Account Scoring Playbook: Prioritize Your Best Prospects

Throwing equal effort at every prospect is like fishing with a net when you should be fly-fishing. Account scoring fixes this. It tells you which accounts deserve your best salespeople, which need nurture, and which should move to a lower tier.

This playbook walks you through building a scoring model that actually works.

What Account Scoring Is (and What It's Not)

Account scoring combines two signals:

Fit Score: How well does this account match your ideal customer profile? Size, industry, technology stack, geographic region: the structural things that make a company a good customer.

Engagement Score: How interested is this account right now? Website visits, email opens, demo requests, content downloads: behavioral signals that indicate buying momentum.

Together: Fit + Engagement = Priority. A great fit with low engagement is a future opportunity. High engagement with poor fit is a distraction.

Building Your Fit Score

Fit score is structural and static. It answers: "If everything else were equal, how good a customer would this be?"

Step 1: Define Your ICP Dimensions

Answer these questions for your best customers:

• Company size: Revenue range, employee count, or both?
• Industry: Which verticals buy? Which don't?
• Technology: What platforms do they use? (Salesforce shops vs. Pipedrive shops often buy differently)
• Geography: Countries or regions you focus on?
• Buying behavior: Do they prefer contract sales? Self-serve? Mid-market vs. enterprise sales cycles?

Write these down as ranges or categories. Don't be vague.

Step 2: Score Each Dimension

Create a point system. Example:

Company Size (Revenue) - $100M+ revenue: 25 points - $50M–$100M: 20 points - $10M–$50M: 15 points - $1M–$10M: 5 points - <$1M: 0 points

Industry - Financial services: 25 points - SaaS: 20 points - Technology: 20 points - Retail: 10 points - All others: 0 points

Geography - USA: 25 points - Canada/UK: 20 points - EMEA: 15 points - Rest of world: 0 points

Technology Stack - Uses Salesforce: +10 points - Uses HubSpot: +5 points - Uses neither: 0 points

Total possible fit score: 100 points.

Step 3: Add Firmographic Rules

Some attributes should be disqualifying. Example:

• If company is in bankruptcy or has been acquired in the last 6 months: 0 fit points (move to "not now")
• If company is a direct competitor: 0 fit points
• If company is in a market you've explicitly deprioritized: 0 fit points

Use these as gates before the point system.

Step 4: Automate the Calculation

Feed company data into your CRM or ABM platform. Most platforms can auto-calculate fit scores based on rules you define. You plug in company enrichment data (from Clearbit, Hunter, or Apollo), and the system scores automatically.

Building Your Engagement Score

Engagement score is behavioral and dynamic. It measures buying signals in real time.

Step 5: Choose Your Engagement Signals

Pick 5-8 signals that correlate with closing deals in your business. Example set:

• Website visits: 1 point per visit (last 30 days)
• Page views on pricing page: 5 points per visit
• Asset downloads: 3 points per download
• Email opens: 0.5 points per open (last 30 days)
• Demo request: 25 points
• Sales call booked: 25 points
• Competitor research: Search for competitor names on your site, 5 points per session
• Time on site: >2 minutes in a session, 1 point

Step 6: Set Decay and Freshness Rules

Engagement scores should degrade over time. A website visit 90 days ago matters less than one last week.

Example rules: - Activity older than 90 days: divide score by 2 - Activity older than 180 days: score expires (0 points) - More than 5 visits in the last 7 days: multiply score by 1.5 (indicates sustained interest)

Step 7: Create Engagement Tiers

Set ranges that map to action:

• Engagement score 50+: Hot lead, hand to sales immediately
• Engagement score 30–49: Warm lead, add to nurture email sequence
• Engagement score 10–29: Cold lead, add to account-based ad retargeting
• Engagement score <10: No recent activity, keep in nurture but deprioritize

Step 8: Track Engagement Across Buying Committee

One person engaging isn't the same as two people or four people. Create a multiplier:

• If 1 person from the account is active: engagement score = base score
• If 2+ people from the account are active: engagement score = base score × 1.3
• If 4+ people from different departments are active: engagement score = base score × 1.6

This indicates broader internal support and higher deal probability.

The Scoring Matrix

Combine fit and engagement into a 2x2 matrix:

                High Engagement    Low Engagement
High Fit        SELL NOW          NURTURE (future)
Low Fit         QUALIFY           DEPRIORITIZE

Sell Now (High Fit + High Engagement)

• Assign to best AEs
• Fast-track to demo or trial
• Allocate 60% of sales time here

Nurture (High Fit + Low Engagement)

• Add to email nurture sequences
• Weekly account-based ads
• Target with thought leadership content
• Check quarterly for engagement lift

Qualify (Low Fit + High Engagement)

• Sales has quick call to understand fit
• If fit improves after conversation: move to Sell Now
• If fit is poor but they're interested: move to lower-tier nurture
• If fit is poor and interest drops: archive

Deprioritize (Low Fit + Low Engagement)

• Passive nurture only (email list, occasional content)
• Review monthly to see if fit or engagement improves
• Reallocate sales resources

Continuous Improvement

Month 1: Establish Baseline

Run your scoring model for 30 days without changing behavior. Track which accounts are "hot" according to the model. Don't change scoring yet; just observe.

Month 2: Validate Against Real Data

Look at the accounts that closed in the last 90 days. What were their fit and engagement scores at the time of first touch? At the time of close?

You'll find patterns: "All closed deals had fit scores above 70" or "All closed deals showed spike in engagement 2-3 weeks before demo."

Month 3: Tune Your Model

Adjust point values based on what you learned. If demo requests should be worth 50 points instead of 25, change it. If accounts in a certain geographic region close faster, weight that dimension up.

Ongoing: Review Monthly

Every month, pull a report: - Top 20 accounts by combined score: Are they in active conversations? - Accounts that dropped from Sell Now to Nurture: Did engagement actually drop, or did we lose visibility? - Accounts that moved from Nurture to Sell Now: How long did nurture take to work?

Share this with sales. Let them challenge the model. "This account scored low but we're in active negotiations" is valuable feedback.

Common Mistakes

Mistake 1: Over-weighting fit. If an account is low-fit but actively buying, don't ignore them. Some of your best customers might not fit the original ICP.

Mistake 2: Engagement score noise. Counting every email open creates too much volatility. Stick to meaningful signals: form submissions, demo requests, extended website sessions.

Mistake 3: No decay. If a prospect engaged 6 months ago and you're still treating them as hot, you're wasting time. Build in freshness rules.

Mistake 4: Static scoring. If you don't revisit your scoring model quarterly, it drifts. Markets change, your product changes, your customer base changes.

Mistake 5: Scoring without sales input. If sales doesn't trust the model, they'll ignore it. Show them the data. Let them see that high-fit + high-engagement accounts actually close.

FAQ

Q: What if we don't have engagement data? A: Start with fit score only. As you build integrations with your website and email platform, layer in engagement signals. Even partial data is better than guessing.

Q: How often should we recalculate engagement scores? A: Weekly is ideal. Real-time is better if your platform supports it. This way, a new website visit or email open updates the score immediately, and sales has fresh data.

Q: Should fit and engagement be weighted equally? A: Not necessarily. Test both: fit-heavy models, engagement-heavy models, and 50/50 models. See which correlates best with your actual closed deals.

Q: What if we have a small customer base and not enough historical data? A: Start with an external benchmark or expert input. Interview your five best customers: Why did they buy? What was the engagement pattern? What company characteristics did they share? Use that to build your first model.

Q: Can we use account scoring for lead scoring too? A: Yes, but account-level scoring is different from contact-level scoring. You need both. A contact might be high-engaged but low-influence. An account might be high-fit but all your contacts are low-engaged.

Next Steps

Build your fit score first. It's the foundation. List your five best customers and five worst customer relationships. What structural differences do you see? Codify those differences into your ICP dimensions and point system.

Then layer engagement signals. Start with the signals you already have: website data, email engagement, and form submissions. Add more as your tooling evolves.

Ready to automate your account scoring? Abmatic AI helps you score accounts by fit and engagement, then prioritize them for your sales team.