My Thoughts on How The AppUp Center Could Totally Freaking Rock
The genesis of this article goes back to Computex 2010 where I saw a few things that got me thinking. (You can read those initial thoughts here) Back then I knew there was more here though, but they still needed time to gel in my mind, there were edge-of-consciousness concepts that remained too slippery to write down in anything better than an outline. In fact, I’ve wound up delaying my own due date for this article like 5 times because the details refused to come into focus. But then a bit of news regarding Adobe Air came out and the pieces finally snapped into place. I think I’m ready to actually write this thing.
But I also need to confess that writing this article has been hard for me for more than the aforementioned reason alone. It feels like hubris to presume that I have anything really valuable to offer here and this whole article constitutes unsolicited advice to a genuine “captain of industry” and that freaks me out more than a little. But my desire to contribute to this thing is stronger than my self-doubt so I’m throwing it out there.
These are my thoughts on how and why the Intel AppUp Center can be one of the coolest things going on in computing in the next 10 years. Intel, if you read this and recognize the Socrates-level wisdom contained herein…send your check to the address at the bottom. Then I’d suggest you see that Adobe pays for half – but I’m getting ahead of myself.
First, a note to the playa haters:
Several folks took at look at AppUp at CES and trolled out some variation on this theme: “Another app store – yawn…”
Let me be clear:
YOU PEOPLE DON’T GET IT.
You’re not even close to getting it, and if you don’t shift your thinking you will completely miss what is about to happen irregardless[i] of AppUp’s flop or flight. And here’s why:
The app store concept is not a product or a service.
It’s a complete reset of the way ALL intellectual property will be sold, shared and distributed.
It will completely reshape the world of books, music and software.
If you see something less than that here you are missing the big picture. Ergo the remainder of this conversation will be lost on you. That said – don’t feel bad. The truth of the matter is we’re talking about a pretty major paradigm shift and stuff like that takes time to get noticed, much more to sink in. I can’t help but feel like the folks who DO get it remain the minority but I don’t say that with any sense of superiority. I’ve come to think this way almost despite myself. Some shifts happen dramatically, I think this one could happen almost without notice. People may look back at their world in another five years and realize that it changed without skipping a beat, and they’ll say ‘…meh.’ After working with Apple these last two years in the iThing ecology I wonder if the explosion of iPhone games and apps took them by surprise too. Why else would the earliest iPhones and iPod Touch models use an old and underpowered version of OpenGL? But regardless of what they expected back then, Apple clearly understands the power if this model now as evidenced by their ‘surreal’ market cap. So do tens of thousands of iThing developers hawking something like 200,000 apps. But more to the point of this article…
Intel understands this too.
When we first learned of the AppUp Center it was back before they officially launched at CES. At that time G:Into The Rain had been in the Apple store for about 6 months and while we were happy with the game – it was aging like any iPhone game. So a chance to stretch G’s lifecycle was a no brainer. But I didn’t get the big picture back then and while we took the plunge to port our game I shared the previously chastised “another app store?” voice in the back of my mind. Still, one word in that initial discussion really caught my ear and started to grow like leaven – OEM. Intel was talking about getting the AppUp center pre-installed on every netbook that comes off the assembly line so the first thing a user sees when they pop open that box from Acer is…wait for it…my game. Sure, it’s my game alongside several others but this was one of those genuine ‘ground floor’ moments where we had the chance to be one of a handful of launch apps when the list of available apps was measured in dozens, not the 100k competitors on the Apple store. What’s more, netbooks were selling in crazy numbers (and still are) so in terms of addressable market, this was a monstrous open door.
Apple has always essentially been, and still is, serving a niche market. While the iPod genuinely came to dominate the MP3 player space, Macs and OS X still only represent ~10% of the personal computer space. iPhones are unquestionably popular but not a majority. Like it or not, the WinTel platform remains the bulk of personal computing across the globe and here AppUp was poised to invade all of those rigs. With the installed-base question wrapped up you can’t help but reckon with a huge and extant body of coders who can crank out apps like a tidal wave…given the right motivation. All of these things together look to me like some very fertile ground. With skinable OEM AppUp stores no matter who wins the netbook war, Intel wins with them, and so do I.
So that’s the place I’m coming from. I think AppUp has a gigantic potential. But I also think that it is a long way from being a sure thing. Going forward in this article, rather than expounding on any one point too much (something I will likely do in the months to come), I’ll be leaning toward bullet points with brief comments.
What follows are my thoughts on a winning strategy for Intel and the AppUp Center.
Here’s What Needs to Happen: (IMHO)
1. Make it Cheap
Intel doesn’t need to attract the attention of established software shops nearly so much as it needs to attract the attention of geeks in garages with time on their hands to experiment. That means facilitating an entire pipeline including hardware AND software that an average team of 1 or 2 can throw down on pure hope and moxie. The biggest single barrier right now is Visual Studio and its $800+ price tag, but we’ll talk about that more down the road.
The best thing Intel can do is to crush the cost barrier for Indie developers. While they’ve done several things to this end like waiving the annual registration fee, the biggest cost barriers are not currently in their hands.
2. Make it Easy. More importantly: make it LOOK Easy
The second most important thing is to take something we think is hard and make it easy. The potential AppUp developer needs to see a robust API/SDK that does a lot of the most common and popular tasks by itself. The value of Apple’s brilliant iPhone SDK cannot be overstated. It gives easy WYSIWYG access to the iPhone features I just read about in Game Informer and now want to play with. Things like the camera and the accelerometer.
• And never, ever, EVER require me to know what a COM object is.
Unfortunately Intel’s current SDK is pretty anemic. It allows me to authenticate my app in the store, but that’s about it. They assume I already know how to build Windows software. (Did I mention the problem with Visual Studio?)
Whether it’s true or false, there is an impression that Windows programming is only for cyborgs. There are words in our vocabulary like “script kiddy” that imply C++/Windows programming is only for the highly educated or masochistic self-taught nerds. That perception may in fact be an insurmountable obstacle if Intel doesn’t step in and go around VS.
One of the iPhone SDKs best strengths is the IMPRESSION that any newbie could build an app. Interface Builder is absolutely brilliant! That one tool alone invited a massive surge of programmers who to that point had felt limited to web or Flash to say, “well this looks like something I can handle.” And so they played around in their spare time. Let me point out that these exact same programmers felt the same anxiety about Cocoa as they did about Windows. But the SDK convinced us that we could make it work by appearing simple. I mean, who doesn’t want to play with a Shark?
And the biggest single item here is a quick and intuitive way to create UI elements. No single tool will get more traction than an AppUp version of Interface builder because most programmers frankly suck at visual design. Giving them the tools that will make their UI at least look nice is priceless.
Note that it really doesn’t even matter if the tools ever get used. There must be the impression that I can create an app successfully and everything I need to be a “real” programmer is right here.
PS: An emulator that mimics netbooks is a MUST HAVE.
PPS: An emulator that mimics other Atom platforms like vehicle systems, Smart TV sets and home environmental control panels…you’ll see wicked innovation.
3. Do Apple One Better
There’s no avoiding the fact that the AppUp Center will be compared to iTunes. Don’t fret about it – own it. Find the places where Apple’s system can be improved upon and deliver that experience first to the developers and then to the customers. It’s completely OK to launch a “me too” product so long as you do it better. And once you’re out of beta (ahem…) you need to decouple AppUp rapidly from iTunes by carving your own path with your own features. I’d also stay away from any of the items that have recently garnered Apple raspberries like iAd.
Here are just a few ideas that could help set AppUp apart:
• Break the link between an app binary and its metadata. A developer wants to churn through keyword iterations quickly without having to upload and approve again. It’s a little thing, but important.
• Make apps giftable. One of the real bummers on iTunes is the way I’m held hostage to their marketing whims and I am really limited in my ability to create direct sales. Every other thing on iTunes can be sent as a gift – why not apps?
• Filter the wheat from the chaff. From the developer perspective, perhaps the most frustrating things about the Apple store is the failure to differentiate between apps with genuine thought behind them and the infinite flood of shovelware. The Xbox Community / XBLA model seems like a good place to start but it could be better. You don’t want to create barriers to developers but if AppUp can circumvent the extreme downward price pressure we feel on iTunes it will compel us to give AppUp a solid look (since I have no way to separate my app from the 3 dozen fart generators). Fix that one problem and you would go a long way to attracting iPhone eyeballs.
4. Never Use The Word “MeeGo”
The vast majority of the population fears change and sleeps with the devil they know. The moment we hear about a new operating system we think about change and we hate change, even if it’s change for the better.
Neither customers nor developers really care what operating system they use. Developers want to know where we can sell our software. Customers want something that stays out of their way. Given those things, the OS is a detail like the L2 cache – only useful in nerd wars. When you start talking about MeeGo as a selling point you invite the lizard brain in me to enter OSZealot mode – and there is no escape from that trap.
Don’t get me wrong; MeeGo ought to be a cornerstone of the strategy, specifically because it avoids Visual Studio[ii]. But you can NEVER mention it as a selling point. The best operating system is the one I never notice; so don’t draw my attention to it.
Nobody knew what Objective-C was three years ago and the only reason anyone cares now is because they have already swallowed the iThing hook. If you want to say “…you don’t need to write Windows code…” then maybe that could be a selling point but the only people who make the business decision based on their favorite kind of variable typing are uber-nerds who ought to like Linux anyway and open sourcers who don’t have (but still seem to want) money.
PS: Be sure O’Reilly has at least 2 MeeGo books in Amazon when you launch. If they don’t, write them. These have to be ready for sale so the developer who says “I’ll give that a shot” can be instantly engaged and invested.
5. (After never mentioning it) Make MeeGo Beautiful and Bulletproof
You cannot underestimate the value of a few gorgeous screenshots, fluid gestures and transitions, all viewed on YouTube in the blessed silence of the bug hunters.
People need to say about MeeGo “It Just Works.” And the best way to do this is to avoid biting off more than you can chew. If you launch a rock solid and pleasantly surprising SDK with say 20 features that absolutely rock you’ll keep us developers busy for 3-6 months as we experiment and flood the AppUp with little nuggets…some of which will be genuine gems. Then you release a rev that unlocks another 10 features which will send us all back to our basements and Tesla coils to experiment some more.
A huge part of this is serious documentation. Lots and lots of samples and explanations. Don’t just write a tech manual but give real-world examples, especially in the places where ambitious developers want to tinker.
Another cannot-be-underestimated element here is the fact that the bulk of your experimental developers will be making games. Gamers are completely intolerant of bugs – finding and talking about bugs is like a special merit badge for gamers. So you can’t afford to have any kind of beta available to the general public. The real release needs to our first impression and it must must must be bug free. (BTW, It’s August, and AppUp is still in beta…I’m just sayin’)
Besides providing exactly the wrong grist for the gamer mills it has a seditious way of reinforcing the Windows-Is-Hard thought. After all, if even Intel can’t make solid software what chance do I have?
I’m sure that must sound dreadfully unfair and please don’t misunderstand me, I’m certain there is a lot going on behind the scenes but again it comes back to a place where perception is at least as important as the technology. I would really suggest that Intel become maniacally intolerant of bugs in their own software. Speed to market is important but a late launch is far more acceptable than a botched launch since the current hardware/software environment takes any mistake and magnifies it 100X. The iPhone 4’s antenna issue cost Apple $10B in market cap in one day.
6. Apple Is Not Your Enemy – Google Is
This may be an esoteric point but I think Intel has more to gain by being Apple’s kissing cousin than its competitor. Apple is buying Intel chips for their Macs and while people keep trying to put the iPad up against netbooks, that comparison already fails to hold water. So except where folks want to compare AppUp to iTunes these two companies have a lot to offer one another, particularly when the Atom gets stuffed in iPhones…where they belong.
Google, on the other hand, is trying to make desktops irrelevant. And with the recent lawsuit surrounding Android, this could be a war worth taking sides in. If Google has its way Intel will sell a lot fewer chips of every variety. If Apple wins, Intel will sell more chips. Seems pretty simple to me.
7. Show Me The Money…and Put That Filthy Lucre Away!
In the end, developers want to get paid for their efforts and that means bringing customers to AppUp. On the one hand this must be the most obvious thing in the world. What may not be obvious is when to bring out the loot bags and how to approach the topic with developers.
Developers are a funny bunch. By and large we consider ourselves a kind of artist. In our minds, we are not motivated by money but we expect to be well paid for our expertise. The trick here is to present the AppUp Center as a powerful profit opportunity without it sounding that way. You have to imply the profit potential. If you come out too strong here you risk provoking the purist. If you don’t make the case clearly enough you risk loosing attention by attrition.
What Intel really needs is the first “I made $1m selling apps in the AppUp Center!” story to make the wire, and then a second within another month. Of course I’m not suggesting that kind of thing be artificially manufactured. It has to be a genuine success and that will only happen when the customers arrive and start buying apps en masse.
8. Don’t Try To Be All Things to All Programmers. Focus and Thrive.
Supporting 300 different languages and frameworks is a disaster waiting to happen. It’s utterly unscalable and invites the kinds of bugs that will do huge amounts of collateral damage to the AppUp brand.
Support MeeGo and Air – drop everything else.
Air is the thing that can happen today. MeeGo is your future divorce from Windows. Just like the OS point in number 4, most programmers really couldn’t care less what language they use. Sure, we all have our favorites but in the end it’s an academic question. You won’t lose anybody by limiting their language options. But if you try to support .Net, Java, C#, RedHat…where does it end? The more languages you support the more difficult it is for you to maintain any kind of quality control. I suspect you may be compelled to support C/C++ at least for a while but then we’re back to the Windows ad Visual Studio problem…
9. Leverage and Cooperate With Existing Services
This is not in contradiction to #8 even if it seems that way. What I’m talking about here is resisting the temptation to bring too many things in-house. A great example would be YouTube (Google being your enemy notwithstanding). It would be a huge mistake to try to make your own YouTube inside AppUp. When you can’t beat them gleefully join and enable them. To be honest I wonder if there couldn’t be an iTunes app inside AppUp to sell music. Even if Intel makes exactly zero cents on the transaction you gain a huge amount of mind share by keeping the user inside the AppUp experience…where they will buy more apps.
10. Plan For, and Encourage, Other Forms of IP
I’m pretty darn certain this is something Intel is already thinking about but it’s worth bringing up. Software is the tip of the iceberg. EBooks, Music, Movies, TV Shows – all of these things exists as digital assets than can be distributed via download. AppUp can replace the vanity press business and do similar things in almost any creative endeavor. I’d suggest taking an X-Prize kind of approach to gather content here. Offer up some kind of absurdly large prize for aspiring authors who publish their book exclusively on AppUp. Stuff like that can prime the pump in no time and really engender competition. But there needs to be a goal somehow, not just a contest. DARPA didn’t pay out to the best robot car. They paid out to the team that cleared a rather high bar.
This One Goes To 11
My last point here is really the biggest one and it deserves some special treatment.
Adobe Air could be the single most important thing that happens to AppUp.
Here’s the deal – there exists a massive horde of Flash/Flex developers who feel disempowered on two fronts. C-flavor programmers have dissed them for years as something less-than. Their work has been relegated to free Flash game sites and ActionScript continues to be mocked as some kind of redheaded stepchild. More recently, this talented body of mad scientists has been slapped by Apple and they are singled out as the only people who don’t get to play in Apple’s walled, moated, and portcullis covered garden.
If Intel makes a big deal about Air and really courts these people, especially if you create that structure that gets these people paid for their work, AppUp will break the tethers.
It looks like this:
1. You don’t need to court iPhone developers.
2. You need to court Flash developers. (and they will be much cheaper dates) These people will be like the school wallflower who never gets asked to dance – any gesture that validates their skills will make them pant in gratitude.
3. When a flood of Air games hits your shelf the customers will come.
4. When the customers come the iPhone developers will come too.
5. For less money and effort – you get way more of what you want.
This was the last piece of the puzzle that really needed to snap into place for me but when I saw it I had a genuine epiphany. Air could be the linchpin here because you don’t have to educate a generation of developers (like you do with MeeGo) and you can turn the sour grapes thing between Apple and Adobe to your benefit. If you’re really good you can play the savior/mediator who steps between those two giants and negotiates a kind of peace.
I’m tempted to go on and on here because of the weight I see in this one point but I reckon the idea is just that simple. Embrace Air, really, really EMBRACE Air, and your app catalog will explode.
There are little things you could do here that would have big payoffs like providing an Intel code signing certificate so broke developers don’t need to buy their own. And adding some kind of Air plug-in to the SDK to make it a 1-clcik integration. Schucks you could build your SDK IN Air, leverage a ton of pre-existing and mature APIs and you’d attract developers like flies.
But There Are Sill Other Problems
While I think those nine things would make for a winning strategy it’s not to say they solve every challenge.
• What’s the compelling reason to buy through AppUp instead of outside of it?
• Is it worth Intel trying to create a vertically integrated experience from hardware to software?
• When I can play 10,000 free Flash games on the same netbook that charges me for Air games, why should I pay?
• Apart from netbooks, the computing continuum that Intel imagines is held hostage by hardware manufacturers who don’t have the same goal or vision.
I’m certain all of these problems can be solved, but they remain problems and most of them are currently beyond Intel’s immediate ability to address.
To wrap up here, my exposure to Intel over the last several months has given me a few momentary glimpses into something I think will be really, really big. I like what the AppUp Center can be and I REALLY like the idea that we get to be part of something (potentially) so formative.
I hope you folks find something interesting here.
One more thing – you don’t need to kiss developers’ asses. You’re Intel for crying out loud – one of the largest and most respected brands in the world. If you invite developers into your bigger story, if you celebrate and lift up the folks who are early adopters and innovators, if you confer honor to developers in a way that treats them like valued allies…that will earn you gigantic points. Going back to the beginning of this article, I certainly saw a business opportunity when we thought about porting it to AppUp. But the truth is it felt like I was being noticed and validated by a company I help in extremely high regard and there is no measure for how much that motivated us. Intel is not an also-ran-me-too competitor in this thing and don’t let anybody tell you that you’re playing catch up. Walk in your strength and lead – we will follow.
i I am completely aware that irregardless is not a word. But some sick, twisted part of me likes to taunt the spelling and grammar wonks…alot. It’s presence here is specifically for you Michael.
ii That’s 3 times Balmer…pay attention or perish.