Friday, November 25, 2011

Reading :: The Lean Startup

The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses
By Eric Ries


Eric Ries had a problem. As cofounder and CTO of IMVU, a Silicon Valley startup, he knew what his customers wanted: to connect instant messaging with virtual reality avatars.

Initially, the idea was to create avatars you could use to interact virtually with your friends. Your friends might be on different IM networks such as AIM, Yahoo Messenger, or iChat – which might sit on your desktop like separate spigots at a soda fountain. But IMVU would interoperate with all of them, serving as a sort of universal service for connecting your friends across all of these networks.

Users hated it. Adoption at first was incredibly low.

In fact, in IMVU’s in-house tests, users enjoyed creating avatars and using them to chat with strangers. They enjoyed meeting new people anonymously. But they did NOT want to tie together their IM networks, and they certainly did not want to allow these strangers in their AIM buddy lists. They were strangers!

Ries says that he and the others at IMVU had a hard time accepting this. He kept dismissing what the test users were saying, assuming that they simply didn’t represent IMVU’s market. But every new batch of test users said the same thing. Meanwhile, IMVU’s usage stats remained essentially flat.


Ries had seen his product as a way to help old friends interact in new ways in the same place. That’s what would make the service sticky. But the test users consistently told him they didn’t want that. They wanted a way to create virtual identities and use them to meet new friends who also had virtual identities. In essence, he wanted a clubhouse and they wanted a costume party. The two objectives implied different products, different services, different marketing strategies – in fact, they implied that IMVU would have to pivot.

As Ries makes clear, they could very well have ignored this issues and consequently failed in the marketplace, chasing after nonexistent users. But instead they decided to pivot. They wrote off the sunk costs in the first option, including a considerable amount of interoperability code that Ries had written himself. And they reformed their strategy.

But they didn’t capitulate – they synthesized their vision with what the customers would accept. That synthesis involved changing the product, of course, but also the process and internal structure of the company. Ries instituted a set of continual feedback metrics to better regulate this synthesis, building a product and service that helped users discover needs they didn’t know they had. As Ries put it:
We adopted the view that our job was to find a synthesis between our vision and what customers would accept; it wasn’t to capitulate to what customers thought they wanted or to tell customers what they ought to want. (Reis 2011, p.50)
It worked. The flat growth became steep growth. And Ries developed a set of techniques for establishing continuous feedback and improvement in startups. These include qualitative and quantitative measures, but also guidance on when and how to pivot. It's a fascinating book: fascinating as in well written and interesting to read, but also fascinating as in sparking ideas for innovating quickly and interpreting feedback on a fast cycle. If you're involved in startups, software development, or other fast-paced innovation, read it.


2 comments:

Ben Linders said...

My experience is that the lean startup approach brings you in contact with your (potential) customers, by piecewise providing solutions for their needs. We're using this approach to develop IT change management training modules, for veranderproject (in Dutch).

Some of our course ingredients fail, they should since we're trying out new stuff (new to our customers that is). We get tremedous feedback, and are improving modules continuously. Together with our customers we develop an understanding what works in change management, and what doesn't. Which helps them to improve the way they run their IT development.

Let me recap some of our experiences with Lean Startup; I'll publish them at my personal blog.

Clay Spinuzzi said...

Fantastic - veranderproject sounds really interesting (from what I can gather using Google Translate) and I very much like the continuous improvement, fail-faster approach you describe. It would be really interesting to apply this same approach in university-level education, although we unfortunately don't have as much discretion in changing our curriculum.

I'll look forward to your Lean Startup experiences - I just put your blog in my RSS reader.