Paralysis by Analysis

In the process of thinking through how discovery-driven planning might apply to our venture Adrianto and I started thinking about some of our key assumptions.  A big assumption that we need to figure out how to test is related to the demand for our rain suits.  More specifically will consumers see the benefits of our design to the extent that they are willing to pay double or triple the price of a generic rain suit.  The most effective way to test this assumption would probably be to set up a stand or hit the streets of Jakarta with a couple of sample rain suits and see how they sell.  Of course this would actually require us to have rain suits already designed and manufactured not to mention that it would help significantly if we were in Jakarta to be able to pull this one off.  So Adrianto and I are brainstorming other ways to start to test the validity of our assumption from here in Pittsburgh without that ability to demo a physical sample to potential customers. 

One method we are considering using to test our assumption is a conjoint analysis.   This is a pretty cool survey technique for simulating markets that Peter Boatwright preaches in his New Product Management (Tepper) class.  This website, http://www.sawtoothsoftware.com/conjoint-analysis-software, has a pretty boring but rather thorough short video of how conjoint works.  For our rain suit project we could easily set up a conjoint survey to determine how much consumers are willing to pay for our rain suit design as well as approximate market share our rain suit could capture relative to the other rain suits on the market.  This would help us eliminate some of the uncertainty surrounding our rain suit demand assumptions.  However one of the limitations for us taking a conjoint survey approach is that we are limited in the extent of which we can actually communicate the benefits of our design in words and pictures, however the results of a conjoint survey would probably be interesting and at least somewhat useful. 

Another limitation to a conjoint survey is that well at the end of the day it is still just a survey which asks people would they buy.  The risk here is that what people say they will buy may or may not match reality when they are actually faced with a purchasing decision.  To overcome this limitation we are considering actually asking people to buy.  We can do this by setting up a website that is meant to look real for our rain suit with pictures, specifications, and an option to purchase the product.  For only a couple hundred dollars we could generate some traffic with google adwords and when users click on the button to purchase our rain suit we will send them to a dummy page that indicates the item is currently out of stock but record their click in a database.  This will allow us to compare how many users who visit our page would actually be willing to buy the product.  Furthermore it would allow us to test various page layouts and prices to see what effect they have on sales.  This approach to testing a market was popularized by Tim Ferris in his NY Times best seller The 4-Hour Workweek.

In either case it begs the question how much time, energy, and dollars do we want to spend researching and testing our assumptions?  Both options I presented are rather resource intensive and the usefulness of the data received from these test is uncertaint and something we need to consider further.