Sorry GoDaddy, too little, too late. But I’ve been meaning to switch for a while now, thanks for finally prompting me to do it.

I’m moving all my stuff from Slicehost to Linode, which I’ve put off long enough because of the above. Vishal is making me use nginx, which is actually surprisingly straightforward.

I quickly ran into a problem, though. mobilefolk.com is a Rails app, with a bunch of back-end stuff we don’t really use, and a whole bunch of static content dumped in the public folder. For example, http://www.mobilefolk.com/whackakitty/. Problem was, that got routed to a Rails action, which doesn’t exist - while http://www.mobilefolk.com/whackakitty/index.html actually led to the right place. I tried rewriting “/” as “index.html”, but predictably, that made the static content work while breaking all Rails actions.

After some unfocused Google searching, the solution is pretty simple:

# check for index.html for directory index
# if its there on the filesystem then rewite
# the url to add /index.html to the end of it
# and then break to send it to the next config rules.
if (-f $request_filename/index.html) {
rewrite (.*) $1/index.html break;
}

(from http://topfunky.net/svn/shovel/nginx/conf/nginx.conf).

Almost makes me want to learn nginx properly!

Today I pushed out the LyricFind app for the Playbook, and so far so good! Joe Styer of PlayBookDaily.com was kind enough to post a great review:

http://www.playbookdaily.com/2011/08/30/blackberry-playbook-lyricfind-app-now-available-access-to-over-1-5-million-lyrics/

And we got covered on CrackBerry.com, which spawned a whole number of reposts and re-tweets. It really is amazing to see the power of a blog with a large readership:

http://crackberry.com/lyricfind-searchable-database-lyrics-your-blackberry-playbook

And a repost somewhere on the Wall Street Journal website. No idea who reads that… but can we say we got WSJ coverage now?

http://onespot.wsj.com/technology/2011/08/30/1ff96/lyricfind-a-searchable-database-of-lyric

If you have a PlayBook, give it a shot - it’s on AppWorld. It’s a great app!

I jotted down a few quick notes of things I found interesting. I didn’t write down the usual stats slides, I’m assuming those are pretty easy to find online. This is more or less a brain dump.

• Monetization - payment through carriers, including in-app purchase
• 80% of users on OS 5 and 6
• App World - 3 million downloads / day
• “Super Apps” - explained what RIM calls Super Apps, which are apps with deep integration in the system, multiple entry points, used every day, etc.
• Integration with OS - integrate with menu items; send items directly to inbox
• WebWorks - near-parity with Java API.

• blackberry.com/developers/paymentservice
• blackberry.com/developers/adservice (developers get 60% - of what? Is it just an AdWhirl that takes 40%?)

• pushing BBM Social Platform. BBM SDK beta available.
• embed chat with friends in app
• use BBM to invite friends to an app / game (oh god.. Mob Wars?)
• peer-to-peer streaming and content sharing
• add data to user profiles (high scores / achievements)
• blackberry.com/developers/bbm

My first thought when hearing about BBM Social Platform was that it will enable some of the worst behavior seen on FaceBook and iPhone (invite spam, newsfeed spam, etc.) Some of the features do sound pretty cool - in terms of gaming, it sounds a lot like Steam, but it can of course be used for non-gaming apps as well.

PlayBook

• showing off multitasking - turned on front-facing camera, started Need for Speed, has both running on simultaneously on the screen. Pretty cool, I don’t think the audience was as awed as he was hoping.
• more showing off multitasking - playing YouTube video in a browser window, showing other apps on the screen, camera app is still running

From the demo, the PlayBook does look very solid.

Tablet OS

• Built on QNX
• WebKit / HTML5 and BlackBerry WebWorks
• Adobe AIR SDK
• “Later this year”:
• Native C/C++
• Java “App Players” for Android 2.3 and BlackBerry 6

• Alliance Program - consumer / enterprise / managed services
• www.blackberry.com/partners

The Q&A session was quite good, the answers were usually more detailed and engaging than my summary (which is not what Tyler said verbatim).

• Will we be able to run iOS apps?
  “Not that I’m aware of, no. We’re excited about Android.”

• What kind of apps would you not develop in WebWorks?
  “High performance, high-end gaming, media experience. Very specialized look & feel. Heavy amount of local storage. Advantage of WebWorks - target Playbook, OS 5 and 6 in one platform”

• Player app for non-native apps - how seamless?
  “Can’t answer all the details yet, announcements to come. As seamless as possible for the developer - repackage and submit to storefront. Seen alongside Air, Webworks apps - seamless. If user doesn’t have player, they’ll be prompted to download it.”

• Can download files locally, like on PC?
  “Not sure.”
  Can I download and store files as part of app?
  “Absolutely, there’s a private file system.”

• Suggestion: be the first in the world to let developers submit video promos with their apps.
  “Great idea, we’ve talked about it in the past”

• Are you ever going to upgrade your JDKs? Sick of using 1.4.
  “Good question, don’t know”.
  Please know soon! (laughter)
  

• Any limitations developing for Playbook? Can we do everything RIM programmers do?
  “Did a good job opening up smartphone, still some left to open up. On Playbook, version 1, so just started adding extensions - initially there will be restraints, a lot more flexibility once we introduce native C++, and enable higher level platforms to call down to native interfaces / components”

• Payment - Paypal, CC, carrier - are they all 30%?
  “Carrier payments are live with AT&T and Telus in US and Canada, working on more. 70/30 everywhere.”

• BB Bridge - can I use phone browser from playbook?
  “Tethering is part of Bridge.”
  Can we use browser to see cache pushed to BB?
  “No.”
  Not like an external screen?
  “No.”

• Enterprise server?
  “Basic policies - does Playbook need password, etc. Bringing 4G this summer.”
  Can you push to Playbook at some point?
  “I would fully expect so.”

• Is PayPal just for buying apps, or are there APIs?
  “We don’t provide specific PayPal integration APIs.”
  Selling physical goods, PayPal site not mobile friendly -
  “Have to do that yourself today.”

• WebWorks vs Air?
  “Seeing both in apps so far, great community of Flash developers who quickly ported their existing apps, seeing a lot more WebWorks now that web developers had some time to ramp up, seeing a good split. A lot of Flash-based games.”

• Are you going going to give examples of low-level code for web works?
  “We’ll take note of that” (tells other RIM guys to write it down)

• Adaptive bit rate streaming support?
  “Trying to remember… can’t remember if we’re supporting at launch. We support adaptive bit rate, no http-encapsulated(?) protocol”.
  Will newer handsets support it?
  “Not supported in 6.0, no near-term advancements to do it”

• Will devs have access to Bridge API?
  “Initially no, something we’re considering”

• Ebook support - can you elaborate?
  “Kobo eBook reader and store will be preloaded. We expect to see third-party stores and readers.”

• Battery performance - Steve Jobs says don’t support Flash because kills battery. How is it on Playbook, and how does use correlate with battery life?
  “Anything high performance will kill battery. Close partnership with Adobe and QNX, one of the advantages was building Flash at low level in OS. Hardware acceleration for Flash videos, lots of optimizations. We’re not concerned about Flash being a significant hog, we’re happy with the performance. Battery life will be consistent and competitive with other tablets”

• AppWorld - updates tedious, also reboots annoying.
  “Would you like an Update All button? What a great idea!” “It’s quite possible that we might be working on that.” (laughter) “Reboots may still be necessary in some cases.”

• Video calling / chat app in Playbook?
  “We expect to do that, also talking to partners who will do that”

• Remote wipe for Playbook?
  “Not sure. Part of why invested so much in Bridge, because emails, etc. will not be stored on Playbook - so you shouldn’t need to wipe. If apps store sensitive data, they’ll have to use access control. We focused on not needing a wipe early on.”

• Does tethering avoid the need for the extra $10 data share plan?
  “Depends on the carrier.”

First, if you missed it the first time, here’s a neat timelapse video of the game being made:

And a shameless plug for a friend: if you like the artwork in Whack-a-Kitty, and need an illustrator for a project, get in touch with Willis: http://itstheburbs.blogspot.com/

The Numbers

Nothing outstanding - we only ever read about the breakaway successes, but I suspect most game launches are a lot closer to this (or worse). The app is free, so 2000 downloads isn’t anything to get too excited about.

The first spike - on the 21st - was “launch day”, most of the traffic came from Reddit. I was hoping the “making of” video would also get some attention on Hacker News, but it languished with about 7 votes and a few positive comments - bad luck/timing, I guess.

The second spike - 25th to 27th - is the famous Christmas effect. I didn’t do any special promotion - as far as I can tell, that is simply organic growth from increased activity in the app store over Christmas. I’m definitely happy I managed to get the game on the App Store before the holidays.

I don’t have any numbers from other games to compare it with, but I’m happy with the 6 minute average session length. That seems reasonably immersive for an iPhone game.

Some data breaking down user activity:

What went right

The making-of video

I thought the video would be a useful promotional tool, and I was right. It’s hard to stand out from the crowd, the video is a good way to offer something different to people. It got about 4000 views in the first few days, and probably more compliments than the game itself

Endless mode

The game features two play modes, “Story Mode” - finite levels with brief intros that tell a short story - and “Endless Mode”, where you play uninterrupted until you lose. Endless mode was almost an afterthought (I think Issy convinced me to include it), but it seems to work much better than story mode.

In retrospect, that should be obvious: endless mode lasts about 5 minutes, at the end you get immediate feedback in the form of a score, you can compare that with friends, and the feedback loop is short enough to allow for friendly competition on the spot.

The stats seem to confirm this: about 87% of Endless Mode games were finished, while only 27% of Story Mode games were.

This seems like a natural area to emphasize and expand in future versions - it IS fun to play and somewhat addictive, so it makes sense to push it more.

High Score leaderboard

GameCenter support was also almost an afterthought - I mostly wanted to learn how to work with it, and lots of games include it, so it just seemed like the right thing to do.

Since the leaderboard is global, this has spawned some serious competition at the office - two friends have been going head-to-head competing for the top spot for “Endless Mode” for a week now, playing a few games every day and occasionally leapfrogging each other. I wasn’t expecting this level of competition, and it’s really fun to watch.

Integration is still really basic - the leaderboard seems like another area to expand. There is no feedback if you get a new high score after finishing a game, unless you open GameCenter and look it up (which is sometimes very slow). Also, sharing scores on Facebook or Twitter, and the ability to challenge friends, could be good.

What didn’t work

Story Mode

As I mentioned above, only 27% of Story Mode games were completed, and only 8 (that’s roughly 0%) were actually won. That’s terrible. The pacing is too slow, and it takes too long (about 15 minutes) to finish, with no easy way to save progress and resume later. The last level may also be confusing, although I suspect few people even got that far.

Getting reviews

I submitted the game to a few iPhone game review sites, I haven’t heard anything back from any of them. Not too surprising, but disappointing nonetheless. I haven’t even come close from reaching all the review sites yet, so I’ll keep plugging away at this.

“Release Date” traffic / getting into a top 100

I made a big deal of having all my promotional material ready for release - the plan was to do a marketing push (to the extent of my abilities) on the day the game hit the app store. The theory was that I had one shot at combining traffic from my release announcements with natural traffic from being on the “Release Date” list on the iPhone, to get enough downloads to maybe sneak onto the bottom of a top 100 sublist.

Turns out, I shouldn’t have bothered. I released the game on the app store the evening before I made the announcements on Reddit etc. In the approximately 12 hours between, the game got a grand total of… about 20 downloads. So much for the “Release Date” list. The fact that the list is alphabetical, and “Whack-a-Kitty” starts with a W, probably didn’t help. I don’t care if it’s absurd, my next game is starting with “A”.

Understanding traffic

Coming from the web world, this is the hardest to get used to: I have no idea where my downloads are coming from!

I can guess at the reasons for the Christmas spike, but other than that I’m in the dark. Are people finding the game through searches in the app store? If so, for what terms? How many downloads are based on word-of-mouth? Is the leaderboard driving any downloads? Am I still getting residual traffic from the sites where I posted on release day? If so, which ones?

Without knowing any of this, it’s hard to understand where my audience is coming from, and how to reach them better.

Inconclusive / What’s Next

Donations model

I’m trying a slightly different model than most games - rather than charging for the game, or slapping some ads on it, I’m asking for donations (”Buy our eternal gratitude”). Basically, pay what you think the game is worth, by choosing one of several in-app purchases. I got 4 or 5 donations so far, two of them from people I know - but overall traffic is still low, so it’s too early to tell how effective it is. I suspect it will be slightly better than advertising. This also seems like an obvious target for A/B testing, if I ever find enough time.

Promotion

This is my main challenge right now - figuring out how to promote the game. I’m mainly a developer, and I know my stuff when it comes to writing code, but I don’t have much experience with promoting an app.

Whack-a-Kitty has been a learning project from the very beginning, and promotion seems to be the next learning stage - I need to try out different things, and learn what works and what doesn’t. There’s still some purely mechanical work I need to do - keep submitting to review sites, post on iPhone forums, etc. - but beyond that I’ll need to get creative.

My sincere thanks to everyone who has helped me with the game, and especially to Willis Bulliner, who made the gorgeous art used in the game.

For now, get the game from the app store - it’s free!

http://www.mobilefolk.com/whackakitty/

http://itunes.apple.com/us/app/whackakitty/id396023651?mt=8

Check out the video, or download the code from the original post.

http://www.androidguys.com/2009/09/14/40bizmodels/
http://www.androidguys.com/2009/09/15/40-android-business-models-part-two/
http://www.androidguys.com/2009/09/16/40-android-business-models-part-three/
http://www.androidguys.com/2009/09/17/40-android-business-models-part-four/
http://www.androidguys.com/2009/09/18/40-android-business-models-part-five/

http://search.twitter.com/search?q=ipad+boston

And, of course, the infinitely more predictable iTampon:

Note that iTampon is trending on Twitter, while iPad is not. Although, to be fair, only roughly 1 in 10 tweets (over a highly scientific 5-minute real time search sample) mention iTampon.

/amused

http://www.mobileorchard.com/developer-to-developer-iphone-app-distribution-without-ad-hoc-provisioning/