Polar Bear Farm

I don’t think it’s any secret to most that the launch of the iPhone SDK and the App Store was rushed. Having seen the state of the iPhone frameworks in version 1.x of the iPhone OS through my jailbreak work, version 2.0 of the OS really did clean a lot things up quite dramatically, there was a lot of work done. It was actually an extremely impressive feat. To anyone who had developed under jailbreak it was clear why Apple had not released an SDK for developers when the iPhone first launched, it simply was not in a state which was releasable, Apple were still working things out themselves.

It’s also interesting how the launch of both the iPhone and the App Store forced Apple to stray briefly from its usual tight lipped policy of never commenting on unreleased products. The iPhone was announced months before it was available, as was the App Store. There are a number of reasons why they really had no choice for this particular product launch. I’m sure one of the reasons for their standard no comment policy is that it forces hard deadlines on themselves. My perception is that Apple don’t like shipping products which are not ready for prime time, they’ve made that mistake in the past and paid for it. Hard deadlines can interfere with that philosophy, hence the first most people know of new Apple products is when they’re released. I think those deadline pressures came very close with the original iPhone, then again with the launch of the App Store and 2.0 firmware.

The App Store has been modeled almost exactly off the iTunes Music Store, in many ways it’s the obvious way to go. It’s got the store front, it’s got the ranking logics, it’s got existing user accounts and billing info, it’s got the reporting, etc etc. The Music Store is a clear success, so why tinker with a winning combo? Well fundamentally, apps aren’t music. It’s that simple. The one price fits all model of music just doesn’t jive with apps. Why? Because apps are wide and varied in their complexity, function, target market size etc. The Music Store is optimised for this one price fits all scheme. The ranking algorithms etc are all perfect for that case. That’s really the first major flaw in the App Store design at the moment:

The One Price Fits All Assumption.

The ramification for this assumption in the App Store is that as sales on the store are driven significantly by visibility. The Top 100 list rules. You need your app on that list! While the exact details of the algorithms used to generate these rankings are secret, it’s clear that it’s overwhelmingly driven by app download count, on some sort of rolling average. The easiest, cheapest way to increase your download count, is by dropping the price of your app to $1. It use to be that developers would change their app to free for a few days, then switch to paid, then ride the Top 100 Paid list that way, but that practice has since been blocked by Apple. $1 is the new free in this game now. So the end result of this one price fits all assumption, is that in general apps tend to converge on one price, and that one price being $1. This isn’t just speculation either. We’ve experimented with pricing a little, and the results are disheartening. As one fellow developer put it: “The App Store makes me feel dirty… dirty as in I just participated in a shit fight and now I have to go home and take a shower.”

First looking at ‘Record‘, our successful audio recording application. On October 16th, we doubled the price from $1 to $2, subsequently download numbers dropped by 59% to 41% of their previous numbers. That’s significant. At the same time we increased the price of ‘Note Pad‘ our iPhone notes replacement application, from $2 to $3, and downloads dropped by 63% to 37% of their previous numbers. On November 22nd we again increased the price on ‘Note Pad‘ from $3 to $4, and once again download numbers dropped disproportionately by 60% to 40% of their previous numbers. Clearly it seems that in our experimenting, it is not possible to maintain revenue while raising prices. These are sobering numbers, and support the idea that the one price fits all assumption, which powers much of the app store rankings, is a positive feedback loop biased towards pushing most app pricing towards the $1 price point. Of course there are other factors involved in this, however I think that this assumption is a dominant factor driving this. So what’s the solution? Easy, tweak the ranking algorithms to try and avoid the current natural discrimination against variable pricing. Simply give more weighting to app revenues, and maybe even bring in ratings into the ranking algorithms to try and push more of the best apps into the list, rather than the just the best of the least expensive apps.

Demo Applications

The ability to offer demo version on the App Store is simply not possible. This also contributes to the low pricing trend. Many customers are reluctant to shell out money, even $5, $10 or $15 for an application they can not test in some way first. ’Lite’ versions are not demos. With all the applications we create, we’re trying to build up trust within our customer base, so that when they see one of our apps on the store that they don’t own, they can purchase it with the confidence that it’ll work, and work well. The ability to offer time limited demos of applications is a simple way to help build this customer trust and confidence quickly. Currently it’s easier to discount the price of applications to a point where people don’t care if they download an app, and it turns out to be absolute crap. It’s simply not a good situation for customers or developers.

Customer Reviews

This is where some improvements have already been made on the App Store, which is great to see! From a developers stand point, while there are many changes which would be good to see, two of the biggest additions to the system would be the ability to flag blatantly false reviews, irrelevant comments, and the like on your own apps and have them investigated in a priority ‘developer reported’ queue across all stores. Currently there is no mechanism for developers to report customer reviews in stores outside the one they hold an iTunes account in, even using the existing reporting methods. Secondly, many customers see the review section as a support forum. It’d be great if developers could have one reply per review, to respond to questions, or to correct inaccuracies.

Of course we as developers need to take responsibility for the success of our applications, we can’t rely on Apple to sell our products by virtue of simply having them listed on the store. However the things i’ve mentioned are significant issues which in my mind need to be improved. Hindsight is 20/20 as they say, but a lot of these flaws really do beg the question, did anyone at Apple really think through the App Store design and operation, or was it just a case of hacking on an apps section to the Music Store to meet a deadline?

Finally I urge developers and customers alike to contact Apple and express your opinions and thoughts through the channels they provide: enhancement requests through http://radar.apple.com Developer Relations contacts, and Technology Evangelists for developers, and Apple’s official feedback channels for customers: http://www.apple.com/feedback/ They can’t improve things if they don’t know about them, so help them out by letting them know what you think.

Layton Duncan

It’s coming up to 5 months since the App Store launched, and things have panned out very different to how I had imagined. You know I had been looking forward to the launch of the App Store. We could finally get our software out there, support it properly, and have all the time consuming overhead of managing payments, license key generation, distribution etc out of our hands, and spend more time doing productive stuff like developing and making our apps better.

We were making good money off our Jailbreak apps selling them for $10 a piece, some people liked them so much they donated over $80 a piece for them. We invested countless hours developing them, not only because it was a fun challenge, but because the support was there from our customers to be able to live off this and be able to develop them full time.

This was a market with at its peak maybe around 1 million phones, users had to pay using PayPal on their iPhone (a horrific experience if ever i’ve seen one), then wait for a license key to be generated so they could enter it into the app on the phone, which would unlock all features or to get rid of popup reminder dialogues. This is a world away from the seamless experience of purchasing songs or apps from iTunes. Naturally we though that the launch of the App Store would not only open us up to a significantly larger market, but that the time consuming process our customers had to go through previously would disappear, and sales would continue heading upwards.

The reality is that we barely manage to match the money we were making with our Jailbreak applications through the App Store. Even at the peak so far, where “Record” was Top 10 paid downloads in 22 countries, and DuckShoot in 10 countries, sales barely reach those seen in Jailbreak.

The point here is that in the current market it is very difficult for professional iPhone development companies who’d love to commit fully to developing great apps for this platform, to be able to do that sustainably. The market is not at a size where $1 apps can really sustain dedicated iPhone development houses, skewing the market towards part time, after hours developers with a day job to support themselves, and App Store income as ‘pocket money’. 

While that in itself is not a problem, we know first hand how hard it is to provide the level of support we need for our products. It’s a time consuming part of the business, one we’d love to improve, but one which is a constant time battle, there’s simply not enough resources, which ultimately comes down to not enough money. This time balancing act is even more apparent for part time developers, and often does end up effecting quality, both in customer support, and continued application development.

But all this isn’t just groaning about the state of things, it’s something to hopefully start triggering thoughts about why the market is currently skewed like this, and how it can be improved for customers and developers alike. We want to be around to see our 2nd year anniversary, still helping to make this great platform even better, but ultimately the market and developers themselves will decide how things play out over time. I’ll add my opinions on things, having seen things from the very start of this game.

I have to say this is a topic which i’ve been reluctant to write publicly about until now (i’m expecting some ‘interesting’ responses), preferring instead to file bug reports, feature enhancements, and talk directly with our Apple Developer Relations contacts about, even emailing the top dog himself on one occasion. Apple does listen, but clearly can not always act as nimbly as some would like. But it’s far easier for them to fix or change things if they get feedback directly through the channels they provide. While the App Store has improved slowly since launch, it had major flaws from day one, which is the topic for part two of this post, along with some cold hard stats on App Store economics.

Layton Duncan

On the 24th of October we will be celebrating our 1st birthday. I thought i’d write a little about how the company started and the experience up to now, not only because hopefully it may be interesting to others, but also because in the day to day business of getting things done, you often forget to sit for a minute and reflect on where you’ve come from. Being the first commercial developer for the iPhone and iPod Touch has made this last year one of the most interesting and enjoyable of my life. I started out a year ago, releasing a small application called “Search”, it was the first application to allow people to search contacts and calendar events by keyword on the iPhone, a feature which is still absent from the phone. It instantly became popular amongst the jailbreak community. At that stage, there was no official SDK, no documentation on how to develop, just header files dumped from the framework binaries from the phone, and the foundations of the iPhones desktop operating system counterpart to go by. But that’s only half the story, none of this would have been possible without the talented and dedicated people who spent countless hour in first jailbreaking, then building a toolchain for those who where chomping at the bit to develop for this platform. While we supported the jailbreak community, even sponsoring a significant repository for applications when it was threatened with closure, for us, jailbreaking was not a rebellion against Apple’s control. It was simply out of a passion for the platform, that we couldn’t wait to write applications to make the phone even better. We assumed that Apple must have been preparing official means of native app development, that was what we were wanting and hoping for, but in the mean time, we couldn’t wait.

Search was such a success, that a week after it was released I decided that I wanted to go take this to the world and exhibit at MacWorld ‘08 in January. I called long time friend Guy Horrocks who had always been interested in entrepanurial startups and asked him if he’d be interested in coming on board and help build this company. A week later we had a 10′ x 10′ exhibit booth booked in San Francisco for MacWorld 08. So we were two guys from the bottom of the world in New Zealand, where the iPhone wasn’t even legitimately available, flying half way around the world to San Francisco to attend the worlds biggest Apple conference, exhibiting software for the worlds most revolutionary mobile platform, which had no SDK, no official way of distributing the software, without even knowing whether there would be native 3rd party support in the future! Looking back on it, it was kind of a cheeky, and risky move, but one which i’ve never regretted.

In the few months before the show, I was playing with various aspects of the phone and software, but the thing I was most interested in, was finding out what the limit of the power of this device was. I spent hundreds of hours playing with layer kit (core animation as it’s now called) in particular the CoverFlow interface. Hours were spent trying to reverse engineer the methods used by the iPod app on the phone trying to get a customizable CoverFlow interface we could use in applications for non music uses. I eventually gave up trying to figure it out, and wrote my own CoverFlow version from scratch, using layer kit. (As it turned out, my previous code which tried to use existing API’s was only a line of code away from fully working!) I even managed to get live video playing in a CoverFlow style interface. The abilities of the phone shocked me. This lead me into thinking about video recording on the phone. Apples iPhone presentation demo rig which mirrored the iPhone display on an external display, and later the Aspeslagh Bro’s (Ecamm) video streaming hack demo, triggered a video obsession in me. Could the iPhone record video? This was a huge technical challenge, after a couple of weeks of sustained work, reversing various frameworks, looking for hints and methods which may be useful, trying many different approaches. The biggest challenge being, how to actually get pixel data to write to the iPhone’s flash memory. I finally cracked it at 5am one morning, after working trough the night and several straight days. An immensely satisfying moment which went on to produce ShowTime, the first video recording application for the iPhone, which was ready for release just a week before we left to exhibit at MacWorld.

MacWorld exceeded all our expectations. Interest in what we were doing and our applications was through the roof, especially from Apple employees. The days, while long and exhausting, were encouraging for the future in this industry. These two applications, Search and ShowTime, were and are still two of the most requested additions to the iPhone, and it’s those two products which truly launched us as a serious iPhone development company, before any sight of the SDK. Search went on to reach over half a million downloads, and ShowTime well over a million downloads, including a significant number from within Apple, in what was an ‘underground’ hacking market. A feat which we’re immensely proud of.

Since then we’ve been lucky enough to have worked with some of the most interesting in the community, travel to various places, and meet with some of the biggest brands in the world! On the back of MacWorld we formed a close relationship with a company called GoGo Apps (now known as Tapulous), after meeting CEO Bart Decrem at the show. Since then we’ve worked along side them writing the initial jailbreak version of Twinkle (a Twitter client with a heavy focus on location), camping out at their Downtown Palo Alto office for the month around WWDC and the lead up to the App Store launch, we had the pleasure of working with their team which at the time included Sean ‘iApp-a-day’ Heber, Thomas Muldowny, Mike Lee, Louie Mantia, and others. We’ve also had the opportunity to travel to various large corporate headquarters and conferences.

Now with our existing products on the App Store, Telegram, Dictaphone, Note Pad, and Duck Shoot and with some longer projects simmering away, we’re thrilled to be celebrating our first year as a native iPhone developer, making us one of the oldest around! Keep an eye out closer to the 24th of October, as celebration plans launch into full swing. So from the team here, Guy, Cody and Myself, thanks to those that have supported us from the jailbreak days, through to our App Store transition. You’ve helped build this company from the most unlikely beginnings! Stay tuned, we’re just getting started!

Layton Duncan