Intent data tells you that an account is in market. Website engagement tells you they are researching. Email engagement tells you they are responsive. But how do you combine these signals into a single number that tells your sales team: “This account is ready for outreach right now”?
An account intent scoring framework does exactly that. It synthesizes intent signals, engagement metrics, and behavioral indicators into a composite score. Accounts above a threshold automatically route to sales. Accounts below the threshold continue nurturing until scores rise.
When built correctly, an intent scoring model reduces time-to-outreach (from weeks to days), increases conversion rates (by engaging accounts at peak buying moment), and scales your sales capacity (by automating triage).
This guide walks through how to build, validate, and maintain an intent scoring model.
Why Intent Scoring Fails (And How to Fix It)
Most ABM teams approach intent scoring wrong: they weight all signals equally, or they over-index on a single signal (like Bombora scores) and ignore everything else.
This leads to:
- False positives: high-intent accounts that never buy (intent data can be noisy; not all category research translates to purchase)
- False negatives: accounts ready to buy that you miss (a company not showing Bombora intent but actively visiting your website multiple times is likely ready to talk)
- Alert fatigue: sales teams receive so many “high intent” notifications that they ignore all of them
The fix is to build a model that accounts for signal quality, signal correlation, and historical conversion data from your own deals. Your model should answer: “Based on the signals this account is showing, what is the probability it will enter the pipeline in the next 30-60 days?”
Core Intent Signals to Measure
1. Intent data subscriptions (Bombora, G2 Buyer Intent, TechTarget)
Intent data providers monitor research activity across web properties, sponsored content, and form fills. When your product category gets researched, providers flag it.
How to use intent data in your scoring: - Bombora surge scores: accounts showing elevated research activity on your category in the past 30 days get points - Competitor research: if an account is researching competitors specifically (not just your category), boost the score - Duration and recency: research activity that started last week is more intent-revealing than research from 3 months ago
Scoring implementation: - Bombora surge score 70+: +30 points - Bombora surge score 40-69: +15 points - Bombora surge score 1-39: +5 points - No Bombora activity: +0 points - (Adjust these thresholds based on your Bombora subscription level and the quality of your data)
2. Website engagement (visits, pages, time on site)
Website engagement is behavioral intent: accounts actively researching your solution.
How to use website engagement in your scoring: - Visit frequency: accounts visiting once per week are more likely to buy than accounts visiting once per month - Page depth: accounts visiting your pricing page, ROI calculator, or integrations page show higher purchase intent than accounts visiting only the homepage - Bounce rate: accounts bouncing after 5 seconds show low intent; accounts spending 2+ minutes show higher intent - Return visits: accounts that visit multiple days in a row show higher urgency than accounts with sporadic visits
Scoring implementation: - 3+ visits in past 7 days: +20 points - 1-2 visits in past 7 days: +10 points - Multiple visits in past 30 days but none in past 7 days: +5 points - Visit to pricing or ROI page: +10 bonus points - Visit to comparison or competitor page: +5 bonus points - No website activity in past 90 days: -5 points (downgrade recent score to account for stale activity)
3. Email engagement (open rate, click rate, reply)
Email engagement shows direct responsiveness. An account that is opening and clicking your emails is more likely to reply to a sales outreach.
How to use email engagement in your scoring: - Email opens: not all opens are equal; an account opening 50% of your emails shows more interest than an account opening 10% - Email clicks: clicking is stronger signal than opening; an account clicking links in email is actively evaluating - Reply rate: any email reply (even a “no thanks” or “not now”) is a positive signal
Scoring implementation: - Email open rate 50%+ and at least 1 click in past 30 days: +15 points - Email open rate 30-49% and at least 1 click: +10 points - Email open rate 30-49% with no clicks: +5 points - Email open rate below 30%: +0 points - Any email reply in past 30 days: +20 bonus points (this is your strongest direct intent signal) - Unsubscribe from email list: -10 points
4. Social and community signals
Social engagement on LinkedIn and community platforms is weaker than direct signals, but still useful.
How to use social signals in your scoring: - Account employees engaging with your LinkedIn content (likes, comments, shares) - Account employees joining your community or Slack group - Account following your company on LinkedIn
Scoring implementation: - 3+ employees from account engaging with your content in past 30 days: +10 points - 1-2 employees engaging: +5 points - Account company following you: +2 points
5. Relationship signals
Some of the strongest intent signals come from existing relationships.
How to use relationship signals in your scoring: - Recent sales conversation: an account that had a meeting with your team in the past 30 days is more likely to move toward purchase in the next 30 days - Demo request: a company requesting a demo has already decided to evaluate - Trial signup: if you offer a free trial, trial signup is a very high intent signal - Referral: accounts referred by customers or partners show higher intent
Scoring implementation: - Demo booked or completed in past 30 days: +25 points - Trial started in past 30 days: +25 points - Sales conversation (discovery, intro, any meeting) in past 30 days: +20 points - Referral from existing customer: +15 points
Building Your Scoring Model: The 5-Step Process
Step 1: Choose your scoring scale
Use a 0-100 scale. This makes thresholds easy to understand: 70+ = high intent, 50-69 = moderate intent, below 50 = low intent.
Start with a relatively simple model (10-15 signals) rather than a complex one (50+ signals). Simple models are easier to validate, maintain, and adjust.
Step 2: Assign point values based on signal strength
Not all signals are equally predictive. Decide which signals are strongest and allocate points accordingly.
Use this ranking to guide your decisions: 1. Strongest signals (15-25 points each): Demo request, trial signup, recent sales conversation, email reply 2. Strong signals (10-15 points each): High email engagement + clicks, multiple website visits in past 7 days, Bombora surge score 70+ 3. Moderate signals (5-10 points each): Single website visits, moderate email engagement, social signals 4. Weak signals (1-2 points each): Account follows you, low email open rate
Step 3: Set trigger thresholds
Define what score requires action: - 70+: “High intent” trigger. Route to sales team immediately. This account is likely in evaluation. - 50-69: “Moderate intent” trigger. Continue nurturing. Route to SDR for light outreach (1 attempt per week). - Below 50: “Low intent” nurture. Automated nurture only until score rises.
You may also create sub-thresholds for different team actions: - 80+: Trigger sales outreach from account executive - 70-79: Trigger SDR outreach - 60-69: Trigger nurture email campaign - Below 60: Automated nurture only
Step 4: Validate against historical deals
Before you use your scoring model in production, validate it against your historical closed deals and lost opportunities.
Pull a sample of deals closed in the past 6-12 months. For each closed deal, assign it a retroactive intent score based on the signals that account showed 30 days before the deal closed.
Validation questions: - What was the average intent score of closed deals? (You want this to be 65+.) - What percentage of accounts that scored 70+ in the past turned into closed deals within 60 days? (You want this to be 20%+ for high-intent accounts.) - Are there closed deals that scored below 50? If so, those are accounts you missed; adjust your model to better capture those signal patterns. - Are there accounts that scored 80+ but never closed? If so, you may be over-weighting certain signals or missing negative signals like unsubscribes.
Adjust your scoring model based on validation results, then re-validate.
Step 5: Document your scoring logic
Create a simple scoring card that documents your model. Example:
ACCOUNT INTENT SCORING MODEL
Version: 1.0
Effective: April 2026
INTENT DATA (Bombora)
- Surge score 70+: +30 points
- Surge score 40-69: +15 points
- Surge score 1-39: +5 points
WEBSITE ENGAGEMENT
- 3+ visits in past 7 days: +20 points
- 1-2 visits in past 7 days: +10 points
- Visit to pricing/ROI page: +10 bonus points
EMAIL ENGAGEMENT
- Open rate 50%+, 1+ click in past 30 days: +15 points
- Open rate 30-49%, 1+ click: +10 points
- Email reply in past 30 days: +20 bonus points
RELATIONSHIP SIGNALS
- Demo booked in past 30 days: +25 points
- Sales conversation in past 30 days: +20 points
- Trial started in past 30 days: +25 points
SOCIAL SIGNALS
- 3+ employee engagements in past 30 days: +10 points
- 1-2 employee engagements: +5 points
DECAY/PENALTIES
- No activity for 90+ days: -5 points
- Unsubscribe from email: -10 points
THRESHOLDS
70+: High intent, route to sales
50-69: Moderate intent, light SDR outreach
<50: Low intent, nurture only
REFRESH FREQUENCY: Weekly
Automating Your Scoring Model
Once your model is documented, automate it using your existing tools:
Option 1: Native ABM platform scoring
Most modern ABM platforms (Abmatic AI, 6sense, Demandbase) have built-in intent scoring. You can configure accounts to be scored automatically based on: - Bombora or other intent data integration - Website engagement tracking (pixel-based) - CRM data (recent meetings, pipeline stage) - Email engagement (if connected to your marketing automation platform)
If using an ABM platform, configure it to auto-score accounts weekly and trigger workflow actions when accounts hit your thresholds.
Option 2: Marketing automation intent scoring
Platforms like HubSpot and Marketo have lead scoring (for individual contacts) but also account-level scoring capabilities.
In HubSpot: - Create an account property “Intent Score” - Use workflows to update Intent Score based on email engagement, form submissions, and lifecycle stage - Connect your Bombora intent data via an integration (native in HubSpot) - Use the account intent score to trigger routing and engagement workflows
Option 3: Custom scoring via Zapier or Make
If your ABM platform and marketing automation platform do not talk to each other, you can use a connector like Zapier or Make to aggregate signals:
- Pull weekly intent data from Bombora via API
- Pull website engagement data from your analytics tool (GA4, Segment, Mixpanel)
- Pull email engagement data from your marketing automation platform
- Calculate intent score as a weighted sum of signals
- Write the score back to your CRM
- Trigger workflows in your CRM based on score thresholds
Skip the manual work
Abmatic AI runs targets, sequences, ads, meetings, and attribution autonomously. One platform replaces 9 tools.
See the demo →Monitoring and Maintaining Your Model
Intent scoring is not a “set it and forget it” system. Monitor and adjust monthly:
-
Track alert accuracy. When an account hits 70+ intent score and you send a sales outreach, do they respond? If response rate is below 30%, your model is generating false positives. Adjust signal weights downward for the signals that account was showing.
-
Track missed deals. Did any accounts convert to customers without ever hitting 70+ intent? If so, you are missing signals. Review their signal pattern and adjust your model to better capture those patterns.
-
Check for signal decay. After an account hits high intent and you send sales outreach, the account score should decay over time if engagement stops. If accounts stay at 70+ scores indefinitely without action, you are missing a decay mechanism. Add a rule that reduces scores for accounts unresponsive to sales outreach.
-
Validate new data sources. As you add new intent data providers or engagement channels, test whether those signals predict deals before adding them permanently to your model.
Example: Full Scoring Walkthrough
Account: Acme Corp ($200M revenue, in your target vertical)
Current signals: - Bombora surge score: 65 (research activity on your category in past 30 days) - Website visits: 4 visits in past 7 days, including 1 visit to pricing page - Email engagement: 60% open rate, 2 clicks on recent email about ROI - Social: 2 employees from Acme have engaged with your LinkedIn content in past 2 weeks - Relationship: Account had a discovery call with your team 10 days ago - No trial, no demo yet
Scoring calculation: - Bombora 65: +15 points - 4 visits in 7 days: +20 points - Visit to pricing: +10 points - Email open rate 60% + 2 clicks: +15 points - No email reply yet: +0 points - 2 employee social engagements: +5 points - Recent sales conversation: +20 points - No demo request or trial: +0 points - Total: 85 points
Action: Score 85 triggers “High intent” workflow. Sales team assigns dedicated account executive. Marketing provides case studies and ROI resources relevant to Acme’s vertical. SDR schedules follow-up demo for next week.
Common Intent Scoring Mistakes
Mistake 1: Over-weighting Bombora, under-weighting website engagement
Bombora is useful but noisy. Many accounts research your category without intent to buy. Website engagement (especially visits to your pricing or ROI pages) is often a stronger signal. Do not assume Bombora is your source of truth.
Mistake 2: Ignoring email unsubscribes
An account that unsubscribes from your email list is actively choosing not to engage. Unsubscribes are intentional negative signals. Penalize them significantly (e.g., -10 points).
Mistake 3: Not decaying old signals
An account that visited your website 6 months ago and then went silent should not stay at high intent forever. Implement decay: reduce scores if no new engagement in past 30 days.
Mistake 4: Treating all demos as equal
A demo is high intent, but a demo from an individual contributor at a $5M company is different from a demo request from a CIO at a $200M company. Add context to your demo signals: weight demos from higher-title roles or larger companies more heavily.
Next Steps
- Document your current signals. List every piece of intent data you currently have access to (Bombora, website analytics, email platform, CRM).
- Assign point values. Use the framework above to create your initial scoring model.
- Validate against deals. Score your last 20 closed deals retroactively. Does your model identify them as high intent 30-60 days before close?
- Automate in your stack. Configure your ABM platform or marketing automation to auto-score accounts weekly.
- Monitor for accuracy. Track whether accounts scoring 70+ actually respond to outreach.
An intent scoring model transforms abstract “urgency” into a number your sales team can act on. When built right, it compresses sales cycles by helping you engage accounts exactly when they are ready to buy.
