignores; if you stay too long in the early stage, you cap pipeline growth at whatever a manual model can handle.