Entering New Market? Release Now (And Release Often)

Today I'm going to talk a bit about software development. This may seem like a off topic subject, but If you think about a little, you'll soon find a connection. After all most of social networks involve software of one kind or another, and they are all developed by a team of developers. Talking about social network without touching on how it is developed would be akin to talking about pastry without baking.


I'll present you with a few pieces of interesting articles, and will make a case for how best to control software release cycle at the end.

Go That Way, Really Fast (And If Something Gets In Your Way, Turn)

Product Release Lesson From Novice Skier

That's what Jeff Atwood practices in building out Stack Overflow. Instead of trying to figure out whether they are on the right path, they'll incorporate user feedback and implement really fast, then ready to listen to users again.

He draws a lesson from Boyd's Law of Iteration: Speed of iteration beats of quality of iteration. Colonel John Boyd made this observation while studying dogfights between MiG-15s and F-86s fighter jets. John's theory for explaining why F-86 pilot consistently beat MiG-15s pilot was F-86s pilot's faster reaction time thanks to hydraulic flight stick which allowed faster reaction time after repeated maneuvers. Although MiG-15s had faster turn speed, F-86s control allowed the pilot to observe, react, and repeat faster due to easier control.

"Fail Fast", Improve, Try Again

Reid Hoffman, founder of LinkedIn, shared his thoughts about initial product launch at Y Combinator Startup School last weekend. He's a firm believer that "fail fast" mantra. He thinks entrepreneurs should not be embarrassed by the first product launch. In fact he advocates that if you are not embarrassed by the first product launch, you've probably waited too long to launch the product.

His message is to "fail fast", but improve fast and try again. He thinks that grabbing the initial market share before anyone else and rising above the rest of competition is far more valuable then launching perfect initial version after competitors' gain traction by their earlier time-to-market.

We Have Always Focused On The [Big] Market

Don Valentine, founder of Sequoia Capital, did a talk last week about how Sequoia's investment philosophy. He hammers on the point that investment decision at Sequoia is made on one single criteria: Is there a big and happening market?

He contends that getting the bright people and hard work are all necessary ingredients, but they are not the central to whether Sequoia decides to invest or not. It's all about big market where there are lots of buyers who will buy the product that you are making.

"We learned that we have never made a bad investment where technology didn't work. It didn't work when they said it was gonna work... What didn't work was the dynamics of the market. We've developed some pretty spectacular things for which there were no buyers... Critical thing is getting a product developed where timing of product's availability and market development is simultaneous... otherwise you are spending lots of money in developing market where you intended not to spend your money on. Invariably we tend to shut those investments down"

Talk is long. Watch the first 5 minutes and skip to 40 minutes into it.

What do all these articles have in common? It is the market, and how no one person can control the market. Instead we have to carefully observe it, listen to it, and anticipate where the market is going to be by the time when product is launched.

Another point is no one gets the market right all the time. Even Don Valentine admits during his hour long talk that 100 out of 500 or so startups that Sequoia invested over the years had to shut their doors due to Sequoia realizing the changing market dynamics. Sometimes the market evolves and buyers are no longer available, or even worse yet, product may be solving the problem that no one is willing to pay for at the time of launch.

Instead of trying to predict the market correctly and get the right product out on the first try, it makes much more sense to get a product out in the market to see whether there is a real market. If there is no market, you'll at least find out quickly and can determine whether to continue or not, i.e. "fail fast". Otherwise, you'll hear what market has to say about the product. If you hear feedback from users, improve and release the new product fast to get the next feedback.

This seems to be the most logical approach to problem that is full of unknowns and prone to changes. Especially when you are talking about emerging market such as social network.