The Hypotheses Scorecard and Hypotheses Dependencies
The first step in the Customer Discovery phase of the Customer Development process is to state and record all the various hypotheses you have about the customer, market, product, price, competition, environment, dependencies, and pretty much everything you can think of. In the course of developing hypotheses for the Yobiz eRetail IPS product, we jotted down over 150 hypotheses.
To keep these hypotheses straight, we constructed a Google doc that categorized each of the hypotheses and assigned to them a unique number. Then, we entered each of these into a spreadsheet so that we can track them and determine whether or not we could confirm those hypotheses.
Here’s a quick snapshot of what this Hypotheses Detail Document looks like:
Hypotheses Aren’t Independent: Hypotheses Dependencies
In doing these hypotheses, we noticed one thing that is not mentioned at all in the Four Steps to the Epiphany book, which serves as our guide during the Customer Development process — these hypotheses are not independent. That is to say, there are dependencies between various hypotheses. If one hypotheses turns out to be false, then others are likely false as well. If one turns out to be true, then some others might turn out to be true.
For example, if you can prove a hypothesis about a particular product benefit, and a specific product feature delivers that benefit, and the first product release (or MVP) implements that feature, then it all flows together. However, if that product benefit turns out to not be proven or even false, then that whole daisy chain is likely false as well. Given the tens of dozens of hypotheses put together, it requires a significant effort to track those dependencies.
Also, what we noticed is that there was a certain hierarchy to the hypotheses. Certain categories of hypotheses in turn depended on others in a logical manner. This means that there is also a priority to addressing the hypotheses. Certain core hypotheses need to be tested as strenuously and early as possible since so many others depend on them.
We actually put together a diagram of these hypotheses dependencies. I think this is the first time this has been done in the Customer Development world, so it would be great to get some feedback on this:
Keeping Track of Hypotheses Dependencies: The Hypotheses Scorecard
Continuing our work in this area, we determined we needed a way to identify just how many dependencies a particular hypothesis had, and likewise, how many other hypotheses depended on that one particular hypothesis. To address this, we built a Hypotheses Scorecard: a matrix of hypotheses mapped against hypotheses with an indicator of “1″ in the matrix cell if there is a dependency, and “0″ if there is not. The resultant sums across the rows and columns identifies just how many dependencies there are for that hypothesis and likewise how many hypotheses are dependent on that hypothesis.
Below is an example of a Hypotheses Scorecard:
Sharing our Customer Development Tools: Should We Build an App?
Sounds tedious? Well it is. But there’s no other way to do this in a quasi-scientific format. We’re actually thinking of turning this and the rest of the tools we’ve developed for our own Customer Development activities into a helpful app to help you manage the whole Customer Development process. If you’re interested in that, please drop us a note — if there’s critical mass, we’ll share the tools we’ve been using — either for free or for a small amount of money, depending on interest and demand. Mind you, this is not our core business, but if the tools have been helpful to us, maybe they’ll be helpful to you.