How to Build an Account Scoring Model for B2B SaaS in 2026

How to Build an Account Scoring Model for B2B SaaS in 2026

You have 500 target accounts. Your team can realistically work 50.

Which 50 should you prioritize?

This is the question account scoring answers. But most teams get it wrong. They build scoring models that are too complex (12+ criteria, hard to manage), too simplistic (only company size and industry), or wrong (measuring activity instead of fit).

A good scoring model is simple. It predicts which accounts have the highest probability of buying. And it tells you something your team doesn’t already know.

This guide walks you through building a scoring model in three days.

Why Account Scoring Matters

Account scoring filters signal from noise. Without it, your team works based on: - Gut feel (sales leader says “those accounts are good”) - Activity (accounts that engage with your content, even if they’re not buying) - Inbound volume (whoever calls first gets attention)

None of these predict which accounts will actually close.

A scoring model predicts. It says: “These 50 accounts are statistically more likely to buy than the other 450.”

With prediction, you can: - Prioritize limited resources (sales time, marketing spend, content) - Identify accounts likely to expand (existing customers with high scores) - Spot warning signs (good fit accounts that aren’t engaging) - Move faster (focus energy on accounts that matter)

Components of a Good Account Scoring Model

A good model has three components:

1. Firmographic fit (company characteristics) - Is the company the right size? - Is the company in our target industry? - Is the company in our target geography? - Does the company have the budget we assume?

2. Behavioral signals (what the company is doing) - Are they visiting our website? - Are they reading our content? - Are they engaging on social? - Have they taken actions that suggest buying intent?

3. Intent signals (external data about buying activity) - Are they searching for solutions in your category? - Did their tech stack change recently (adding relevant tools)? - Are they hiring in relevant roles (e.g., hiring a VP Sales if you sell a sales tool)?

Most weak models overweight behavior (activity = interest). Good models balance all three.

Step 1: Validate Your ICP (Ideal Customer Profile)

Before you score anything, validate that you know who your ideal customer is.

ICP Definition Exercise (1 hour)

Gather your sales, marketing, and CS leaders in a room. Answer these questions:

Company size: - What’s the minimum employee count? (e.g., 50 employees) - What’s the maximum? (e.g., 500 employees) - Why? (budget size, organization complexity, user count)

Industry/Vertical: - Which 3-5 industries see your fastest sales cycles? - Which have highest deal sizes? - Which have highest churn (avoid)? - Which have best expansion/upsell?

Geography: - Which countries/regions are you prioritizing? - Why? (go-to-market strategy, support capacity, regulations)

Job titles of buyers: - Who is the primary buyer? (e.g., VP Sales, VP Marketing) - Who are influencers? (e.g., Sales Operations, RevOps) - Who are blockers? (e.g., IT/Security, Finance)

Use case/pain point: - What problem are they trying to solve? - What’s the urgency? (new budget, existing tool dissatisfaction, new team)

Financial indicators: - What’s your assumed deal size? ($50K, $500K?) - Can your ICP afford it? - Do they have budget? (Series A vs. pre-seed affects this)

Technology stack: - What tools do they already use that integrate with yours? - What tools are they using that suggest they have your pain?

Write this down as “ICP 2.0.” This is your foundation for scoring.

Step 2: Build the Firmographic Score (40% weight)

Firmographic score is objective. Either a company matches your ICP or it doesn’t.

Firmographic Scoring Card

Create a spreadsheet or tool that captures these criteria:

Company Name | Size (Y/N) | Industry (Y/N) | Geography (Y/N) | Budget Signal (Y/N) | Firmographic Score
Google       | Y          | N              | Y               | Y                   | 75% (3 of 4)
Acme Corp    | Y          | Y              | Y               | Y                   | 100% (4 of 4)
Startup XYZ  | N          | Y              | Y               | N                   | 50% (2 of 4)

Scoring logic: - Each firmographic criterion = 25 points - Total firmographic score = 0-100

Include these criteria:

1. Company Size (25 points) - 50-500 employees: Full 25 points - 20-50 employees: 15 points (smaller, may have lower budget) - 500-1000 employees: 20 points (larger, slower to move) - >1000 employees: 0 points (not in ICP) - <20 employees: 0 points (not in ICP)

2. Industry Vertical (25 points) - Tier 1 vertical (SaaS, Financial Services): 25 points - Tier 2 vertical (Healthcare, Manufacturing): 15 points - Tier 3 vertical (Other): 5 points - Not relevant: 0 points

3. Geographic Location (25 points) - US: 25 points - UK, Canada: 20 points - EU, APAC: 10 points - Rest of World: 0 points

4. Revenue/Funding Signal (25 points) - Raised Series B+: 25 points - Raised Series A: 20 points - Pre-series funding: 10 points - Profitable with $10M+ revenue: 25 points - Bootstrapped <$5M revenue: 5 points - Unknown: 10 points (benefit of the doubt)

Tool for firmographic data: - ZoomInfo, Apollo, Clearbit: Subscription data services - LinkedIn: Manual research or API - Company website and Crunchbase: Free

Step 3: Build the Behavioral Score (30% weight)

Behavioral score measures engagement with your company and content.

Behavioral Scoring Card

Track these activities over the last 90 days:

Company Name | Web Visits | Demo Attended | Content Downloads | LinkedIn Engaged | Social Score
Google       | 15         | 0              | 2                 | 3                | 40
Acme Corp    | 45         | 1              | 5                 | 8                | 75
Startup XYZ  | 3          | 0              | 0                 | 0                | 10

Scoring logic:

1. Website visits (10 points total) - <5 visits in 90 days: 0 points - 5-10 visits: 3 points - 10-20 visits: 6 points - 20+ visits: 10 points

2. Content engagement (10 points total) - Downloaded whitepaper or guide: 4 points (per download, max 4) - Attended webinar: 5 points - Read 3+ blog posts: 2 points - Other content interactions: 1 point each (max 10 total)

3. Demo or product trial (10 points total) - Attended live demo: 7 points - Signed up for trial: 8 points - Scheduled follow-up: 10 points

Total behavioral score = 0-100

How to Capture Behavioral Data

• Website behavior: Google Analytics 4, account-based analytics platforms (Terminus, Demandbase)
• Email engagement: HubSpot, Marketo, or your email platform (opens, clicks)
• Demo and trial: Calendly and your product (form submissions, trial signups)
• Social engagement: Manual LinkedIn review (quarterly) or CRM notes

Important: Behavioral data gets stale. Update at least monthly.

Step 4: Build the Intent Score (30% weight)

Intent signals tell you who’s actively buying right now.

Intent Signals to Track

1. Third-party intent data (if budget allows) - Bombora, ZoomInfo, 6sense: Which accounts are searching for your solution? - Cost: $5K-50K/month depending on scale - Score: If on intent list = 50 points, if top 10 intent companies = 100 points

2. Tech stack changes (free to low-cost) - Is the company adding relevant tools? (Slack, Salesforce, HubSpot) - Use Hunter.io, Apollo, Crunchbase to see recent hires in relevant roles - Score: Tech stack addition or relevant hire = 20 points

3. Hiring signals (free) - Are they hiring a VP Sales? (suggests sales team expansion, need for tools) - Are they hiring a Sales Ops person? (suggests process maturation) - Use LinkedIn job changes, LinkedIn ads (“we’re hiring”), or company career page - Score: One relevant hire = 15 points, multiple hires = 30 points

4. Business events and announcements (free) - Series A funding (suggests new tools investment) = 30 points - New product launch (suggests company growth) = 15 points - New market entry (suggests expansion needs) = 20 points - IPO, acquisition, rebranding = 20 points - Use Crunchbase, PitchBook, or news alerts

5. Account owner/contact changes (free) - New CEO or VP Sales (often trigger tool refresh) = 20 points - Use LinkedIn alerts or quarterly Zoominfo updates

Intent Scoring

Option A: Third-party intent data (if you have it) - Account on intent list: 50 points - Account in top 50 intent ranks: 75 points - Account in top 10 intent ranks: 100 points - No intent data available: 25 points (benefit of doubt)

Option B: Self-built intent signals (if no third-party data) - Track each signal (hiring, funding, tech stack change, announcement) = 15-30 points each - Max out at 100 points total

Total intent score = 0-100

Step 5: Combine Scores Into Account Score

Weighted Formula

Account Score = (Firmographic × 0.4) + (Behavioral × 0.3) + (Intent × 0.3)

Example:

Acme Corp: - Firmographic: 100 (perfect ICP fit) - Behavioral: 75 (strong engagement, attended demo) - Intent: 50 (third-party intent data, but not top 10) - Total score: (100 × 0.4) + (75 × 0.3) + (50 × 0.3) = 40 + 22.5 + 15 = 77.5

Startup XYZ: - Firmographic: 50 (partial fit, smaller company) - Behavioral: 10 (minimal engagement) - Intent: 80 (just raised Series A, strong hiring) - Total score: (50 × 0.4) + (10 × 0.3) + (80 × 0.3) = 20 + 3 + 24 = 47

Recommended Thresholds

• 80+: Tier 1 (Top Priority)
• Action: Direct sales outreach, custom content, executive touch
• Effort: 5-10 hours/month per account

• Time to qualification: 4-8 weeks

• 60-79: Tier 2 (High Priority)

• Action: Templated outreach, nurture, group webinars
• Effort: 2-5 hours/month per account

• Time to qualification: 8-12 weeks

• 40-59: Tier 3 (Standard Priority)

• Action: Email sequences, self-serve content, community engagement
• Effort: 0.5-2 hours/month per account

• Time to qualification: 12+ weeks

• <40: On-Deck

• Action: Monitor, re-score quarterly
• Effort: Minimal
• Time to qualification: When they move to higher tier

Step 6: Build Your Scoring Dashboard

Create a live view that updates weekly or monthly.

Minimum Viable Dashboard

A spreadsheet with these columns:

• Company name
• Industry, size, geography
• Firmographic score (0-100)
• Last website visit date
• Content downloads (count)
• Demo attended? (Y/N, date)
• Intent signals (Y/N, which ones)
• Behavioral score (0-100)
• Intent score (0-100)
• Overall account score (0-100)
• Tier (1, 2, 3, or On-Deck)
• Sales owner (if assigned)
• Last update date

Sort by score descending. Update monthly or quarterly.

Manual vs. Automated

Manual (spreadsheet): 100 accounts max - Good enough if you have <100 accounts - Takes 4-6 hours/month to maintain - Team doesn’t forget to check it

Semi-automated (HubSpot, Marketo, Salesforce): - Sync CRM and web analytics - Auto-calculate firmographic score - Manual intent data input (ZoomInfo API possible but complex) - Takes 8-10 hours to set up, 2-3 hours/month to maintain

Fully automated (intent platform + ABM solution): - Tools like 6sense, Demandbase, Terminus do all this - Cost: $20K-100K+/year - Good if you have 500+ accounts to score

Recommendation: Start manual. Move to semi-automated when you hit 100 accounts and have budget.

Step 7: Update the Model Quarterly

Scoring models decay. Change in your business means change in your ICP.

Quarterly Scoring Review

1. Review win/loss data - Which accounts that scored 80+ actually closed? - Which accounts that scored <40 unexpectedly closed? - Adjust scoring weights based on misses

2. Update ICP - Did our best customers differ from our ICP definition? - Did our worst customers share hidden characteristics? - Update ICP in scoring document

3. Refresh firmographic data - Company sizes change (hiring, acquisitions) - Funding changes (new series, IPO) - Re-score all accounts

4. Adjust thresholds - Are you covering too many accounts at Tier 1? Raise threshold to 85. - Are you covering too few? Lower to 75. - Thresholds should cover 10-15% Tier 1, 30-40% Tier 2, 50-60% Tier 3.

FAQ: Building Account Scoring Models

Q: How many criteria are too many? A: 6-8 total (firmographic + behavioral + intent). More than that and your model becomes unpredictable and hard to manage. Start simple. Add complexity only if it improves accuracy.

Q: Do we need third-party intent data? A: No. It helps (maybe 10-15% accuracy lift), but hiring signals and tech stack changes work too. Start free. Add third-party intent if you have budget.

Q: How do we handle existing customers? A: They should always score highest on behavioral signal. They have engagement. For firmographic, score based on expansion/upsell potential, not initial fit.

Q: What if our sales team disagrees with the scores? A: Review 5-10 accounts where you disagree. Update your scoring criteria to match reality. Your model should explain why an account scored low.

Q: How often do we re-score? A: Monthly updates to behavioral signals. Quarterly updates to intent and firmographic data. Annual full review of model accuracy.

Q: What if we score 100 accounts all as Tier 1? A: Your thresholds are too low. Raise them. Or your firmographic criteria are wrong. Revisit your ICP definition.

---

Next Steps

1. This week: Define your ICP with sales and marketing leaders.
2. Next week: Build your firmographic scoring card.
3. Week 3: Add behavioral tracking to your CRM or spreadsheet.
4. Week 4: Implement intent signals (free: hiring, funding; paid: ZoomInfo if budget allows).
5. Week 5: Create your first account score list and validate with sales.
6. Quarterly: Review model accuracy against win/loss data and adjust.

Account scoring is your filter. Build it, use it, improve it.