Nielsen Ad Intel: an expert system that shouldn’t need an expert

Stakes

Ad Intel is a powerful advertising-intelligence product that only experts could use. Casual users — marketers building a pitch or an RFP response — lost days to pivot tables and exports,, and the business was losing accounts to competitors.

The bet

If the product answered the questions casual users actually bring, in the order they bring them, the same data would reach a decision in hours instead of days.

Three moves

Mapped time to insight. Traced the current workflow step by step against a competitor’s. Why: it showed where the days went.

Researched in rounds. Internal experts first, then casual-user panels with low-, medium- and high-fidelity prototypes, each round narrowing the next. Why: cheap rounds early, expensive ones only where the answer mattered.

Turned findings into a plan. An ideal-customer profile, an impact chain from activities to revenue, a risk model and a prioritized release roadmap. Why: research that doesn’t become a sequence of decisions doesn’t ship.

What changed

This case is a strategy engagement: the deliverable was the direction and the roadmap, not a shipped metric.

What I’d do next

Instrument time to first insight in the product so the next release can be measured, not argued about.