How to Implement Lead Scoring: The Fit + Intent Model

Build a lead scoring model that combines fit (firmographics) and intent (behavior) so your team spends its time on the leads most likely to convert. This guide covers the point framework, the thresholds, the decay logic, and the CRM setup for both HubSpot and Salesforce. It is the hub of a five-part cluster: the ready-to-paste HubSpot lead-scoring template, the Salesforce lead-scoring template, a deep dive on how score decay works, and how to turn scores into a prioritized BDR queue.

Why does lead scoring matter?

Lead scoring matters because it tells reps which leads to work first. Teams that qualify by gut feel tend to spread effort evenly across leads of wildly different quality, while a calibrated model concentrates the fastest response on the leads most likely to close. Fit-only scoring tends to miss high-intent buyers outside your stated ICP, and intent-only scoring tends to flood sales with unqualified browsers, so effective models combine both. The claims below are directional, drawn from widely cited lead-management research and our own engagements, not a controlled study.

• Commonly cited figures link mature lead scoring to higher MQL-to-SQL conversion than no scoring at all (directional, not a guarantee).
• Reps who lack a way to prioritize tend to lose a large share of their time to unqualified leads. The AI GTM Engineer can price that cost for your own funnel.
• Fit-only scoring tends to overlook high-intent buyers who do not match the ICP on paper.
• Intent-only scoring tends to route unqualified leads to sales, since raw activity does not confirm budget or authority.

The scoring tiers at a glance

The table below shows a typical tier structure. The conversion column is illustrative: it reflects the direction we see in engagements, not a measured guarantee, and you should replace it with your own conversion-by-tier data once you have a month of history.

How do you set up a scoring model?

You implement lead scoring by defining fit criteria (firmographic and technographic signals), engagement signals (website visits, content downloads, email opens), and optional third-party intent data. Assign weighted point values to each signal, set threshold scores for the MQL and SQL stages, and automate routing based on the total score. The six steps below walk through it end to end.

1. Define fit score criteria

Identify firmographic attributes that correlate with closed-won deals. These rarely change and indicate alignment with your ideal customer profile. The point values below are an illustrative framework capped at 50 points; tune each weight against your own win-rate data.

• Company size (0 to 10 points): score your sweet-spot employee band highest, adjacent bands lower, and clearly-too-small companies at 0.
• Industry (0 to 10 points): give your core ICP industries the full weight, adjacent verticals a partial score, and low-fit sectors 0.
• Revenue range (0 to 10 points): score the ARR band where you close fastest and retain longest highest. Your sweet spot is usually not the largest companies.
• Tech stack (0 to 10 points): reward stacks you integrate with cleanly (for example Salesforce or HubSpot) and discount legacy or no-CRM setups.
• Growth stage (0 to 10 points): weight the funding or maturity stage that maps to active buying, and score pre-revenue at 0.

How to apply revenue-range scoring in your CRM: map company ARR to point values using your historical win-rate data, ordering the rules from highest-value band to lowest. CRM scoring rules evaluate top to bottom and apply the first match, so the best-fit ranges should be evaluated first. In HubSpot, use workflow branching; in Salesforce, use ordered flow criteria. The HubSpot template and Salesforce template give you the exact fields and rules to paste in.

Pro tip: run win/loss analysis on your recent closed deals, calculate win rate by each attribute, and assign higher points to the attributes with the strongest win-rate correlation. Let the data set the weights.

2. Define intent score criteria

Track behavioral signals that show buying interest. These change frequently and indicate "ready to buy now." The weights below are an illustrative framework capped at 50 points.

• High-intent actions: a demo or trial request, repeat pricing-page visits, a case study or ROI-calculator download, or a completed product demo video. Weight these heavily.
• Medium-intent actions: webinar or event attendance, repeat website visits within a month, multiple email opens, or a whitepaper download.
• Low-intent actions: a single email-link click, a one-off site visit, a blog read, or a single email open. Weight these lightly.
• Negative signals (subtract points): an unsubscribe, a careers-page-only visit (a job seeker, not a buyer), or a spam complaint. These pull the score down.

Illustrative intent calculation: a lead who requests a demo, visits pricing several times, downloads a case study, opens a handful of emails, and returns to the site multiple times would accumulate points across those actions and hit the 50-point cap. Cap repeat low-value actions (for example, count only the first several email opens) so a newsletter reader cannot inflate the score. This example is illustrative, not a measured outcome.

The blind spot in every intent score: most behavioral signals come from anonymous traffic you cannot attach to a lead record. Page views, pricing visits, and repeat sessions do not score if you do not know who is behind them. De-anonymizing site visitors is what gives the intent model actual input. See the de-anonymization ROI benchmark for how that surface is sized.

3. Set scoring thresholds

Combine fit plus intent into a total 0-100 score, then define tiers to route leads into the right stage of your sales process. The tiers and actions below are a starting structure; the conversion notes are directional.

• Hot lead (80 to 100): high fit plus high intent. Route to sales immediately with a Slack alert and a sub-5-minute response target. This is where speed matters most.
• Warm lead (60 to 79): moderate fit and intent, or high fit with low intent. Add to a nurture sequence and follow up within a day.
• Cold lead (40 to 59): low fit or low intent. Keep in marketing nurture, no sales contact yet, and watch for the score to climb.
• Disqualified (0 to 39): poor fit or negative signals. Suppress from outreach and archive.

Routing rules do not matter if no one acts on them. The widely cited Lead Response Management study (Oldroyd, 2007) reported roughly a 21x lift in the odds of qualifying a lead when an inbound inquiry is worked within 5 minutes rather than after half an hour, which is why hot leads need an automated handoff rather than a Slack-then-forget. Turning those tiers into an actual work order is the job of BDR queue routing from lead scores.

4. Configure score decay

Reduce the intent score over time so stale leads do not stay "hot" forever. Without decay, a lead who visited pricing months ago still reads as hot today, which sends reps chasing interest that has gone cold.

• Intent score decay: a common rule is to reduce the intent score by roughly 10% per week of no activity, so it approaches zero after several inactive weeks. Any new engagement resets the decay and adds points.
• Fit score (no decay): keep it constant unless firmographics change. Refresh it only when enrichment data updates, since company size, revenue, and industry rarely move.

Illustrative decay math: a lead sitting at a 40-point intent score with no activity would step down week over week under a 10% weekly rule. Those figures are an example of the mechanic, not a measured result. For the full treatment, including half-life models and re-engagement resets, see lead score decay explained.

5. Implement in your CRM

Configure automated scoring in HubSpot or Salesforce with workflows and formula fields so points are assigned without manual entry.

HubSpot implementation:

1. Create score properties: "Fit Score" (number 0-50), "Intent Score" (number 0-50), and "Total Lead Score" as a calculation of fit plus intent.
2. Build a fit-score workflow triggered on contact create or update, setting the fit score from company size, industry, and revenue via if/then branches.
3. Build an intent-score workflow triggered on page views, form submissions, and email opens that increments the intent score by the point value.
4. Build a decay workflow on a weekly batch trigger that reduces the intent score for leads with no activity in the last seven days.
5. Create a routing workflow: when the total score crosses your hot threshold, assign the rep, send a Slack notification, and create a "call now" task.

Salesforce implementation:

1. Create score fields on the Lead object: "Fit_Score__c", "Intent_Score__c", and a "Total_Score__c" formula field.
2. Build a record-triggered flow on lead update with decision branches for firmographics that set the fit score.
3. Use Pardot or Marketing Cloud to track web activity and push intent-score increments to Salesforce via API.
4. Build lead assignment rules that route leads above the hot threshold to the AE queue with an alert.

Both platforms work, and the copy-paste field lists live in the HubSpot template and Salesforce template. If you are on a newer, schemaless CRM, the same fit-and-intent structure ports over with formula fields rather than workflow plumbing.

6. Validate and refine

Track conversion by score tier and adjust point values and thresholds against real data, not assumptions. The AI GTM Engineer can measure the impact on pipeline velocity as well as raw conversion.

• Compare tiers to your own baseline: after a month of data, check whether hot leads actually convert better than warm, and warm better than cold. If the ordering does not hold, your weights are off.
• Watch the score distribution: if a large majority of leads land in the hot tier, the model is too generous and the threshold should rise.
• Tighten when hot leads convert poorly: raise the hot threshold or trim the points on the single action that is over-weighted.
• Re-weight non-predictive attributes: if a firmographic is not correlating with wins, lower its points; if a specific action keeps preceding closed deals, raise it.

Fit score vs intent score: what's the difference?

Many B2B scoring models fail because they blend fit and intent into one number. A large enterprise that visited your blog once ends up with the same score as a small startup that visited pricing three times. The dual-score model separates them so your team knows whether a lead should buy (fit) versus wants to buy now (intent).

The key insight: a lead with high fit and low intent is a future customer, so nurture them. A lead with low fit and high intent is a risky deal, so qualify harder before spending AE time. Only leads that clear the threshold on both dimensions should be routed to sales immediately.

What does a lead scoring quick reference look like?

A practical reference splits 100 points into two halves: a fit score (max 50) from firmographic signals like company size, industry, revenue, tech stack, and growth stage, and an intent score (max 50) from engagement signals like pricing-page visits, demo requests, and content downloads. Combine the two, then route once the total crosses your thresholds.

• Fit score (max 50): company size, industry, revenue, tech stack, and growth stage, roughly 10 points each in the illustrative framework.
• Intent score (max 50): demo request and repeat pricing visits weighted highest, content downloads and webinar attendance mid-weight, single opens and blog reads lowest, with a decay rule applied on inactivity.
• Total equals fit plus intent: hot leads (80+) get immediate sales contact, warm (60-79) enter nurture, cold (40-59) stay in marketing, and disqualified (under 40) are suppressed.

Beyond fit and intent: the pain-first evolution

Fit plus intent is the right starting point. It is also the wrong ending point. Two prospects with identical firmographics and identical engagement scores can have completely different operational pain. One is drowning in a broken pipeline and ready to buy anything that fixes it. The other is fine and will politely ignore every sequence you send. Standard scoring cannot tell them apart.

Pain-first ICP scoring is the next evolution. Instead of weighting fit and intent roughly equally, it weights pain signals heavily and firmographics lightly. A representative rubric:

• Pain relevance (largest weight): has the prospect publicly acknowledged a problem your product solves?
• Signal recency: how fresh is the pain signal? Recent signals get full weight and older ones fall off fast.
• Buyer signal strength: the title seniority and buying authority of the person showing the pain.
• Firmographic fit (capped low): capped so a perfect firmographic match with zero pain still scores low and gets rejected.

Two hard floors disqualify regardless of total score: some pain must be discussed, and the signal must be recent. A near-perfect prospect with a months-old signal is still disqualified, because stale pain is not actionable. Pain is the signal; firmographic is the filter. Read the full pain-first ICP scoring framework.

Methodology and limitations

This scoring model is based on our hands-on lead-scoring implementations for B2B companies, refined through real-world deployment at companies roughly in the $1M to $50M ARR range across a range of industries. Point allocations and threshold ranges are calibrated directionally against conversion data from those implementations, not a controlled study, and we recalibrate as new client results come in.

Every point value, tier, and conversion note in this guide is illustrative and meant to be tuned against your own data. Any figure framed as "commonly cited" is directional and drawn from widely cited lead-management research, not a proprietary dataset, and outcomes vary by ICP fit, offer, deal size, and sales process.

Frequently asked questions

Sources and references

The guidance here is directional, drawn from widely cited lead-management research. Point values and multipliers are illustrative, not guarantees.

• Lead Response Management study (Oldroyd, 2007): reported that working an inbound inquiry within 5 minutes rather than after half an hour is associated with a large lift, roughly 21x, in the odds of qualifying the lead. (The 2011 Harvard Business Review article "The Short Life of Online Sales Leads" reported a separate, smaller lift in qualification for contact within an hour versus later.)
• Gartner and Forrester lead-management guidance: research on lead scoring models, MQL definitions, and how top-performing teams prioritize pipeline.
• HubSpot and Salesforce lead-scoring documentation: practical guides to implementing fit and intent scoring in the CRM with rule-based and predictive approaches.