Building the Account Score

A scoring model that ranks accounts by purchase likelihood is one of the highest-value assets a sales team can build. This lesson walks through building one from scratch.

Step 1: Define Your Scoring Attributes

List every attribute that correlates with a successful closed-won deal in your business. Start by analyzing your last 30 closed-won customers and looking for patterns. Common high-correlation attributes:

• Industry (specific verticals you win most in)
• Employee count range
• Technology stack matches
• Funding stage and recency
• Hiring signals (specific titles being hired)
• Website visit behavior (page type, session depth, return visits)
• G2 intent match

Step 2: Assign Weights

Not all attributes predict equally. Industry match and page visit behavior typically carry the most predictive weight. Assign points based on your analysis: industry fit (0 or 25 points), employee range (0 or 15), tech stack (0 or 20), pricing page visit (0 or 30), return visit (0 or 15), etc.

Step 3: Validate Against Historical Data

Score your last 50 closed-won and 50 closed-lost deals retroactively. If your model is working, closed-won deals should average 20–30 points higher than closed-lost. If they don't, adjust the weights until they do.

Step 4: Set Action Thresholds

Define: Score 80+ = immediate outreach (same day). Score 50–79 = standard queue (within 48 hours). Score 30–49 = low-priority queue (weekly review). Score below 30 = no immediate action, monitor only.