Learning faster

Post created for Technology, Policy, and Public Service Innovation Class.

I did some work at medical device company that sold a number of urology products. One of their products was a non invasive testing machine. Its target market was doctors in a certain specialty with private practices. The device sold for a couple of thousand dollars and it had a disposable that was used every time the doctor used the machine to test a patient. We were pricing the disposable at around $50. This product had been around for a few years, and there were a few doctors who had it, but they rarely ordered new disposables.

The company had just received capital from investors who wanted the company to use it to figure out the right sales model for the product.

My role, although I was not the only one assigned to work on this, was to help the company figure out how to make this product as profitable as possible.

We had a couple of hypotheses as we started to consider a new sales strategy. The most important was that the potential of the disposable revenue stream dwarfed that of the sale of machines. A practice that used the machine twice a week for two years would provide us with as much revenue as the initial sale of the machine!

What we needed to learn was, how do we sell as many disposables as possible? We did test a number of strategies in attempt to figure out this question, but if we had applied a more rigorous approach to hypothesis testing we could have figured it out much faster.

Our first attempt to increase disposable sales was to rent out the machines and give economic incentives for practices that used more disposables. This turned out to have little effect on the problem, which we could have learned much faster with A/B testing. Rather than offering a rental program to all of our clients we should have segmented our population to be able to compare disposable use for those who owned and rented the machine. With this approach we could have more quickly moved to the next potential strategy.

Through this strategy and others we finally realized that the practices we were selling to had almost no change in their use of our product given any potential economic incentive. Doctors often said to us in sales meetings that they are very sensitive to the price of products and their ability to get insurance reimbursement for particular tests or procedures. But their behavior indicated that they were, ironically given their mercenary rhetoric, mostly acting in a way that was best for their patients, independent of the price of delivering service.

Had we learned this earlier we could have much more quickly pivoted to a strategy that highlighted the clinical benefit of our device, which was a non-invasive alternative to an existing test, and abandoned strategies that focused on pricing. Had we been A/B testing our economic strategies from the beginning, we would have much more quickly discovered our client’s motivations.