Big Data is not like a Big Mac
Lots of companies have experimented with Big Data solutions because they were told it is the 'next revolution' (just like data mining before it...).
However, there are 2 truths that few could deny:
1. The number of customers who have try 'big data' needs is quite small - small enough that I don't think there's room for the kind of growth that Hortonwork's early investors were betting on.
2. The techies who have to sell this stuff haven't figured how how to apply their big data capabilities to solving enough concrete business problems (all the clients I worked with who wanted this stuff already had a good idea of what they needed it for).
So extremely niche market, over-funded startup and a bunch of geeks who struggle to actually demonstrate use cases. That's why they had to IPO... their investors needed to unload their shares to the Big Stupid.