Edge AI Commercialization Strategy

An early-stage company in the tech industry had a working technology — an on-device (“edge”) AI agent — and needed to decide how to take it commercial: who the target customer is, which capabilities matter most, and how to price and go to market. Over two academic quarters, our team ran the market and customer research to inform that decision.

A mixed-methods, hypothesis-driven study. The work spanned:

• Competitive and market-landscape research.
• Around 50 expert interviews across executives, operations and IT leaders, and technical staff.
• Two surveys, analyzed with hierarchical and K-means clustering to build behavioral customer segments rather than demographic ones.
• Opportunity scoring and a Van Westendorp price-sensitivity analysis.
• Porter's Five Forces, SWOT, and TAM/SAM/SOM market sizing.
• A multi-year financial model.

I led the secondary research on the regulatory and technology landscape and ran a portion of the primary interviews, then contributed to the thematic analysis and the customer segmentation that came out of the survey data.

What a technical team believes makes their product valuable, and what the market actually values, can diverge — bridging that gap is much of the work.

Clustering on attitudes and behavior gives a sharper, more actionable target than demographic segmentation.

The research has to test the market on its own terms, not the client's. A client's assumptions about who wants the product and why can quietly shape which questions get asked and how the answers get read — the study has to be designed to resist that and validate demand without the bias.

Stating a study's limitations plainly matters as much as the findings themselves.