Microsoft Dynamics 365 Customer Experience Analyst : Configure predictive lead and opportunity scoring models

How a Microsoft Dynamics 365 Customer Experience Analyst would configure predictive lead and opportunity scoring models step-by-step:

1. Understand the Business Goals

Before setting up, clarify:

• Lead scoring goals: e.g., increase conversion rate, prioritize hot leads.
• Opportunity scoring goals: e.g., focus sales on deals with the highest probability to close.
• KPIs: time to conversion, win rate, revenue uplift.

2. Ensure Prerequisites

•  Dynamics 365 Sales Insights license enabled.
•  Leads and opportunities have sufficient historical data (minimum \~40–100 records with meaningful patterns).
•  Sales team actively tracks lead qualification, opportunity stages, and outcomes.

3. Enable Sales Insights

1. Go to Power Platform Admin Center → Environment.
2. Select environment → Settings → Features.
3. Turn on Sales Insights.
4. Accept terms and provision AI models.

4. Configure Predictive Lead Scoring

1. In Sales Hub → App Settings → Predictive Lead Scoring.
2. Click Set up.
3. Select the qualifying criteria (e.g., lead source, industry, budget, interactions).
4. AI analyzes historical lead conversions to train the model.
5. Review top contributing factors (e.g., email open rate, account size).
6. Save and activate the model.

5. Configure Predictive Opportunity Scoring

1. In Sales Hub → App Settings → Predictive Opportunity Scoring.
2. Click Set up.
3. Choose factors from historical opportunities (e.g., estimated revenue, close date, contact engagement).
4. Review AI’s correlation analysis for win probability.
5. Activate the scoring model.

6. Deploy & Surface Scores

• Lead score: Appears on lead forms and views (e.g., “Score: 87/100 — Hot”).
• Opportunity score: Appears on opportunity records with reason codes.
• Create views like:
• Hot Leads (Score ≥ 80)
• Opportunities Likely to Close This Month

7. Monitor & Adjust

• Review scoring model accuracy in the model performance dashboard.
• Re-train the model as more data comes in.
• Adjust criteria if business priorities change.

8. Real-World Impact Example

• Before scoring: Sales reps chase leads randomly → 20% conversion.
• After scoring: Focus only on leads with score ≥ 75 → 35% conversion.
• Revenue impact: Faster close cycles, higher win rates, better resource allocation.