by: Neil Harris, President & COO of HeroEngine

What is the Price of Failure?

In the James Bond movies, if you work for SPECTRE and fail, the boss pushes the button and you fall into the shark tank or get electrocuted.  The price of failure is death.

I once worked for a very large company that shared that culture.  Everything you did had to be perfect, or else; because after all, we were a very large corporation and we had to protect the corporate image.

I came into the company from a direct marketing background.  We built failure into our marketing methodology.  For example, if you were selling magazine subscriptions you would send out 80% of your mail with a package where you knew exactly what kind of response rate to expect, then you would print 4 batches to 5% each.  The extra batches might be as simple as changing the color of the headline or as radical as a completely new package.  If one of the tests worked better than the main mailer then next time that would become the main mailer.

When I explained this to my finance guy he was horrified.  From his (and the company’s) perspective, what I was doing was spending money on things that might fail.

Failure was not an option.

Fail Small, Win Big

For me, though, failure was a requirement.  If I wasn’t failing then I wasn’t learning.  Failing wasn’t the goal, of course, but it was a side effect of the learning process.  If James Bond didn’t succeed he tried something else until he found something that worked.

You want to fail at a small scale so when you find something that works you can scale it up for a big win.  Marketing guru Seth Godin (who by the way, in his earlier days, founded online games company Yoyodyne), echoed a similar sentiment about failure in a recent blog post. Godin advocates that failing early is crucial and it is these “initial interactions with the market” and your “ability and willingness to appear stupid in front of others” that lead to the big wins.

The company I worked for spent their time doing research, conducting surveys, hiring consultants, the usual corporate methods.  The time spent building and measuring test packages turns out to be much faster and cheaper.  But in that organization it was not culturally acceptable.  It was risky.  Risk was evil because risk implies the possibility of failure.

Which is why I’m not there any more… and why that division is not there anymore either, by the way.

For Online Games Companies, The “Try-Fail-Iterate” Model  is Crucial

In an internet company testing is fast and cheap.  You need to track what you’re doing and be prepared to try new ideas and variants of old ideas.  You can try new things every day and within a short time have enough data to improve your responses.

In an online game company you have every opportunity to optimize your game by testing.  What should you charge for a virtual item?  Zynga likes A/B tests where they try different prices to see which gives the best return.  Build those tests into your system, don’t assume you know what something is worth to a player.  Don’t just playtest – test everything you can think of.

There is a reason why James Bond beats SPECTRE, after all.

About the author: Neil brings more than 30 years of gaming, software and technology management experience in his current role as COO of Idea Fabrik and President of HeroEngine. Prior to Idea Fabrik, Neil held executive and management positions with Simutronics, GE, Atari and Commodore.  Neil has a BS in Business & Management from the University of Maryland and an MBA from Wharton’s University of Pennsylvania.