[I went up to Winnipeg a few weeks ago to give a talk about iPhone game design and marketing.  It ended up being a 6 hour long affair where I covered pretty much everything I know about anything.  These are my notes from that talk.]

DISCLAIMER: While everything I included in the talk has actually worked, in practice, for a real company, that does not mean it will work for you!  I'm proud to have advice that I can share that is not purely theoretical, but in my experience each game company is a really different place, and exceptions abound.  Proceed not with caution but with logic and critical thought!

Finally, at least as a blog writeup, I won't be expanding on these topics in text anytime during the near future.  What follows are simply the notes I took while structuring and designing the talk, and I included handwritten notes I made just prior to actually giving the presentation.  The presentation was somewhere around SIX HOURS LONG, which would make this article insanely lengthy were it to include full treatises on each topic.  What I've included here are simply the meatiest, most important nodes, organized chronologically from before you are even an iPhone developer leading up to sustaining your small but successful company.

BECOMING AN APPROVED APPLE IDEVICE DEVELOPER

> regular access to an Intel-based machine running OSX 10.5 (Leopard)
    - can totally be a "hackintosh"
> register as an iPhone developer
    - free!  this is how first version of Wurdle was created
    - http://developer.apple.com/iphone/program/start/register/
    - grants access to iPhone SDK and the simulator
> join the iPhone Developer Program
    - $99 (i think this is annually - if this is a problem, quit now!!)
    - http://developer.apple.com/iphone/program/
    - gives access to the Development Certificates
        > necessary to sign an app and get it running on an actual device!
> IF YOU WANT TO SELL APPS, open an account at a big-name international bank
    - need a SWIFT code for Apple to wire you the money
    - give your account info to Apple
    - verification can take about a week
        > do it ASAP if you plan to sell stuff
        > stressing over bank paperwork during launch != fun
    - this can be done later if you want to start with free apps
> create a Development Provisioning Profile
    - access from iPhone Developer Program Portal
    - allows you to sign apps that will run on development devices only
    - must provide device IDs of all devices.
    - all this certificate/provisioning stuff is a huge pain in the butt.
> eventually you even get to make games!
    - XCode is where you write the codes
    - Objective C, C++ are the codes
    - Apple UI API, OpenGL are the libraries
    - Unity & GameSalad can help if that's too intimidating!
    - PNG files for graphics, m4a for sound


DESIGNING GOOD MAINSTREAM GAMES


IDEVICE CONSTRAINTS AND OPPORTUNITIES

> iDevice speedbumps
    - DOES NOT RUN OSX
    - slow file system access
        > lots of wurdle's development was dedicated to dictionary load time
    - touch data frequently unreliable
        > games that require a lot of CPU can trash touch events
        > great games (and apps like Safari) sometimes lose touch-end events
    - simulator & device performance varies
        > real dev must be done ON DEVICE
        > simulator is still useful for quickly checking
            logical functionality or asset positioning
    - small screen + fat fingers + dense LCD = tricky
        > images look smaller than on your monitor
        > "fingertip" is roughly 40px x 40px (official apple guideline)
    - hard to get game-console-like performance
        > not impossible by any means, but not going to happen
            without a lot of cleverly optimized code in most cases
> iDevice opportunities
    - lots of popular input mechanisms
        > big button, fake controller, mouse, tilt, local touch moves, etc
        > WARNING: some of these require a LOT of fine-tuning to feel good
            - spent weeks (!!) on the control scheme for our racing game
    - huge audience of gamers and non-gamers
        > if same percentage of smart people in both groups, and there
            are more non-gamers, then i want non-gamers playing my game!
        > actually, i want EVERYONE playing it (thus the accessibility concerns)
    - apple provides UI layer that can go over your OpenGL stuff
        > used for wurdle, canabalt dashboard, etc
    - built-in support for two player bluetooth gaming
        > but NOT for 4 players as Apple claimed
    - network connectivity and web browser widget
        > makes web-based high score display cheap and easy!
> project scope
    - goal: smaller, better games
    - iDevice playtimes tend toward smaller & bite-size anyways
    - procedural generation is good option for a lot of reasons
        > see upcoming issue of Game Developer for more info!


SOUND AND ATMOSPHERE

> this is a game designer's secret weapon!
    - people are VERY sensitive to sound
    - has everybody seen the mary poppins horror trailer?
> SOUND BASICS
    - music
    - ambience
    - effects
    - chiptunes
    - editors
    - synths
    - trackers
    - sequencers
> music
    - collaborating with musicians
    - free retro music tools: famitracker, musagi
> sound effects
    - collaborating with engineers
    - SFXR: indispensable retro effect synth
    - audacity: decent free editing tool
        - can distort and clean up sounds, very helpful
        - can tweak stereo positioning too
    - downloading and re-engineering free fx from the web
        > this takes a surprisingly long time
    - crappy home foley!
    - randomize when possible (e.g. randomly choose from 4 diff footsteps)
> cannot stress how important it is to ship with the RIGHT sounds


MARKETING YOURSELF AND YOUR GAME

> basics
    - these are things you should do/care about no matter what:
    - GOOGLE ANALYTICS
    - website (ideally with playable demo in flash/unity or HD video)
    - aggregators
        > no control, but you can make it easier for people
        > stumbleupon.com -> adamatomic.com = 190,000 hits
    - twitter
        > they provide a simple http-post based API
        > also provide a big 'real' API
        > twitter audience tend to be mavens
        > adamatomic.com 13,000 hits
        > canabalt.com 41,000 hits
        > syndicates to facebook for extra 50% hits
    - youtube channel, blogs, etc
        > these take up real time/effort but are still 'affordable'
        > blog can even be on someone else's site (i blog on gamasutra)
        > Wolfire are kind of the experts at this stuff
            - they are selling a game that isn't even done yet!
> publisher/PR-oriented
    - sign a deal with either a PR firm or a publisher
        > e.g. Triple Point, Chillingo, NOT NGMoco :(, NOT Publisher X
    - they take a combination of up front money and royalties usually
        > $3000 to $30,000, 5% to 30%
    - they have hookups at Apple, advertising deals with PocketGamer & TA
> reputation-oriented
    - building up your rep can take years
    - most canabalt traffic goes through my own website (10 to 1)
    - increase via blogging, public speaking, free games, free tools
        > Flashbang is awesome at this!
    - devote serious time to "networking" with other devs AND journalists
        > NOTE: if you approach it like speed-dating you will fail!!
        > "networking" is douche-speak for "making friends"
        > game developers are SUPER interesting people
            - a lot of journalists are pretty rad too!
        > i only do this casually and am on a first-name basis
            with at 150+ industry people


SELF-PUBLISHING TO THE APP STORE

> pick a launch date (friday after 10 days after submission)
> update your twitter
    - even ask people to follow you in case there are submission issues
> run a mailing list
    - 3,000 clickthroughs from 30,000 emails from 300,000 customers
> get advance 'ad hoc' copies to journalist friends
    - GRAPHICS MATTER, most blogs feature screenshots only still
    - same with the app store for that matter
> cold emails DO NOT WORK EVER
> getting app store visibility without having editor hookups
    - TWO LISTS
        > editor lists (hot new games, what we're playing, etc)
        > sales lists
            - top downloads
            - top revenue
    - need to sell I think 500+ copies to make it on the top 100 genre lists
    - need to sell probably 1000+ copies to make it on top 100 games
    - this is why all this launch stuff matters so much!
        > If you get 1000 purchases on launch day,
            then you are APP STORE VISIBLE without editor help!
        > this also improves chances that you will get noticed
    - this is not as easy as it sounds but totally possible
> pricing (GASP)
    - must be a dollar to make it in the top 10
    - must be more than a dollar to sustain a company outside the top 10
    - may notice difference in reviews for $1 versus more


SMALL BUSINESS SURVIVAL BASICS

> what is the point of starting your own company?
    - legal protection
        > ostensibly.
    - doing things your way for once, taking risks, betting on yourself
    - control over who you work with (more on this later)
    - financial success that enables your extracurricular activities
        > subsidize your creative passions:
            - making games too weird to sell
            - cooking, making music, traveling, etc
        > this is relatively impossible at a big company
            - except maybe universities where you get summers/sabbaticals
> commercial responsibility
    - sorry but I am going to harp on this!!
    - your company is not a way to legitimize the things you love
        > this will spoil your loves and stress you the F out
    - work on projects that you think will sell
        > this doesn't mean you can't take risks
            - quite the opposite actually
            - this is why you have your own company after all
        > it DOES mean you need to honestly assess your ideas
    - prioritize your projects based on budgets
        > more on this later!
> paperwork
    - taxes, paying contractors, filing whatevers, updating crap
    - this is not trivial, chews up at least 100 hours / year
> collaboration
    - a big part of running your own company is choosing who you work with
    - opportunities to hire and fire people; godlike powers O_O
    - best advice: only work with people who are better than you.
    - seek especially people who complement your weaknesses
    - work hard to increase your own skills so that collaboration is a CHOICE
    - if you ARE hiring someone:
        > pay them well
        > pay them upfront
        > give them royalties
> overhead
    - small company can get by on small audience
    - do you need an office?
        > can help with communication, time management
        > lot of overhead, same cost as 1 or 2 employees!
        > if you DON'T need an office, consider timeboxing or SOMETHING
    - do you need a server?
        > yes.
    - do you need insanely high speed internet for $1000/yr?
        > mmmmmm probably not
    - do you need a $4000 Mac workstation, or a $1000 hackintosh?
> philosophy
    - realism and objectivity are king!  no place for optimism OR cynicism
    - everything in business can be divided into two groups:
        > things you have control over
            - quality
            - schedule
            - budget
        > things you don't have control over
            - appeal
            - sales
            - visibility
            - contests/competitions
            - equipment functionality
    - i find it is safer and more useful to plan for the worst case
        > if you don't have control over it, assume it will fail
        > THIS IS NOT CYNICAL, this is USEFUL
    - hedge your bets by taking the things you CAN influence VERY seriously
        > pick the right project
        > make a great game
        > get it done on time
        > don't run out of money
> budgeting tips
    - MAN-MONTHS: this is basically a made up unit
        > ostensibly it refers to the cost of one person working for one month
        > this varies wildly from company to company obviously
        > it is more important to understand how much real work is done
            in a man-month than how much a man-month actually costs (at first)
    - budgeting is more about precision than accuracy
    - don't worry about calculating exact dollar expenses for now
    - pretend you have 5 game ideas
        > 4 of them need 6 man-months
        > 1 only needs 2 man-months
        > MAKE THAT ONE FIRST
    - pretty much that's it.
    - eventually you might need to know dollar amounts
        > e.g. if you don't want to run out of money on accident
    - man-months are tricky units!
        > need accurate (i.e. impossible) estimates of production
        > need to incorporate the aforementioned overhead plus taxes etc
        > even on a 1 person game this can get annoyingly out of hand
            - e.g. you are working on a simple game that will take 1 month
            - say you spend $2000/mo on rent, food, video games, and liquor
            - first, this is more like $3000 before taxes
            - do you need any new equipment?  are you paying for music?
            - what if you get sick for a week?
> licensing basics
    - last but not least!
    - licensing is a great way to make money off a project after it's done
    - involves trading the rights to your game to another company
        > usually for combination of up-front and royalties
    - everyone wants to make money off of your popular fun game!
        > this is fine, if YOU make enough to help stabilize your operation
    - everyone wants to make money and NOT SHARE IT EVER
        > THIS INCLUDES YOUR LAWYER BTW
        > things to look out for in licensing contracts
            without buying your lawyer a new car:
            - no limit on sequels/"deluxe" versions
            - clear, exclusive list of platforms
            - length of contract term
            - right to final approval of game and marketing materials
            - your company name appears in game and in ads
            - undefined terms used loosely (e.g. "exclusivity")
            - terms of license fee calculation (net vs gross, etc)
        > YOU HAVE TO LOOK FOR ALL THIS EVERY TIME
            - Even if my dad runs the other company?
                > YES.
                > Sometimes these contracts have 3 or 4 year terms
                > Anything can happen in 3 years.  I started 2 companies!
                > What if somebody buys your dad's company?
        > if the licensor tries to guilt trip you or placate you,
            tell them to eat it and find somebody else to exploit.
            - contracts must be fair to BOTH parties