Wednesday, March 16, 2011

Gameday Tycoon: Lean Startup Challenge

At my social sports game startup, Gameday Tycoon, we make damn sure no time is wasted "spinning our wheels."

The truth is that initial customer development is difficult in games. You can perform full blown studies to identify a market that loves your game idea, but the end product can still be a total flop. The reason? Games are more than just ideas, they’re experiences.

So can a game startup be lean? Is it possible? Yes and yes. A game startup can start off by identifying the core experience, building a minimum desirable product (see Andrew Chen's ebook), testing it with users, etc... Lean is just way of thinking and a set of tools to help you act lean.

When friends and family ask what I've been doing since Gameday Tycoon released a few months ago, my answer is always "making it better." Usually I need to further explain what this means,
let me give you an example from last week and show you how we're "lean":

At this stage of Gameday Tycoon, we only care about one thing -
engagement. Engagement is why users pay, invite friends, revisit your app, send nasty emails about bugs-to-be-fixed, and many more.

But let’s step back a little bit first. Before a user becomes engaged, he/she first needs to be initiated. The reason initiation is important is because users need to be setup for engagement. A good initiation experience converts a first timer into an engaged user (a repeat user).

In Gameday Tycoon, we consider ‘initiation’ choosing an assistant and completing step 1 & 2 of our mini-tutorial. It takes around 5-10 minutes to complete and is essentially a guide thru the core game experience:

    1. Choose an Assistant
    2. Mini-tutorial 1
        - Pick on a live game
        - Collect Cash
    3. Mini-tutorial 2
        - Purchase a team
        - Send front row tickets to friends


Let's track our users to see whether they convert thru our 'initiation' funnel in Mixpanel:


WOWZA! 2 of 117 people ‘initiated’ on March 8th! Now that is simply horrendous conversion… simply unacceptable!

So what should I do?
I can inspect the first-time game experience myself and pick out all the things “I think” users are getting stuck at, OR go out and talk to real users to find out how I wasted 5 minutes from each of those 115 user's life. I hope you can guess what I chose to do :)

I went out and got 50+ students on MIT campus to demo my app. Want proof?

                                             

Okay, so lots of things went wrong. Here are the major problems (and what we did to fix them):
  • Users didn’t feel obligated to finish the tutorial (award them with bonus incentives)
  • Users didn’t know where to click (guide them with big green arrows)
  • Users were shown sports leagues they didn’t care about (ask for their favorite league up front and personalize the experience from there)
  • Users said the screen was cluttered (stage new features as they make progress)
  • Users not interested in sports (ignore them, I can’t make everyone happy)

Was it a lot of work? Absolutely. Would it have been more time and energy wasted guessing what the problems were? You bet.

Now let’s check out my ‘initiation’ funnel on March 15th…

Not too shabby huh? Mini-tutorial 1 conversion increased from 29.06% to 57.14%. Mini-tutorial 2 conversion increased from 1.71% to 33.33%.


Geckoboard shot:


Moral of today’s story:
  • Set goals for your users
  • Create a funnel with these goals
  • Track the funnel with numbers
  • Improve numbers by talking to users
  • Rinse and repeat until you achieve success

Was the story exaggerated? A little bit.
Is it real? As real as the startup.
Are we making money? Yup.
Are we looking for mentors? Yes.

P.S Thank you Noah @Appsumo for this opportunity. Check out his site for great deals on web apps.

2 comments:

Brad Mills said...

Erik, I love what you're doing. I'm a fellow contestant in the Lean Startup Challenge with my Facebook related product CoinWhale.com

What you're going through reminds me a lot about what we did with our Facebook games re: user testing. Check out bradmills.tumblr.com to read my entry.

Basically, we learned the lesson after spending 6 figures and wasting 12 months of arguing over assumptions instead of testing. So you're definitely on a great start doing it lean from day 1.

Questions for you:

1) How long did it take you to build this game?
2) Do you use Kontagent for analytics? (you seem very data driven)
3) Do you want to get paid to get more paying users for Gameday Tycoon? Email me at brad@coinwhale.com to talk about this!

Some beta testing notes for you:

I just installed the game and played it, and I didn't actually realize I was in a tutorial, not sure if that's good or bad, I didn't know I had to 'complete' the tutorial...but the flow went really well except:

Asked to pick assistant
Asked to pick my league (NHL)
Asked to place some bets.
Ran out of cash and it said I could refer 2 friends to get more, so I clicked that button and this is what I was presented with:

http://lts.cr/T5v (something's up with our invite friends box on that step.)

anywho I close the box and continue, did the imaginary collections, and then I'm stuck at this step:

http://lts.cr/T5x

Mark said...

Would you mind sharing a few pointers? I have very limited experience with flash games. In a few small projects, I've gone with Kongregate and Mochi, either of which will send some traffic just for a game being new. Unfortunately, if the initial rating is poor, the game just disappears. In the case of my most recent game, I improved it dramatically over a dozen patches, but over 95% of all the traffic it ever saw was in the first couple of days (version 1.00)
What's your general workflow for releasing games? What sites to you start with and where do you put them after making improvements?