Launching a new product into a high-trust market is a different problem from launching into a competitive but familiar one. Healthcare, financial services, legal — these are categories where the buyer's default position is skepticism, and where generic messaging doesn't just underperform, it actively damages credibility. The product might be genuinely innovative, but if the go-to-market approach doesn't reflect a real understanding of the buyer's world, it signals that the company doesn't understand the problem it claims to solve.
For AI products specifically, the trust barrier is even higher. The technology is unfamiliar to many buyers, the stakes of getting it wrong feel significant, and the market is full of overclaiming. Companies that cut through do so by demonstrating they've done the work to understand who they're selling to — their specific fears, their specific language, and the specific moment in their life or workflow where the product actually helps.
Awarded NIH Grant to help older Americans age in place independently using privacy-protected, AI-based monitoring.
CherryLabs was bringing an AI healthcare product to market targeting elderly users — a segment with distinct communication preferences, high caution around technology adoption, and very specific decision-making dynamics. And the decision-making dynamic here had an additional layer: it wasn't the elderly users themselves who were the primary buyers. It was their adult children — who researched the options, evaluated the product, and made the purchase, while consulting their parents on whether to proceed. Two audiences, two sets of concerns, one acquisition strategy that had to speak to both.
The challenge was building validated acquisition campaigns from scratch, with no prior customer data and a 30-day window to generate meaningful results. That required getting close to both groups first — understanding their language, their concerns, and the framing that would make an unfamiliar technology feel relevant and safe.
The work started with direct research — conversations with elderly early adopters in the healthcare space, and with their adult children who were the actual purchase decision-makers. Both groups were mapped separately: how elderly users talked about their situation and what made a technology feel approachable, and how their children evaluated options, what reassurances they needed, and what would make them confident enough to buy. That research became the raw material for concept development.
Fifteen promotional concepts were developed and tested across different angles — varying the framing, the tone, the specific benefit highlighted, and the level of technical language used. Rapid iteration meant concepts that didn't resonate were cut quickly, and the ones that did were developed further. The result was a validated messaging foundation that reflected the actual language and concerns of both segments.
Acquisition campaigns were built directly from that foundation. The targeting was specific, the messaging was tested, and the leads that came in reflected the quality of the underlying research.
Fifteen concepts in three months sounds productive. In practice, it wasn't efficient. Some concepts were weak from the start — but instead of cutting them early, I tested everything in sequence and lost time on hypotheses that were never going to work.
The lesson was straightforward: kill criteria need to be defined before the test starts. Now I set explicit thresholds at the beginning — if a concept isn't generating signal after a defined number of iterations, it's cut. The decision is made on criteria, not on gut feel mid-process.