Postmortem: Inside Adept's Trixel For iPhone

[In an illuminating postmortem, Adept Games designer Daniel Boutros recounts several important lessons learned from the development of the studio's debut iPhone game Trixel, explaining his experience and suggesting tips for increasing your success in the App Store.] In the last few months I, like many other iPhone developers, have obsessed over the iPhone charts and how they work, from a "gameable" perspective and an audience perspective -- what does the audience respond to? What organically climbs without marketing support or built-in license or franchise awareness? At Adept Games, we developed an original logic puzzle game called Trixel. Our challenge was to develop a solid game in two months and achieve success without any money spent on ads or marketing. All PR was based on me phoning and emailing friends and press. That is all we've relied on for the success we've achieved thus far. Those goals were typical of an enthused indie team on its first project. Our targets were to achieve the following: 1. Quality: Make an accessible yet deep game, that plays very well and is polished. Does the market mostly respond to shallow games, or were shallow games just the only ones available? We figured we'd throw a game with depth in the market and see if that changed anything. 2. Accessibility: Make sure the concept is easy to grasp, visually and interactively. To cover ourselves, and to welcome a greater audience, we would focus on a simple premise and control system, with depth primarily driven by the level designs. 3. Prettiness: People like pretty shiny things. Popcap has been great at designing pretty shiny things that are appealing to many. Peggle is a great example of this. PopCap also expresses the shiny element through its sound design. We would go with a shiny Web 2.0 look, as it was too short a dev cycle to experiment with something truly unique that would also appear polished, and we couldn't secure an in-house artist until mid-way through development. Experimentation on the art side was evaluated as being too risky. 4. Showmanship: People like to show off and compare sizes, so to speak; give them something so they can do that. We would have a performance-focused scored mode called "Race the Clock." A move counter would challenge perfectionists and a timer would be for ninja-finger-tappers. A score table, of course, was a given. 5. Value: We figured 50 levels would sound great and give great value, provided we could release them as well-tuned and as polished as possible. Eran and Paul from Adept set the software up so I could easily add, play with, and tune level content. What Went Right, And What Went Wrong? 1. Quality: Make an accessible yet deep game, that plays very well and is polished. Right: Critics responded very well. User reviews averaged at 4.5 until we dropped the price and got knocked to 3.5, thanks to the dreaded one-star reviews from the software delete prompt. We felt we nailed making a solid game and our feelings were confirmed by great critic reviews. Wrong: Due to a short timeline, we had limited rounds for iteration. Our first prototype was bland as hell, but our programmer Eran recognized this before he presented the running code, and introduced some super tile concepts such as the diagonal tile and the warp tile. We began riffing to get to the final selection we have, although though this was at odds with keeping the UI streamlined, and therefore we may have created a higher barrier of entry by default. We had to be careful when using screenshots of the hard levels, since they could scare users off. 2. Accessibility: Make sure the concept is easy to grasp, visually and interactively. Right: The concept was simple, and after settling with a final tutorial method after a few tries, we got mostly strong feedback that it worked: Flip tiles, then tap a tile adjecnt to you. The UI design got part of that right. Wrong: Due to lack of iteration time, we got the UI looking clean, but it was not entirely representative of the game concept. By looking at the screenshots, can you guess what you have to do and what your limitations are? Probably not exactly. We were misjudged by a few based on our screenshots. When they gave the game a chance, they came back positive. For example, Touch Arcade user VoodooVyper said, "I'm not a fan of these types of games," but came back later to say, "Wow, I was pleasantly surprised. Before playing the game, I read a comment that stated the game was something like Lights Out. I really can't stand that game, hence why I claimed I had no interest in this game. But I tried it out, and I enjoyed it...What's great about it is that it isn't Lights Out, and is nowhere near as frustrating as it...it really surpassed my expectations." Short version, our UI mis-sold us. 3. Prettiness: People like pretty shiny things. Right: We got it looking shiny -- Web 2.0 button shiny. Wrong: Pretty is subjective. Strangely, most girls who gave us feedback thought it was very pretty, but some guys were really turned off by the art style. The style was too plain to some, and therefore the gameplay was judged to be similarly monotonous. Plus, marketing to girls wasn't in my plan, as I have no contacts at female lifestyle publications, and we didn't expect them to be so receptive to the game. 4. Showmanship: People like to show off and compare sizes, so to speak; give them something so they can do that. Right: "Race the Clock" turned out fun with move counts as the score and a timer as another barometer of skill. Wrong: We didn't have time to add an online score table, which meant dick-swinging would be a solitary affair -- although that didn't hurt Traffic Control at launch. Luckily, our programmers set things up so we could add this with an update if we could put all the other pieces together. 5. Value: We figured 50 levels would sound great and give great value, provided we could release them as well-tuned and polished as possible. Right: We ended up with 100 puzzles and four 12-puzzle-in-a-row challenges, which was above scope and, according to the reviews, provided hours of fun, making it well worth our original price of $2.99. Wrong: Value is not perceived the way we expected on the App Store. If your game is the hot thing, it'll sell no matter the price. Fieldrunners was a small game for $4, but Time Magazine's "Best of the Year" accolade shot it sky-high and gave it a boost that still has it in the 1,000+ sales per day Top 50 list. Was our value proposition just not clear? Probably, since online reviews and user reviews point to it being worth more than the asking price. What Did We Learn? Despite learning all those lessons the hard way, Trixel has done okay -- which was down to succeeding on the quality side, which led to the exposure we needed to jump up that chart. We got a lot of great reviews and that helped us get noticed by the guys at CO-OP on Area5.tv. They featured our game in their iPhone segment, and our sales shot up that week. The week after, Apple placed us in its "What's Hot" window, which features on the front page of iTunes and the App Store. A week on that list brought us higher and got us a phone call from Apple as well as a placement on the "New and Noteworthy" list, which then shot us even higher up the charts into the top 40 paid games. However, we mainly saw spikes, not steady inclines. This leads me to believe there is something missing on the visibility side of our game, and that could be down to Trixel's icon and name. Its icon is eye-catching but not descriptive. The name "Trixel" seems to be memorable, it doesn't mean anything. Nobody knows what the hell it means. "iFlip" was our first name choice, but most "i+Verb" names are taken now, and we had to alter it just before submission. My guess is that those who didn't see something inspirational, relatable, or desirable in the icon and name didn't care for it and therefore did not click. Those who clicked through to the page saw pretty yet motionless screenshots. We improved those later, but because we didn't have time to refine the perfect UI, our screens could be regarded as abstract to new viewers who wouldn't bother reading the application description. We also had a stack of great press quotes in our page text. This leads me to speculate most users are concerned less with reviews than with perceived value, which is first judged by the screenshots and only as an afterthought by the text. I speculate that those few thousands who did buy Trixel came directly from the reviews -- the very same ones quoted on our App Store page -- and that those customers did care about quality in their assesment of game value. I believe few came from the Apple storefront, because sales spikes timed with specific pieces of coverage -- as well as, of course, the "What's Hot" and "New and Noteworthy" page placements. Another consideration is that since we have had so far only four weeks worth of sales, our data sample may not be comprehensive enough to establish strong certainties. These are simply educated guesses. Final Thoughts My final conclusion is that if you stick to the following guidelines -- based purely on my own personal theories from experience -- I believe you'll increase your chances of success on the App Store. 1. Have a catchy name, but make it descriptive. Or give your name a descriptive subtitle. Case in point: Flight Control or Brain Teaser. "i+Verb" names are mostly taken and will require a name check the day you plan to submit. Expect that your logo could be the last thing to go live, unless you have a pool of money and lawyers and can trademark the name at the early stages. 2. Release a competent product. By competent, I mean the game meets expectations. I expect to see a level of production in an EA-published iPhone games that stands above a Gameloft-developed game. If Secret Exit make a game, I expect it to be artful and well-produced. If a game is $5, I expect it to be above Flash game levels of depth, and so on. If your graphics look simple in style, you may show competency in simplicity but not in depth. Elegance or stylized simplicity may communicate depth. 3. Ship with as refined a gameplay interface as possible. Your screenshots should immediately demonstrate what the game does. What's the control mechanism? What's the goal? What's the success condition and the failure condition? Are there status mechanisms I can see? The style of your UI can flavour the context with further depth and character. 4. Understand the App Store audience's perceived understanding of value for money. True value is perceived through the question, "Will I be able to play this?" If it's out of their realm of understanding, it's not getting bought, unless it's very desirable and is recommended by another person or renowned source, such as a Time magazine, which honored Fieldrunners as the game of the year. 5. Give the social aspect focused thought. Mafia Wars on Facebook is a killer because it builds adding players into its reward structure in a well-tuned way that's obviously been through numerous iterations. A high score table is also fine, but it's all relative to the product and the suitability of that feature. Flight Control is a simple, well-paced game, and has a high score table. Mafia Wars has social interaction and a detailed set of comparable status indicators. 6. Think of your App Store icon, name and game page as a game retail box. What does the game do? What is it called? Why do I want it? Abstract imagery gets you lost easily, as people will (mis-)read into things we take for granted. A strong art style could allow for more subtle messaging. 7. Don't release into a PR vacuum. We got lucky. But if you're building off sweat and without pay, you need to be self-aware enough NOT to shoot the game out just because you're desperate to get money back as soon as possible. You need to make sure people know your game is coming, get them excited for it, and then make sure you get a PR guy who can get your reviews timed in large chunks. PR really is a full-time job. Hopefully this helps you. Why am I sharing it? If more great games are out there, our art form will grow quicker in all the right places and we'll all see more cool stuff before we die. [Dan Boutros was the creator and designer of the critically acclaimed Trixel for iPhone, and has a career path that's looks more like a dart board with random jobs and company names than a straight line. He sells and sometimes fleshes out game concepts for a living as a designer, and currently has a concept being developed into a flick at Union Entertainment. He also invented choco-bacon, and cheese-on-chicken-curry on toast.]