Account Scoring Framework for ABM in 6 Steps

Account Scoring Framework for ABM in 6 Steps

You can't chase every account. You have 50 salespeople. There are 50,000 potential accounts in your market. You need to score them.

Good scoring separates the $10M opportunity from the $50k dead-end. It keeps your team focused on accounts that can actually close.

Most teams either ignore account scoring (spray and pray) or build overly complex models (no one uses them). This framework splits the difference.

Why Lead Scoring Fails for ABM

Lead scoring measures buyer intent at the person level. Did this person visit your pricing page? Download a guide? Open an email? Score them 1-100.

Account scoring measures organization-level buying signals. Does this account have budget? Is the buying committee active? Are they evaluating competitive solutions?

Lead scoring tells you if a person is interested. Account scoring tells you if the organization will buy.

ABM lives at the account level. You need account scoring.

Step 1: Define What "Good" Looks Like

Before you build a scoring model, define your ideal customer profile (ICP).

Your ICP is a description of your best accounts. Not a person. An organization.

Criteria to consider: - Company size (employee count, ARR, funding stage) - Industry and vertical - Geography - Pain points (you solve X, Y, Z) - Technology stack (uses tools you integrate with) - Buying behaviors (buys software at your price point, sales cycle length)

Example ICP: "Series B or C SaaS companies, 50-500 employees, in fintech or insurtech, built on cloud infrastructure, with revenue ops teams, in North America, spending $100k-500k annually on go-to-market tools."

Now find 10-15 of your best customers. How many match this ICP? 80%? Great. You know what you're looking for.

Step 2: Identify Firmographic Attributes

Firmographic data is what you can look up about an account without any interaction.

Build a list of firmographic attributes:

Weight each attribute. Company size matters more than tech stack for most businesses.

Step 3: Identify Behavioral Signals

Behavioral signals are what an account does with your company.

Possible signals: - Website visits (especially pricing, product pages) - Content downloads - Webinar attendance - Email engagement (opens, clicks) - Trial signups - Demo requests - Sales calls scheduled - Active evaluation (competitor research, product searches)

Which signals matter most? Start with these five:

1. Demo request = High intent (10 points)
2. Trial signup = High intent (10 points)
3. 3+ pricing page visits in 30 days = Mid intent (5 points)
4. Webinar attendance + follow-up email open = Mid intent (5 points)
5. More than 5 website visits over 60 days = Low intent (3 points)

Document what each signal means. "3 pricing page visits" is specific and trackable. "Shows interest" is vague and won't help your sales team.

Step 4: Build Your Scoring Model

Combine firmographic and behavioral signals into one account score.

Simple formula:

Account Score = Firmographic Base + Behavioral Points

Firmographic Base: ICP match - Perfect ICP match: 40 points - Good match (80%): 25 points - Fair match (60%): 10 points - Poor match: 0 points

Behavioral Points: Activity in last 90 days - Demo request: 15 points - Trial signup: 15 points - 3+ pricing visits: 8 points - Webinar + open: 8 points - 5+ website visits: 5 points

Score ranges: - 50+: High priority (hand to sales immediately) - 30-49: Medium priority (nurture with content, schedule check-in in 30 days) - 10-29: Low priority (add to nurture list, revisit in 60 days) - 0-9: Not a fit (don't reach out)

Step 5: Set Up Tracking and Updates

Your scoring model is only useful if you actually track it and update it regularly.

Create a simple dashboard in your CRM or data warehouse:

Update behavioral scores weekly. Update firmographic scores quarterly (company size, funding changes happen slowly).

Use automation if you can. Most CRMs and marketing platforms can auto-update scores based on website visits and email engagement.

Manual monthly updates work too. Takes 4 hours a month. Worth it.

Step 6: Validate and Refine

After three months, look back at your highest-scoring accounts.

Questions to ask: - How many of your highest-scoring accounts moved to opportunities? - What was the average deal size? - How long was the sales cycle? - Did your sales team find these accounts valuable?

Example findings: "Our 50+ score accounts had 35% conversion to opportunity. Average deal size: $250k. Sales cycle: 4.5 months. Our salespeople loved these accounts. Keep the model."

Or: "Our 50+ score accounts had 8% conversion. Average deal size: $60k. Sales said half the accounts weren't even a fit. Model is broken. Let's re-weight."

Common refinements: - Add new behavioral signal (if you discover buyers watch demo videos before requesting a demo) - Change score thresholds (if your "high priority" threshold produces too many accounts) - Adjust firmographic base (if smaller accounts close just as well as your ICP) - Remove noisy signals (if a signal doesn't correlate with deals closed)

Common Mistakes

Mistake 1: Too many signals You build a model with 17 different behavioral signals. No one understands it. Sales doesn't trust it. You stop using it.

Fix: Start with five signals. Add more only if they improve accuracy.

Mistake 2: Weighting signals equally A demo request is not the same as a website visit. But you weight them the same. Noise wins.

Fix: Weight signals by predictive power. Ask yourself: "If an account does X, what's the probability they close?" Higher probability = higher weight.

Mistake 3: Not using ICP data You build a beautiful ICP. Then your scoring model ignores it. You waste time on bad-fit accounts.

Fix: Firmographic scoring should be at least 40% of your total score. Make ICP matter.

Mistake 4: Changing the model constantly You tweak the model every week. You never collect enough data to know if it works.

Fix: Run a model for at least 90 days before you refine it. Let data accumulate.

Tools to Implement This

You don't need expensive software. Options:

Option 1: Spreadsheet - Manual data updates - Formulas to calculate scores - Weekly sorting and prioritization - Works for 500-1000 accounts

Option 2: CRM native - Salesforce, HubSpot, Pipedrive all have scoring features - Auto-track website visits, email engagement - Built-in reporting - Scales to 10k+ accounts

Option 3: Intent data platform - 6sense, Demandbase, TechTarget track buyer intent at company level - AI-powered signal weighting - Expensive (starts at $30k/year) - Best for enterprise sales teams

Start with a spreadsheet or CRM native feature. Upgrade to intent data once you validate the model works.

Key Takeaways

Account scoring separates good opportunities from time-wasters. Combine firmographic (ICP match) and behavioral (activity) signals. Start with five signals. Weight by predictive power.

Update weekly. Review quarterly. Refine based on closed deals.

A simple, useful model beats a complex, unused model every time.

Build your first model this week. Load your target account list. Run scoring. Hand your top 50 accounts to sales.

Measure what happens next.

---