Infrastructure for Lead Scoring & Prioritization

Lead scoring requires that opportunity stages, required CRM fields, and basic qualification criteria are documented and practiced. The baseline confirms proposal templates and pricing guidelines exist. However, the deeper logic of what makes a lead 'strategic' or 'high-fit' remains in partners' heads — the AI scores on observable signals (firmographics, engagement), not on undocumented strategic intent. L2 is sufficient because the system scores what is captured, not what is tacit.

Lead scoring requires systematic capture of firmographic data, engagement signals, and win/loss outcomes with consistent metadata. The CRM must capture lead source, engagement touchpoints, proposal submission, and outcome for every opportunity — not just memorable ones. Template-driven capture ensures the training dataset covers the full distribution of leads, including the small deals and early-stage losses that reveal disqualification patterns.

Lead scoring models require consistent schema across opportunities: Account.Industry, Account.Size, Opportunity.Source, Engagement.Touchpoints, Opportunity.Outcome. The CRM's existing structure (Account → Contact → Opportunity) provides the foundation, but all fields must be standardized — not proliferating practice-specific custom fields. Consistent schema enables the model to compute features like 'time-to-proposal' and 'engagement density' reliably across all leads.

The lead scoring system must query CRM opportunity data, pull engagement signals from marketing automation, and access website behavior data via API at scoring time. Modern CRMs like Salesforce provide the API access needed. The system scores leads as they arrive — requiring real-time API queries to assemble firmographic, engagement, and historical context for each new lead without manual data assembly.

Lead scoring models need event-triggered updates when win/loss patterns shift — new service lines enter the portfolio, pricing changes alter competitive position, or a new market segment emerges. When the firm wins a cluster of deals in a new industry, that signal should update the model's industry-fit scoring without waiting for quarterly review. Event-triggered maintenance keeps score calibration aligned with actual market performance.

Lead scoring requires API-based connections between CRM (opportunity data), marketing automation (engagement signals), website analytics (behavior data), and email platforms (open/click tracking). These systems must share context — the score for a lead combines firmographic fit from CRM with engagement density from marketing automation and behavioral signals from web analytics. The baseline confirms CRM-to-marketing-automation integration exists, covering the primary data flows.