FordDirect, Ford's joint venture with its dealers, deployed AI agents in Domo to automate the complex process of analyzing why dealerships left their AdVantage digital advertising program. The system tackled a real operational problem: with over 3,000 Ford and Lincoln dealers, manually analyzing performance data for the 97 at-risk or terminated dealers in 2025 was a nightmare of data integration across different timelines, metrics, and external market factors like seasonality and tariffs.

This is exactly the kind of unglamorous but valuable AI work that actually moves businesses forward. Instead of chasing chatbot demos or flashy consumer apps, FordDirect focused on a specific workflow pain point where AI could deliver measurable ROI. The agent compares three months of dealer performance before and after leaving AdVantage, generating standardized reports that sales teams can actually use in dealer conversations. Brendan Sullivan, FordDirect's director of advertising analytics, chose Domo over Databricks specifically for its data governance capabilities — a smart move when you're dealing with dealer relationships worth millions.

What's missing from the coverage is the success rate — did these AI-generated insights actually win dealers back? The story focuses on the technical implementation but glosses over business outcomes. Also unclear is how much human oversight the agent requires, or whether it can handle edge cases like dealers who left due to ownership changes or market exits rather than performance issues.

For developers building similar systems, the key insight here isn't the AI itself but the process: start with manual workflow analysis, build solid data foundations first, and focus on governance from day one. The real innovation is turning scattered business intelligence into repeatable, scalable insights that non-technical teams can trust and act on.