Revenue AI Strategy
Open the platform
All posts
L4 Maturityinbound 6 min read

Account-Based MQA Scoring: Moving from Lead Queues to Ranked Lists

Learn how to build a predictive MQA scoring engine using 6sense, Demandbase, and CRM automation to prioritize high-intent accounts over low-value leads.

Run the playbook

Account-based MQA scoring (L4)

First-party + 3P signals feed a model that scores accounts (not leads) by ICP fit + intent + engagement. Sales gets a ranked list, not a queue.

Why this matters

The traditional "lead" is a lie in B2B. Large accounts don't buy in a vacuum; they buy in committees. When your Sales Development Representatives (SDRs) are chasing a single marketing-qualified lead (MQL) who downloaded a whitepaper, they are missing the three other stakeholders at that same company researching your competitors on G2.

The cost of sticking to person-level MQLs is a 30-50% waste of sales capacity. Your AEs are spending their most expensive hours cold-calling people who have no intent, while "dark social" and third-party research happen right under their noses. By moving to Account-Based MQA (Marketing Qualified Account) Scoring, you shift from a reactive queue to a prioritized target list. Companies implementing this L4 maturity model typically see a 15-20% lift in MQA-to-SQO (Sales Qualified Opportunity) conversion rates because sales only touches accounts where the "smoke" of intent has turned into the "fire" of a buying signal.

How it works

1. Standardize the MQA Scorecard

Before touching any software, you must align on what "good" looks like. Assemble a task force (Head of Growth, VP Sales, RevOps). You are building a three-dimensional scoring matrix:

  • Fit (40% weight): Pull a 24-month "Closed Won" report from Salesforce or HubSpot. If your winners are $50M+ SaaS companies with 200-500 employees, those are your 100-point fit accounts.
  • Intent (40% weight): Identify 20-50 high-intent keywords. If someone is searching for "[Competitor] pricing" or "[Your Category] integrations," that is a high-value signal.
  • Engagement (20% weight): These are first-party actions. A demo request is 50 points; attending a webinar is 10.

Owner: RevOps & Sales VP | Time: 4-6 hours.

2. Train the Predictive Model

Now, move from theory to automation. Use a platform like 6sense, Demandbase, or ZoomInfo Operations to build a predictive model.

  • Upload a CSV of your last 12 months of 'Closed Won' vs. 'Disqualified' accounts. This allows the AI to learn the behavioral patterns of buyers versus lookers.
  • The Formula: (Fit Score * 0.4) + (Intent Strength * 0.4) + (Engagement Velocity * 0.2) = Total MQA Score.
  • Pro Tip: If you have fewer than 50 won deals, skip the AI and use a heuristic (rules-based) system in Clay or your CRM until you have more data.

Owner: Marketing Ops | Time: 3-5 hours.

3. Implement Signal Decay Logic

A high score from six months ago is a false positive. You need a "decay engine."

  • In HubSpot or Marketo, create a workflow that reduces an account’s engagement score by 10% Every 24 hours if no new activity occurs.
  • Ensure your intent platform is set to "Daily Refresh."
  • Crucial Step: Use a tool like Clearbit or 6sense to filter out your own employees’ IPs so your team’s internal research doesn't trigger a "Hot Account" alert.

Owner: Marketing Ops | Time: 2 hours.

4. Automate High-Intent Routing

The goal is to move data to the rep, not the rep to the data.

  • Use LeanData or Salesforce Flows to set a threshold: If Score >= 80, trigger an immediate notification.
  • Push these alerts via Rattle or Slack.
  • The Message Template: "🔥 HOT MQA: [Account Name] hit score 85. VP of IT searched for '[Your Product] vs [Competitor]' 3 times today. [Link to Outreach Snippet]."
  • Accounts scoring 50-79 should be dropped into an automated HubSpot nurture sequence, not an AE’s task list.

Owner: RevOps | Time: 4 hours.

5. Build the Sales Command Center

Kill the "New Lead" email. Reps should start their day in a Salesforce Dashboard filtered by Account Score (Descending).

  • Include columns for "Top Intent Keywords" so the rep knows why they are calling.
  • Use tools like Momentum.io or Fathom to capture notes from initial calls and see if they validate the MQA score.
  • Hold a bi-weekly feedback loop. If an AE says "This MQA was garbage," go back to Step 2 and adjust the weights.

Owner: Sales Ops | Time: 3 hours.

Tools you need

  • Intent/Scoring: 6sense, Demandbase, or ZoomInfo.
  • Data Enrichment/Orchestration: Clay (for custom scoring logic), Clearbit.
  • Routing/Alerts: LeanData, Rattle, Slack.
  • CRM: Salesforce or HubSpot.

KPIs to track

  • MQA → SQO Conversion Rate: Should see a >15% improvement over the old MQL model.
  • Pipeline from MQA: Total dollar value of opportunities opened via high-score alerts.
  • SDR/AE Efficiency: Calculated as (Pipeline Created / Focused Hours).

Common pitfalls

  • The Notification Flood: If you alert reps for every score change, they will mute Slack. Only route the top 5-10% of accounts.
  • Low Data Volume: Don't use a "Predictive AI" model if you only have 20 customers. It will hallucinate patterns. Stick to manual point-scoring.
  • The "Shadow" Fit: Ignoring the tech stack. If your product only works with AWS, an account searching for "Cloud Security" that runs on Azure is a 0, no matter how high their intent.

When to graduate to the next level

You are ready for Level 5 (Autonomous GTM) when your MQA scoring is so accurate that you can begin using AI Agents (like Lindy or Claude Code scripts) to automatically draft and send initial hyper-personalized outreach based on the intent keywords before an SDR even opens their laptop.

abmscoring6sensedemandbase

Ready to ship it? Open the playbook

Account-based MQA scoring (L4)

Step-by-step instructions, the tools to use, and the KPIs to watch — already wired into the Revenue AI Strategy workspace.

Open playbook