How transaction-based testing can better predict customer behavior
Posted on Harvard Business Review: June 13, 2011 3:15 PM
"Our new service tested off the charts—more than 5% of customers we surveyed said they would buy it," one reader at a not-for-profit organization emailed me. "We launched it. The results were disappointing."
Remember, customers lie. Not maliciously, but they lie. The reader learned that lesson the hard way. For his next idea he wanted to run a transaction-oriented test, but struggled with how to design and execute one.
Transaction-based learning need not be overly complicated. Consider these four options:
1. Ask potential consumers to sign up for something. We recently developed a novel idea for a consumer electronics company. The idea would take about 18 months to fully commercialize, and would require substantial investment. How could the company tell if it would be worth it? The team created a mockup of the idea and showed it to consumers. At the end of the demonstration, they asked consumers if they would be willing to pay a nominal fee to be part of a "private beta test" once the offering when it was ready. It was a great way to determine if consumer excitement would translate into even modest payments.
2. Test the idea among your employees. Most organizations have an untapped asset walking their hallways every day. In Procter & Gamble facilities you often will see project teams pitching new ideas outside the corporate cafeteria. If your own people won't buy it, why would consumers?
3. Introduce a "10% version" of the offering in a highly contained market. Imagine you had an idea for a one-week seminar that intersects strategic innovation and design thinking, bringing together world-class speakers, global attendees and unique tools. Participants would build businesses in real time and get feedback—nd potentially funding—from venture capitalists. Sounds like a great idea, but would anyone come? One way to learn is to hold a half-day seminar, with a pretty good speaker and "good enough" tools in a single city.
4. Hold a "sign-up seminar." A few years ago we helping a startup develop an idea for a medical tourism business where customers seeking elective procedures would fly to exotic but lower-cost destinations. As my colleague Matt Eyring described in a recent Harvard Business Review article, we had meticulously studied the market, and the projections looked solid. One of our colleagues suggested that we hold seminars in locations replete with the target customers to gauge interest. When no one showed up, we realized that the stated interest was unlikely to translate into real demand.
These "low resolution" tests aren't perfect, but they are better than simply asking a customer, "Do you want this?" Remember, discount what customers say. Pay very careful attention to what they do.