Maybe it's possible to have too many developers

I never thought I'd say that, but Apple's making me wonder. Apple says 100,000 developers have already downloaded the iPhone SDK (link). For comparison, it took Palm a couple of years of heavy evangelism to hit the same milestone. That's a deceptive comparison, though -- Palm was a small and relatively unknown company at the time, whereas Apple is a huge brand, with a large base of current Mac developers that it can bring over to the iPhone.

So the process was quite a bit easier for Apple. But still, criminey, 100k is a lot of SDK downloads in just four days. Apple has clearly struck a nerve.

In addition to the 100k statistic, the Apple press release included new endorsements from companies like Intuit, Namco, PopCap (Bejeweled, baby), and SixApart. It's a very interesting variety of companies.

Which raises an interesting question: How in the world will all those developers find an audience? Based on the current iPhone installed base, Apple already has one developer for about every 40 iPhone users. Let's be optimistic and say every user will spend an average of $20 a year on applications (a figure much higher than we've seen on any other mobile platform). Apple keeps 30%, so there will be a total of $560 in revenue per year available to each iPhone developer (not to each app, to each developer). The iPod Touch will increase that number a bit, but not enough to change the math. Either Apple has to grow the installed base enormously, or there are going to be a lot of iPhone developers going hungry.

18 comments:

Stefan Constantinescu said...

the maths in your post is right which leads me to believe one of two things:

1. there will be an unlocked iPhone option.

2. the iPhone family will grow.

not having 3G is stopping a lot of people from adopting this and being locked to one operator isn't too attractive either.

don't forget that a lot of people downloaded this just to play with it and don't really plan on making any applications at all.

june can't come any sooner.

Michael Mace said...

>>don't forget that a lot of people downloaded this just to play with it and don't really plan on making any applications at all.

Very true. And that was true of Palm OS as well. But the number of downloads of the Apple SDK is still growing rapidly, and I probably overestimated the amount that the average user will pay for apps. If the iPhone is at all like the Palm experience, a lot of people won't buy any apps at all, which will pull down the average.

It's an interesting problem for Apple. Because of the power of the Internet, they can scale their developer community at web speed. But they are selling a hardware device, and constrained by carriers. They can grow their customer base only at phone speed.

Unless they license their mobile OS, which I don't think Apple is going to do. But if they did...that would be Interesting.

Chris Dunphy said...

"If the iPhone is at all like the Palm experience, a lot of people won't buy any apps at all, which will pull down the average."

The iPhone has done a great job of being nothing like the Palm experience. The best thing Apple has done is NOT copy Palm, and I expect that they are actually going to do an on-device app store right, and they will create something that average users will actually explore and use.

Look at how well they have managed growing the music store, training people to buy songs on their PC. Apple has now even passed Best Buy, and is the #2 music retailer in the US. I expect that within 18 months (if not less), Apple's iTunes app store will be selling more mobile applications than all the other "stores" combined, unless the competition starts figuring out the real issues here.

I tried for years to build a platform standard way for PalmOS developers to easily sell direct to users, and we never managed to pull all the pieces together. But it looks like Apple just might.

- Chris Dunphy // www.twostepsbeyond.com

Anonymous said...

Does this announcement change the dynamic between Apple and Nokia any? (War Between Nokia and Apple) If Nokia's purchase of Trolltech and Qt was for developers, knowing that Apple would be releasing its SDK soon, then at 150,000 developers it still has more than Apple. (Qt Product Information) The question for Nokia now is, how quickly can it develop an SDK for S60, S40, and N*** series devices and create an integrated music-videos-applications store. The battle is really heating up, with Apple using one phone with a firmware upgrade and the iTunes store to try and overcome Nokia's distribution channels and phone portfolio...personally, I hope to have a Nokia WiMAX tablet and VOIP phone that can use the Maemo platform and Debian repositories and the Sprint Xohm WiMAX service :-)

Brandon

Anonymous said...

Hey, it's basic economics; to stop going hungry, all the developers have to do is develop something people might actually want - an app for reading eBooks springs to mind - and will happily pay for.

Seriously, how many widget-alike apps do we need for currency conversions, weather forecasts and world time?

Oh yes, and it wouldn't hurt the developers cause to remember that there are quite a few iPhone users outside the US, too...

Michael J. said...

$20 per year per user is far too low. If you think of the iPhone as a miniature computer, instead of a beefed-up mobile phone, then a better comparison would be $/yr/usr for Mac OS - which has got to be far higher than $20. There are going to be a lot of $1 and $2 and $5 impulse-purchase apps (like songs or mini-albums on iTunes) and they will add up.

And, don't divide by the 100K SDK downloads, but by some % of that - possibly as low as 10% or 20% of that.

The implicit thinking in your post is, "Who will be the next Microsoft in mobile," whereas a lot of developers just want to earn a living - and $50K or $100K is enough. Just like the Mac enables smaller, independent graphic artists (for example), the iPhone SDK enables small developers.

A few will hit it big. Most of the ecosystem will be small potatoes, and lovin' it.

Dave Haupert said...

I thought the number of downloads was interesting as well. But two takeaways:

1. While Palm was able to get over 100K SDK downloads, there were never 100k apps, more like 5000-10000, and considering that developers like myself released a dozen or more on our own, it seems that there were likely about 3000 active developers of commercial/shareware/freeware. That was certainly sustainable with the Palm userbase, so hopefully it will be with Apple as well.

2. There seem to be a lot more iphone users than there are iphone users! I read that StyleTap mentioned they had over 500k views of a youtube video demo, read that the unlock software was used more than that, and yet as a percentage these just seem way too high. I think developers of other platforms and users of other platforms, whether iphone users yet or not, all are watching very closely.

Tim Meyer said...

Clearly there will be a lot of iPod Touches (I have not seen any market share estimates) so the target market will be growing rapidly on that front as well. Many apps will work fine without a cellular network.

Another factor to consider in the revenue calculation are ongoing revenues from a service provided through an app (which would probably be free). Not sure how such billing events are triggered through the appstore, I have not heard of any API for this.

Albert said...

Spanish translation posted here.

Regarding what Stefan says in the first comment,

"not having 3G is stopping a lot of people from adopting this and being locked to one operator isn't too attractive either."

I tended to think that, too, but the lack of 3G has not really stopped the 'gray' market of one iPhone from every four being unlocked. People here in Spain seem to be perfectly happy using the iPhone when WiFi is available and just talking and texting when it is not.

As per the mobile apps, the commercial market is almost non-existing in Spain. I'm curious to see if iTunes changes that.

Olivier said...

I think what is wrong is thinking that each SDK download equals one "commercial" developer. I downloaded the SDK, and know a few others who did, but really more out of curiosity.

For a while, Palm was counting every SDK registration and download as a developer, but in reality you would never find more than 2000 people at the dev conf, and really was the same crowd of 500 core developers. (and i bet mike you know all of them)

Now, There's no doubt the iPhone is reaching beyond the "mobile developer" crowd; reaching core Mac developers, and probably a good amount of the web developers as well.

Michael Mace said...

Thanks for the interesting comments, folks.


Brandon wrote:

>>Does this announcement change the dynamic between Apple and Nokia any?

I think Apple has tremendously raised the ante for anyone who wants to have a thriving native mobile apps economy -- Nokia, Microsoft, etc.


>>The question for Nokia now is, how quickly can it develop an SDK for S60, S40, and N*** series devices and create an integrated music-videos-applications store.

And figure out how to certify those apps, and negotiate the studio deals to get videos into the store, and...

I think one of the biggest challenges for Nokia is that Apple keeps adding more stuff for Nokia to respond to before Nokia is through responding to the previous one.


>>personally, I hope to have a Nokia WiMAX tablet and VOIP phone that can use the Maemo platform and Debian repositories and the Sprint Xohm WiMAX service :-)

That would be fun to play with.


Michael wrote:

>>$20 per year per user is far too low....There are going to be a lot of $1 and $2 and $5 impulse-purchase apps (like songs or mini-albums on iTunes) and they will add up.

The last time I checked, the average iTunes user was buying three songs a month (link). That's a total of about $36 a year. So I don't think my guess of $20 a year for apps is at all unreasonable.

But it'll be very interesting to see what happens. Personally, I hope you're right.


>>The implicit thinking in your post is, "Who will be the next Microsoft in mobile,"

Actually, no it's not, and I'm sorry you got that impression.


>>whereas a lot of developers just want to earn a living - and $50K or $100K is enough.

My point was that even that goal may be difficult for a lot of developers unless Apple grows the iPhone installed base very rapidly. Which maybe they can do.

A lot of people are enthusiastic like you. There's a nice commentary here.

The trouble starts when 100,000 developers each assume that they can get "just" one percent of the market. You run out of percents after a while.


Dave wrote:

>>While Palm was able to get over 100K SDK downloads, there were never 100k apps, more like 5000-10000

Actually, the figure was about 20,000 (I was the guy doing the counting for a long time). But your point is a good one.


>>That was certainly sustainable with the Palm userbase, so hopefully it will be with Apple as well.

Hope so.


>> I think developers of other platforms and users of other platforms, whether iphone users yet or not, all are watching very closely.

I agree strongly.


Tim wrote:

>> Another factor to consider in the revenue calculation are ongoing revenues from a service provided through an app (which would probably be free). Not sure how such billing events are triggered through the appstore, I have not heard of any API for this.

I'm not sure if the app store will handle that sort of thing at all. There are a lot of unanswered questions once you start to dig into it -- for example, how exactly will they certify, and does that $99 fee cover all of your recertification for a year? What if you turn your code every two weeks (which a lot of web-oriented developers do)?


Olivier said:

>>For a while, Palm was counting every SDK registration and download as a developer, but in reality you would never find more than 2000 people at the dev conf, and really was the same crowd of 500 core developers. (and i bet mike you know all of them)

And it was a privilege to hang out with them, too. They were a great group of people. I hope Apple can re-create that sort of atmosphere.

What got to me about Apple's numbers was that it took Palm years to get to 100k downloads, and Apple did it in four days. Extrapolate out Apple's rate of SDK distribution, and they're on the road to having a much larger developer base than Palm had. And it'll be very hard for four million units, or even ten million, to support all of those people.

I'm not saying that no developers will be successful, but in every gold rush most of the prospectors don't make any money.

Krakowian said...

A few other points to take into consideration:

1. Free apps. I'm guessing we will see tons of free apps, and people doing things to get around the $99 fee, like grouping together under an umbrella.

2. Vertial market apps. I'm thinking real estate, medical, business, sales, etc. These things will be going for a lot more than $20 I bet. :-)

3. Niche market apps. IMO, a bit different from vertical market. I'm thinking thigs like translation software (Hoping to see Lingvosoft enter the iPhone market, myself), or Bible software.

4. Games and toys. I think we can see a bunch of fantastically original games that will command a higher price. However, here, the competition will be very high, and there will be losers, but I can imagine that the winners will be true winners. It will be interesting.

In any case, I don't see the high number of SDK downloads as a bad thing. It means healthy competition.

Thanks again for a though-provoking article.

-Jon

andy said...

While 100,000 download is impressive, Google's Android SDK has been downloaded 750,000 times (since Nov 07). I am more eager than ever to see what Android and its open-source apps can do for mobile.

http://technology.timesonline.co.uk/tol/news/tech_and_web/article3554119.ece

Jane Sales said...

I am 5 of those downloads. If everyone had the same difficulties I did, then the figure can be reduced substantially.

Anonymous said...

Well if this is the new "Gold Rush" then the numbers make sense... Many try and few succeed. I think we can all agree that it will be really hard to come up with 100,000 (even 50,000) unique mobile applications that are good enough for people to download (and I am not even talking about buying).

Of course developers have to make money and monetizing mobile application is even harder. My guess is that there will be a lot of "causalities" here...

Monty Alexander said...

IF Apple accepts that 100,000 developers have downloaded the iPhone SDK, its not amazing in current time, while the population and the mobile phones user are rapidly increasing day by day, and apple has kept a strong base in mobile phone market whether in India, US, Uk or any other country

gibtang said...

The number of downloads is certainly surprising, but not all will actually create something using the SDK and I know quite a few who just downloaded it out of curiosity. And those 100,000 are only numbers from the official website. There are plenty of torrent seeds on the internet where you can download the SDK since people who are in other countries such as Singapore are unable to download the SDK from the official website since the download link only seems to work for users in certain countries.

Anonymous said...

Sorry I'm late everybody ;)

At least one analyst is predicting that huge numbers of iPhones will be sold in the next couple of years, in the order of 45 million. Of course, this number is probably inflated, but it still would change the numbers in your equation.
Additionally, a good portion of the people who develop for the platform won't charge anything for their apps. At least, that's what Jobs predicted at the announcement.

In spite of this, it reminds me of something I heard in a college ag course about corn plants. It's my understanding that if we stopped cultivating corn that it would become extinct within a few seasons. The reason being that corn ears have been developed to produce too many seeds. When the ear falls off of the plant at the end of the season, too many plants grow in the same concentrated area, depriving each other of moisture, nutrients and sunlight. The plants would never get tall enough to reproduce.

Crappy analogy I know.

Still, there will likely be a multitude of disappointed developers.
-Brad

P.S. Subtract one from your list of downloads, I downloaded it strictly out of curiosity. Never programmed anything in my life ;)