ABM Account Scoring Framework

Introduction

Account-based marketing (ABM) starts with a simple problem: your GTM team has infinite targets and finite resources. You can’t execute a personalized, multi-threaded, account-orchestrated motion on 10,000 accounts. You have to pick.

Account scoring is the lens through which you make that pick. But most scoring frameworks are built wrong. They’re either:

1. Too simplistic (fits any lead that meets a basic firmographic filter).
2. Too rigid (built once, never updated).
3. Too siloed (sales doesn’t trust the criteria; marketing doesn’t understand the business logic).
4. Too late (you score the account after months of inbound activity, by which time the decision is already moving).

This guide walks you through building a scoring framework that avoids all four pitfalls. We’ll cover the mechanics of scoring, how to integrate intent data, how to version and refine your model over time, and how to get both sales and marketing aligned on the output.

Why Account Scoring Matters

In a traditional lead-generation motion, scoring measures likelihood-to-convert: What’s the probability that this lead, once touched, will become a customer? That’s a funnel-optimization problem.

In ABM, scoring answers a different question: Is this account worth our ABM investment? That’s a resource-allocation problem.

The distinction matters. A lead might convert despite not being a strong ABM target (e.g., a small, low-touch deal). Conversely, a huge enterprise account with 0 current inbound demand might be a perfect ABM target.

Scoring in ABM should predict: - Strategic fit (does this account match our ideal customer profile?). - Revenue potential (what’s the ceiling on contract value?). - Accessibility (can we realistically get in and orchestrate multiple buying groups?). - Timeliness (is there purchase intent activity now, or are we cold-outreach only?).

A well-built scoring model helps you: - Prioritize your research and account planning work upfront. - Allocate account executives and marketing resources to accounts where the effort will compound. - Create a repeatable process so you scale ABM without adding proportional headcount. - Build a shared vocabulary between sales and marketing about what “good” looks like.

Core Components of an Account Scoring Framework

1. Firmographic Scoring

Firmographic data describes the company: industry, size, location, funding status, technology stack. These are usually sourced from enrichment providers (ZoomInfo, Apollo, Hunter, or intent-data platforms).

Firmographic scoring asks: Does this company fit our ICP (Ideal Customer Profile)?

Common firmographic variables: - Industry vertical. - Company size (employee count or revenue range). - Geographic region. - Revenue stage (profitable, funded, bootstrapped). - Technology adoption (has they implemented a CRM, CDP, marketing automation platform?).

Build your firmographic criteria by analyzing your best-fit customers: - What industries dominate your wins? - What company-size range has the longest contract value and fewest deal cycles? - Are you primarily selling to certain regions? - Do your customers tend to have specific technology profiles?

Implementation tip: Don’t score 0/1 (yes/no). Use a scale (0-100 or 0-25, depending on your system). A company that partially fits your ICP should score proportionally.

2. Intent Scoring

Intent data captures behavioral signals that a company is actively evaluating solutions in your category. Common intent signals include:

• Website visits from company IP addresses.
• Downloads of content or assets.
• Attendance at webinars or live events.
• Search-behavior signals (keywords the company has searched).
• Job postings (hiring for roles that correlate with buying).
• News and funding events.
• Technology changes (implementing new tools, sunsetting others).

Intent scoring is crucial because it tells you when to reach out. An account might have a perfect ICP fit, but if there’s no purchase intent activity, you’re likely early in their buying journey.

Map intent signals to scoring weights:

High-intent signals (strongest predictor of immediate opportunity): - Website visits + form submissions. - Document downloads. - Webinar attendance. - Specific keyword searches related to your solution.

Mid-intent signals (indicates growing interest): - General industry searches. - Job postings in related functions (Sales Development, Revenue Operations). - Technology stack changes.

Low-intent signals (early-stage awareness): - News mentions. - Funding events. - Attendance at industry conferences.

Implementation tip: Decay intent signals over time. A website visit from 30 days ago is less predictive than one from 3 days ago. Use a rolling window (e.g., 30-day recency) in your scoring model.

3. Relationship and Engagement Scoring

These signals measure your existing footprint and traction within an account:

• Number of engaged contacts at the account.
• Tenure and seniority of those contacts (first-touch vs. multi-touch conversations).
• Open rate and click rate on emails sent to the account.
• Response rate to outreach.
• Alignment of engaged personas (are you talking to economic buyers or influencers?).

Relationship scoring recognizes that some accounts already have momentum. If marketing has been nurturing 3 contacts at an account, and your sales team has had initial conversations, that account is higher-priority than a cold target with equal fit and intent.

4. Technographic Scoring

Technographic data describes the technology stack an account uses. Common sources include:

• Browser-based stack detection (G2, Crunchbase, Clearbit, Apollo).
• Job descriptions and company websites.
• Intent-data platforms with technology signals.

Technographic scoring asks: Does this account’s current tech create a wedge for us to enter, or a moat we can’t overcome?

Examples: - If you sell a Salesforce app, accounts that already run Salesforce score higher. - If you sell a demand-gen platform, accounts that already use marketing automation (HubSpot, Marketo) score higher. - If you sell a replacement for a competitor’s product, accounts using that competitor score higher.

Building Your Scoring Model

Step 1: Define Your Tiers

Most ABM strategies use 3-5 account tiers:

Tier 1 (Largest ABM investment): Named accounts with executive-level sponsorship, custom marketing campaigns, dedicated ABM teams. Typical criteria: ICP fit > 75, intent activity in last 30 days, revenue potential > $100K ARR.

Tier 2 (ABM-light): Accounts that fit your ICP, show intent, but may have lower revenue potential or fewer relationship touchpoints. Executed by shared ABM resources or demand-gen teams. Typical criteria: ICP fit > 60, intent activity in last 60 days, revenue potential > $30K ARR.

Tier 3 (ABM-nurture): Accounts in your target market with weak current intent. These are early-stage plays, ideal for nurture and thought-leadership content. Typical criteria: ICP fit > 40, no intent filtering, used for bottom-funnel nurture.

Tier assignment should be: - Rules-based (not manual). Define the scoring thresholds upfront. - Transparent. Every account can explain which criteria elevated or lowered its score. - Dynamic. Accounts should re-tier quarterly or whenever significant intent or engagement activity occurs.

Step 2: Assign Weights

Your scoring model is a weighted sum of your component scores:

Overall Account Score = (Firmographic × 40%) + (Intent × 35%) + (Relationship × 15%) + (Technographic × 10%)

These weights are illustrative. Your weights depend on your business model:

• If you’re selling to net-new markets, intent might weight heavier (40-50%).
• If you’re selling to existing customers in new divisions, relationship scoring might weight heavier.
• If you’re selling a point solution with a specific tech stack requirement, technographic might weight heavier.

Implementation tip: Start with conservative weights. Bias toward firmographic fit first, then layer in intent and relationship. You can refine weights after 2-3 scoring cycles.

Step 3: Integrate Intent Data Properly

Not all intent data is equal. Consider:

1. Zero-party intent (strongest): Contacts who’ve engaged with you directly (email opens, webinar attendance, form submissions).
2. First-party intent (strong): Your own website activity, CRM data, sales interactions.
3. Second-party intent (moderate): Data shared via partnerships or integrations.
4. Third-party intent (weakest): Purchase-intent platforms, keyword search data, publicly available signals.

Use a mix, but don’t over-index on third-party signals. Your own engagement data is more accurate.

Step 4: Set Scoring Thresholds

Define the thresholds that move accounts between tiers:

• Tier 1: Overall score >= 75
• Tier 2: Overall score >= 55 and < 75
• Tier 3: Overall score >= 30 and < 55
• Out of scope: Overall score < 30

These thresholds will change. After your first 2-3 months of scoring, analyze your closed-won accounts. What was their average score at first contact? Adjust thresholds to match your observed data.

Refining Your Model Over Time

Account scoring isn’t build-once. It evolves.

Monthly Model Review

1. Identify accounts that won deals in the last month.
2. What was their score 3, 6, and 12 months before close?
3. Did higher-scoring accounts have shorter sales cycles?
4. Were there accounts that won despite low scores? Investigate why (missed signal? external factors?).
5. Were there accounts with high scores that stalled or churned? What changed?

Use this analysis to adjust weights and signals.

Quarterly Re-scoring and Re-tiering

Re-run your scoring on your entire account database. Accounts move between tiers as intent and engagement shift. This prevents score decay (accounts that you scored 6 months ago may have completely different intent signals now).

Annual Audit

Once a year, review your entire framework: - Do your tier definitions still reflect your GTM strategy? - Have new intent signals emerged that you should incorporate? - Has your ICP evolved? - Are there firmographic criteria that don’t correlate with wins anymore?

Adjust and republish.

Aligning Sales and Marketing on Scoring

The biggest risk in account scoring is building a model that marketing loves and sales ignores.

Prevent that:

1. Involve sales in the design. Don’t build scoring in marketing and present it to sales. Co-design. Ask: What signals do you see in your own deals that predict success? What account characteristics make your deals easier or harder?

2. Make scoring transparent. Every ABM email and sales conversation should reference the account’s tier and key scoring signals. “We’re prioritizing you as a Tier 1 account because of your company size, recent expansion into [market], and active engagement with our content.”

3. Create feedback loops. Sales should be able to flag accounts they believe are mis-scored. Create a simple form or Slack channel for feedback. Review feedback monthly.

4. Celebrate Tier 1 accounts. Once an account is scored, make sure your GTM team treats it accordingly. Tier 1 accounts should get faster response times, more personalized outreach, and executive touchpoints. If they don’t, sales won’t believe the scoring system.

FAQ

Q: How often should we re-score accounts?

A: Minimum quarterly. For high-velocity teams with mature intent data, monthly re-scoring is ideal. The goal is to catch intent peaks (when accounts are most engaged) and ensure you’re not spending effort on accounts where intent has decayed.

Q: What if we don’t have intent data?

A: Start with firmographic and relationship scoring. Build a framework that works with your current data infrastructure. Once you implement an intent platform, layer it in. Intent data is a multiplier, not a prerequisite.

Q: Should we score accounts that are already customers?

A: Yes, separately. Use a upsell/expansion scoring model for existing customers, weighted toward expansion opportunity (team size, product usage, feature adoption) rather than new-customer ICP fit.

Q: How do we handle accounts that are too small to find firmographic data on?

A: Set a minimum company size threshold. Accounts below that threshold stay out of your ABM database (they might be good for self-serve or low-touch). For accounts where data is incomplete, build a process to flag and enrich them. Don’t score incomplete profiles; it introduces bias.

Q: What if sales overrides the scoring system and decides to pursue a low-scoring account?

A: That’s fine. Not all account decisions can be rules-based. Build a “sales override” workflow that lets sales elevate accounts, but require them to document the business rationale. Track overrides and use them as feedback to improve your model.

Q: How do we handle scoring for accounts in different product lines or segments?

A: Build separate scoring models. Your enterprise SaaS product might have different ICP, intent, and revenue-potential criteria than your SMB product. Trying to score both on a single model dilutes the signal.

Conclusion

Account scoring is the foundation of ABM. It forces you to articulate what “good” looks like, creates transparency between sales and marketing, and gives you a repeatable process to scale.

Build conservatively, measure ruthlessly, and refine quarterly. Your first scoring model doesn’t need to be perfect. It needs to be better than your current account-prioritization process (which is probably “largest companies first” or “whoever has the most recent inbound”).

Ready to implement a scoring framework that works across your sales and marketing teams?

Book a demo with Abmatic AI to see how intent data and account analytics accelerate your ABM motion.