This multi-part series examines the history of Knockabout Games, a mobile games startup I co-founded in 2002, near the start of the pre-iPhone “first wave of mobile gaming”.  The start of the series can be found here, and last week’s post on how Knockabout’s plan played out can be found here.

So some of the things we did worked well.  It didn’t play out perfectly though.  A number of problems presented themselves:

• Our costs rose with each round of titles too.  Not as much as the funding, but it was a couple years before we finally closed the gap.

• We chose to emphasize quality over everything else.  We had a limited number of resources, so that meant delivery date was not #1 (we would rather slip than be less than great).  I had a theory that, assuming you have the talent to deliver, the difference between a good title and great title in the traditional game space was perhaps 2x in development cost.  But it was 5x in sales, and those sales would cure all sins.   Except in mobile it wasn’t 5x, it was more like 1.2x, which doesn’t justify the development math.  We were making licensed sports titles for the #1 publisher in the space and garnering the top review scores.  It didn’t matter.  Consumers had trouble differentiating titles prior to purchase (14 characters for your game’s title, and no screenshot or other information), and didn’t seem to value them very differently beyond that (not enough for word of mouth to take off).  The COO for one of the major mobile publishers told me flat out that he’d rather have two mediocre titles for the same budget as one great title.

• We were slow.  This turned out to be Knockabout’s achilles heel.  Yes, we sacrificed time for quality.  But it went beyond that.  We couldn’t hit a date to save our lives.  I don’t have a good explanation for that.  In my first company we had taken people who had never shipped before and got them to hit every date, so part of me thinks I took that for granted — after all, our team had a ton of experience shipping product.  Or maybe our choice to go with more junior managers was wrong.   Those all feel like thin excuses though.

• Platform support, over time, became the critical skill to have as a developer.  We were building reference builds — i.e. a version of the game on 4 or 5 handsets that represented a sample of the possible memory and screen configurations.  The publisher would then send those off to a shop in India or Russia to have them ported to 200 handsets at $300 a pop (these days it’s more like 5000 handsets).  We couldn’t compete with that porting price.  But the developers who won were the ones who built a framework that could spit out ports at a push of a button.  The publisher would pay them the porting money, and they’d have no additional cost.  Of course, they had to limit their products somewhat, or choose ones that didn’t push the envelope like ours.   But by the time we realized we had to take this approach it was too late.

• We eventually hit a development budget cap.  We were getting the top rate in the space, but the return wasn’t there and publishers stopped increasing their per title development budgets.

All of this led to some difficult decisions.  We were trying to split our time between our own games and those for our publishers (our own titles were self-funded, we always dedicated resources to our publishers for the duration of our deals).  But as deadlines slipped, we applied more of our resources to our contractual commitments, and eventually had to can everything but our pinball game.  Bye bye breakout and blackjack.

And since we were bootstrapped, we had no capital to address the porting framework problem or to bring in more resources.  We were living a typical hand to mouth existence, with payroll dependent on hitting every milestone.  There was no buffer, and at 16 people, we couldn’t personally cover the difference.

I have quite a bit of game design experience, and that’s the role a played for Knockabout.  But I was also the CEO, which meant I was responsible for various business development and operational functions, plus managing the guys managing the products.  In addition to not having the time to actually do all that, it created a really awkward situation for the producers:  I was their boss, but I was also the guy on their team they needed design work from.  I had a good rapport with individual developers, but it undermined the managers.

Hiring junior managers was probably a poor choice too.  We saved money, and the guys were talented.  But I never spent the time to help them grow into their roles, throwing them in the fire and devoting cycles to business and design issues.

Our assumptions about distribution were also a little off.  While many of the carriers were encouraging developers to go through publishers, Verizon asked us point blank why we were using JAMDAT and not going to them directly.  In the long run I think that would have been hard to sustain (and we certainly couldn’t have gone direct to every carrier), but we might have been able to capture more short term revenue.

There were some unexpected staffing complications as well.  We gave a small signing bonus to one engineer we hired;  he took it, quit a month later and refused to repay it.  Another engineer came to work for us and quit after two weeks because a competing company made a better offer.  That’s random stuff that just happens:  both guys came in on personal recommendations from people already working for us.  In a startup you always have to expect the unexpected.

Next time:  Knockabout’s demise.

This multi-part series examines the history of Knockabout Games, a mobile games startup I co-founded in 2002, near the start of the pre-iPhone “first wave of mobile gaming”.  The start of the series can be found here, and last week’s post on Knockabout’s equity and staffing plan can be found here.

Sure enough, our method of meeting publishers long before pitching worked well to establish our presence in the space.  By the six month mark, if I was introduced to a new publisher, they had almost always heard of Knockabout in advance.  We hadn’t even shipped a title yet.

Early on we signed up with JAMDAT to build the next version of their solitaire game.  Their first one was a great seller, so despite the lack of a brand, it seemed like good bet (and in the early days of mobile, just having a simple name for a title — e.g. solitaire or racing — was often more effective than long brand names that didn’t fit well on the deck listing).  A few months after that they gave us their NBA title to build.  In both cases, we overspent relative to what they funded.

We eventually signed on with a few other publishers, mostly to build games based on movies or sports.  And we added JAMDAT’s NFL property to our list.

For our own IP, we built three titles:  pinball, breakout, and blackjack.  We received almost a dozen offers to distribute the pinball game, and ultimately signed with JAMDAT.   Not only did they offer the best deal, but it made sense given how much work we were doing for them on their major brands.

We also signed a royalty-based OEM deal with Motorola to put our pinball game on a number of their handsets.

We constantly kept in touch with publishers and handset providers we weren’t working with, always looking for worthwhile deals.  Our steady workflow, particularly with high profile publishers like JAMDAT, gave us walk away power in contract negotiations.  We even dropped a potential deal with THQ in the late stages of negotiations because they refused to discuss a few critical parts of their contract (if anyone ever tells you “that’s just the way it is” in a game development contract, you should tell them “just where it is” they should shove it).

As expected, we overspent on our titles.  We got the results we expected too:   the games were very high quality, consistently rated #1 by reviewers in their respective categories;  and each subsequent deal we did was higher than the previous, often doubling in size.

We managed our publisher relationships very well too.  When things did go wrong, they were supportive and worked with us to find a solution.  In one instance, where we knew we were going to miss a movie launch date, the publisher pulled the title and sent the assets off to a chop shop in Russia to get a quick product out the door.  But they also immediately called us to try and lock us down for another title before we went to another publisher, going so far as to send us a milestone check before we’d even determined what the game was or had a contract in place.

Most of our staff were software engineers.  We had one full time artist who was crazy fast, and the platform limitations put a severe limit on how many assets we could provide (max download sizes ranged from 100k – 400k).

It didn’t play out perfectly though, a topic I’ll cover next time.

This multi-part series examines the history of Knockabout Games, a mobile games startup I co-founded in 2002, near the start of the pre-iPhone “first wave of mobile gaming”.  The start of the series can be found here, and last week’s post on Knockabout’s business strategy can be found here.

Funding and Equity

Knockabout was mostly owned by myself and Monty Kerr.  Monty’s role was largely silent — he had an existing game development business to run still — but he had quite a bit of extra office space, desks and computers for Knockabout to borrow.  He also had spare resources (people) for us to use on a part-time basis, to complement the rest of us who weren’t going to draw any salaries for a while.   Since game development is primarily a labor intensive activity — almost all the expense is tied up in payroll and the space to house everyone — this took care of the bulk of our costs and allowed us to bootstrap the business (we later leased our own space and in an unusual reversal, sublet to Monty’s company).

We also gave a small amount of equity to three seasoned game industry veterans we thought would be critical to our company:  a director level manager to oversee the specific titles, a technology director, and the CTO of Monty’s current business (who was to stay there and migrate to Knockabout later).  We wanted these folks to feel and behave like owners, not employees, since they were so important to our success.

And we gave them ownership straight up — no options, no vesting.  We did have an option plan drafted, but never actually used it.  Our take was that options, in 2002, still had a negative connotation left over from the dotcom bubble.  There was no point in handing them out if no one would take them seriously.  So for most of the staff we had a simple, and generous, bonus plan:  any time profit distributions were made, 50% of it had to go to the staff.

Development Staff

For staff, we only wanted to hire coders and artists with prior game development experience.  The platform was going to be a nightmare as it was — I didn’t want people learning how to make games for the first time as well.  I just wanted experienced game developers who were good engineers (or artists who had strong illustration, modeling and animation skills) — that was the harder thing to find;  any good engineer could learn a new platform easily.

On the flip side, we skimped on management.  I wanted producers who were former developers so they could communicate well with the team.  That struck me as particularly important since the teams were so small (often one engineer).   So we would hire guys who knew how to make games, but were trying to transition from a technical role into a management one.  I was to personally train and mentor them in those roles.

That wraps up the planning phase for Knockabout.  Next time I’ll talk about what actually happened.

This multi-part series examines the history of Knockabout Games, a mobile games startup I co-founded in 2002, near the start of the pre-iPhone “first wave of mobile gaming”.  The start of the series can be found here, and last week’s post on Knockabout’s product strategy can be found here.

Business Strategy

Building relationships were important for our long term exit strategy too.  By becoming indispensable to publishers, we hoped to create intangible value on top of explicit revenue generation.  And in circumstances where revenues aren’t as high as anticipated, having added value in other areas can often trigger a very profitable acquisition (my first company sold this way).

To get those relationships going, we had a three step approach:

• Step one:  introduce ourselves.  Don’t be arrogant game developers coming to teach mobile how real games are being made.   Be humble and acknowledge that while we know games, we really don’t know mobile and would like to learn.  Spend the first month or so meeting with publishers and asking them questions.  Let them bestow their wisdom upon us (people love to talk about themselves, or in this case, about their companies).  Don’t pitch anything, but leave them with a business card.  The goal was to create familiarity first;  we’d come back later and pitch.
• Step two:  show some demos.  After a couple months, we’d come back and show some early demos.  We would seek distribution relationships, but would also inquire about high profile brands they needed addressed.
• Step three:  close a couple deals over the following three months, both for distribution of our titles and third party development of theirs.  Make sure to diversify our risk by working with multiple publishers.

And of course, step four would be to then deliver the highest quality game we could and up the ante on the next round of development and distribution deals.  And step five would be an acquisition by one of our publishing partners.

I should also note that we didn’t cold call anyone.  For phase one, our plan was to always get an introduction from someone into the company (if we didn’t know someone already there).  Whatever your role in the game industry, after a few years you end up either knowing someone at every company, or being one degree removed (friend of a friend).  Like most industries, we work in a small and nepotistic space.

We also thought it was important to talk to more than publishers.  We spent a lot of time with handset makers and platform technologies (like Qualcomm), building those relationships.  They were great at making further introductions to publishers and other business opportunities.  And we talked to developers.  They’re a good source of information, and by sharing, we could all raise the quality of the deals that were out there.  The traditional game space has always felt a bit like a community — open and helpful — and we thought mobile would benefit from the same approach if we could nudge it in that direction.

Next time:  Funding, equity and staffing.

This multi-part series examines the history of Knockabout Games, a mobile games startup I co-founded in 2002, near the start of the pre-iPhone “first wave of mobile gaming”.  The first part can be found here.

Product Strategy

The common perception of mobile phones at the time was that they were essentially 15 year old computers.  This led to such nonsense as porting old titles to the device and severely underestimating hardware capabilities.

The reality was very different.  Yes, the processor in the mobile device was very much like a computer from 15 years earlier (although the range was astounding — some devices were orders of magnitude more powerful).  But the screens, while not large in picture resolution, were 12 to 16 bit in color depth and the pixel pitch was a good as any modern LCD.  The keypad, with its 100ms – 250ms latency was like something out of the seventies (i.e. press a button and it would be 250ms before your software would get the notification).   All of this meant developing games that understood this unique combination of hardware.

Misconception of hardware didn’t fully explain why mobile games were sheer crap though.  Despite the odd platform, it was actually capable of far more than what was on the market.  What we observed was that publishers and developers were unwilling to spend the money to fully take advantage of the device.   To some extent this makes sense:  the space was new and no one was quite sure where it was headed.  But it was also a very narrow view, examining current sales to determine funding for tomorrow’s products.  In a brand new space growth is rapid, often double digits on a monthly basis.  Even with two to three month development cycles, this is a short sighted strategy if you want to remain competitive.

We also noticed that the value chain was really scattershot.  The carriers were the gatekeepers, but all kinds of companies were participating in multiple layers, combining development with self publishing, or content creation with handset manufacture.  There was also a crazy matrix of devices and carriers, hundreds of each and an absurd number of combinations.

We decided to look ahead.  We thought this would all shake out, with different parties dominating different parts of the value chain, and a structure not too dissimilar from the traditional game space (substitute carrier for retailer, pre-iPhone, and that’s pretty much what happened) .  So we decided to focus on one piece — content development — and be better than anyone else at it in anticipation of what was to come.

We knew it would be important to own our IP, but we also knew we’d need relationships with strong publishers (who would have the distribution relationships at a time when app stores were controlled by carriers).  In order to insure they’d give our self-funded products the right amount of attention, we’d need to build products for them based on their IP (this would also help raise our profile as a developer).  In theory, if we did good work and they wanted to keep working with us, they’d want to keep us happy by being fair with our own titles.

That was the other part of our product strategy:  build the best possible product for the device, not the budget.  Even if we were building a game to be owned by a publisher, we’d spend some of our own money to make up the shortfall of what they were paying us.  Why?  We wanted to hook them on the quality.  Product cycles were short — we could eat some of the cost and raise our rates with each subsequent iteration.

So we decided to spend half our cycles on self-funded IP we would own, and the other half on publisher brands that were funded in the typical game industry publisher-developer model (advances against royalties that would never earn out).  We decided that for publisher IP, we’d only work on very high profile or high selling brands (preferably both).

Next time:  the business strategy.