Predictive Outreach AI
How Skapa Transformed Stakeholder Education and Outreach with Predictive AI
Top-Line Metrics
- Predictive risk identification —
Skapa built an AI model that identifies high-risk excavator and member segments, enabling our customer to target outreach where it matters most. - Data-driven resource allocation —
Marketing and outreach resources are now prioritized based on predictive insights rather than broad assumptions, maximizing the impact of every campaign. - Performance recognition —
High-performing members are identified and rewarded, reinforcing safe behaviors and strengthening engagement across the state.
About the Client
Statewide 811 Utility Protection Center
The Problem
Our customer's marketing outreach programs relied on generic safety messaging and broad prioritization. This limited the effectiveness of communications and prevented marketing and business units from focusing on members with higher risk profiles.
The organization initially considered using AI to predict damages from incoming tickets and historical records. While prediction was possible, prediction alone wouldn't tell anyone what to do next. Without a way to translate data into action, the insights were useless.
What they needed was a system that could prioritize and make insights actionable. One that could help their marketing and business units prioritize outreach, recognize high-performing excavators, and reduce risks where possible — without requiring manual analysis every time.
The Solution
Skapa partnered with this notification center to implement an AI-driven intelligence system to support targeted marketing and outreach.
The solution consisted of a three-layer model:
- Layer 1 — Clean | Data Normalization: Standardized excavator records to merge variations of the same company name and ensure accurate reporting.
- Layer 2 — Segment | Peer Group Classification: Grouped excavators into categories such as electric utilities, water/sewer utilities, and specific contractors to enable comparisons within peer groups. This made it possible to compare apples to apples: a small contractor isn't evaluated against a large utility. Performance and risk are assessed within the right context.
- Layer 3 — Grade Layer | Risk Scoring: Assigned A–F risk grades based on historical damage data and key contributing factors, with risk calculated as probability × impact.
With these layers in place, Skapa enables our customer to:
- Target outreach to high-risk excavators
- Recognize and reward high-performing excavators, reinforcing behaviors that reduce damages
- Prioritize marketing and safety resources efficiently and based on real data
- Provide actionable data that business units could use directly