Account Scoring Methodology: Building a Rigorous Ranking System
Account scoring is the backbone of ABM prioritization, combining firmographic fit and behavioral intent to predict pipeline conversion. Without it, you're choosing accounts based on intuition or company size alone, wasting resources on low-fit targets.
Why Account Scoring Matters in ABM
In traditional lead scoring, you optimize for lead-to-opportunity conversion rate. In account scoring, you optimize for account-to-pipeline conversion and deal size.
Account scoring helps you:
- Prioritize accounts most likely to convert
- Allocate limited outreach resources efficiently
- Identify which accounts to expand within your existing customer base
- Measure how market dynamics (intent, hiring, funding) change account priorities
- Enable sales to focus on hot accounts today vs. warm accounts for tomorrow
Step 1: Define Your Ideal Customer Profile (ICP)
Before scoring, define who you want to win. Your ICP is the starting point.
ICP dimensions to include:
Firmographic Attributes
- Annual revenue (range)
- Company size (headcount range)
- Industry/vertical
- Geographic region
- Growth stage (early-stage, growth, mature, enterprise)
- Use case suitability (e.g., are they using legacy systems you displace?)
Technographic Attributes
- Current tech stack (competitors, complementary tools)
- Cloud adoption level
- CRM platform in use
- Other relevant tools in your ecosystem
Outcome-Based Attributes
- Customer type (e.g., must have dedicated marketing ops team)
- Buying motion (e.g., self-serve, enterprise sales)
- Deal size your product can support
- Expansion potential
Action: Write down your top 5-10 ICP characteristics. Rank them by importance.
Step 2: Create Your Firmographic Scoring Component
Firmographic scoring measures how well an account matches your ICP.
Set up a points-based system:
Revenue
- Target revenue range: 100 points
- Revenue within 20% of target: 75 points
- Revenue within 50% of target: 50 points
- Below/above target but still viable: 25 points
- Wrong revenue band: 0 points
Company Size
- Target headcount range: 100 points
- Within 25% of target: 75 points
- Within 50% of target: 50 points
- Viable but not ideal: 25 points
- Wrong market (too small, too large): 0 points
Industry
- Tier 1 target industries: 100 points
- Tier 2 target industries: 75 points
- Adjacent industries with fit: 50 points
- Other: 0 points
Geography
- Priority regions: 100 points
- Secondary regions: 75 points
- Lower-priority regions: 25 points
- Unlikely regions: 0 points
Total firmographic score: Out of 400 points (normalize to 0-100)
Example: Account with: - Revenue in target range (100 pts) - Headcount 40% above target (50 pts) - Tier 1 industry (100 pts) - Priority geography (100 pts) - Total: 350/400 = 87.5 firmographic score
Step 3: Add Behavioral Scoring
Behavioral scoring captures account engagement with your content and brand.
Behavioral signals to track:
| Signal | Points | Rationale |
|---|---|---|
| Website visit (last 30 days) | 10 | Awareness phase |
| Multiple page visits (10+) | 15 | Active interest |
| Content download | 20 | Active research |
| Webinar attendance | 25 | Intent signal |
| Demo request | 50 | Sales conversation |
| Sales email open | 5 | Engagement |
| Sales email click | 10 | Higher engagement |
| Website return visit | 8 | Repeated interest |
| Competitor comparison content | 20 | Evaluation phase |
Decay by recency:
- Activity in last 7 days: 100% of points
- Activity in last 14 days: 80% of points
- Activity in last 30 days: 60% of points
- Activity 30+ days ago: 20% of points
Rolling 90-day calculation: Every account gets a behavioral score calculated from recent activity. Update weekly.
Example account: - Website visit (last 7 days): 10 pts - Content download (last 14 days): 16 pts (20 × 80%) - Webinar attendance (last 5 days): 25 pts - Total behavioral score: 51 points
Step 4: Integrate Intent Signals
Intent data predicts accounts actively buying. This is your strongest conversion signal.
Intent signal categories:
First-Party Intent
- Content consumption (your website)
- Email engagement (your emails)
- Events (attended your webinar)
- Sales conversations initiated
Third-Party Intent
- External intent data vendor signals (job postings, earnings calls, funding announcements)
- Technographic changes (new tools adopted)
- Industry reports consumption
- Competitor website visits (requires intent data partnership)
Zero-Party Intent
- RFP responses
- Pricing page engagement
- Product trial signup
Scoring intent signals:
| Signal | Points | Signal Type |
|---|---|---|
| High intent vendor signal | 40 | Third-party |
| Rising buying signal velocity | 30 | Third-party |
| Competitor mention in content | 20 | Third-party |
| Product trial signup | 50 | Zero-party |
| RFP response | 60 | Zero-party |
| Pricing page engagement | 35 | First-party |
Velocity matters: 3 intent signals in one week = 50 points. 1 signal per week = 15 points.
Example account: - Vendor data shows high intent (40 pts) - Rising velocity (3 signals this week) (30 pts) - Downloaded pricing guide (35 pts) - Total intent score: 105 points
Step 5: Build Your Composite Score
Combine all three components into a final score.
Weighting framework:
| Component | Weight | Points |
|---|---|---|
| Firmographic | 40% | 0-100 |
| Behavioral | 30% | 0-100 |
| Intent | 30% | 0-100 |
Final score formula:
Final Score = (Firmographic × 0.4) + (Behavioral × 0.3) + (Intent × 0.3)
Example account: - Firmographic: 87 - Behavioral: 51 - Intent: 80 - Final Score = (87 × 0.4) + (51 × 0.3) + (80 × 0.3) = 34.8 + 15.3 + 24 = 74.1
Skip the manual work
Abmatic AI runs targets, sequences, ads, meetings, and attribution autonomously. One platform replaces 9 tools.
See the demo →Step 6: Define Account Tiers
Create tiers based on final scores. Different tiers get different strategies.
Recommended tier structure:
| Tier | Score | Strategy |
|---|---|---|
| Tier 1 (VIP) | 85+ | Personal outreach, dedicated rep, executive engagement |
| Tier 2 (Growth) | 70-84 | Account-based campaigns, personalized content, group events |
| Tier 3 (Market) | 50-69 | Targeted ads, general nurture campaigns, webinars |
| Tier 4 (Pool) | <50 | Exclude or minimal activity, no personalization |
Step 7: Validate Your Model Against Historical Data
Before deploying, test your methodology against accounts you've already won or lost.
Validation process:
- Score your last 50 closed-won accounts
- Score your last 30 closed-lost accounts
- Score a random sample of 50 non-sales accounts
Measure:
- Average score for won accounts
- Average score for lost accounts
- What score threshold separates won from lost?
Example results:
- Won accounts average score: 78
- Lost accounts average score: 52
- Non-sales accounts average: 45
- Target sweet spot: 70+
If won accounts have low scores, your model needs adjustment. Re-weight components or add missing signals.
Step 8: Set Up Continuous Scoring
Automate your scoring so it updates regularly.
Frequency recommendations:
- Behavioral scores: Update weekly (new web visits, email engagement)
- Intent scores: Update weekly or daily (intent data feeds)
- Firmographic scores: Update quarterly or when company data changes
- Final score: Recalculate weekly
Use tools that:
- Automatically track web visits and engagement by account
- Integrate with your CRM
- Ingest third-party intent data feeds
- Push updated scores to your CRM for sales visibility
- Tier accounts automatically based on thresholds
Step 9: Create Account Movement Rules
As accounts move through their journey, their scores will change. Define how that triggers action.
Movement rules:
- New Tier 1 account: Alert sales immediately; add to top rep's queue
- Account dropped from Tier 1 to Tier 2: Sales should continue nurturing; decrease touch frequency
- Account dropped below 50: Move to general nurture; pause ABM campaigns
- Account jumped from below 50 to Tier 2+: Trigger re-engagement sequence; add to campaign
Step 10: Measure and Refine
Your model improves over time. Track what actually predicts wins.
Monthly refinement meeting:
- Which scored-high accounts closed? Which didn't?
- Were there accounts you missed (low scores, but won anyway)?
- Did any behavioral or intent signals NOT predict conversion?
- Should we adjust weights or add new signals?
Quarterly model refresh:
- Re-run validation against latest closed deals
- Are Tier 1 accounts converting at expected rates?
- Should we raise or lower score thresholds?
Common Pitfalls
Using vanity metrics: Company size alone doesn't predict fit. Prioritize outcome-based metrics.
Ignoring account velocity: A score of 78 this month means nothing if it was 45 last month. Velocity signals intent.
Over-complicating the model: Start with 3-4 key signals per component. Add complexity later.
Forgetting to validate: Never deploy a scoring model without testing it against historical data.
Static scoring: Update weekly. Markets change, buying signals change, accounts evolve.
Scoring Model Checklist
- [ ] Defined ICP with 5-10 key firmographic/technographic/outcome characteristics
- [ ] Built firmographic component (0-100 scale)
- [ ] Built behavioral component with web/email/event signals and recency decay
- [ ] Integrated intent signals (first-party, third-party, zero-party)
- [ ] Weighted components (firmographic 40%, behavioral 30%, intent 30%)
- [ ] Defined account tiers (Tier 1, 2, 3, 4 with clear thresholds)
- [ ] Validated model against last 50 won and 30 lost accounts
- [ ] Set up automated weekly scoring in your tech stack
- [ ] Created account movement rules (when to alert sales)
- [ ] Scheduled monthly review to refine based on results
A solid account scoring model takes 4-6 weeks to build and validate. Once it's live, you'll see immediate benefits: higher rep productivity, faster sales cycles, and more efficient marketing spend. Keep iterating based on what you learn.
