Marko Karppinen

Right now, I have a hundred apps on my iPhone and none of them are at version 1.0.

All of those apps started at 1.0. Then one of two things happened. Either the app got a great initial response, and that justified continued development. Or, the first version tanked, and that justified continued development.

Either way, for successful apps, 1.0 was just the beginning.

Unfortunately, most companies looking to build an app are not thinking past the initial release. The 1.0 is what they are building and what they have a budget for. Everything they want from the app has been laid out in the RFP.

But this is the worst way to approach building apps because all of the value in an app project will be realized after 1.0. Wouldn’t you want to get there faster, with more of your budget still available to deploy after the 1.0 has validated (or indeed invalidated) your original assumptions?

So, a recommendation: Don’t spend more than half of your budget on 1.0. That’s when the real work starts.

None of the above is app-specific, of course. In the startup space, for example, a whole lean startup movement has emerged around Minimum Viable Products and the key insight that product development is mostly about learning through experimentation. But apps are also unique in a couple of ways.

First, the whole mobile app infrastructure is heavily geared towards thinking about apps in terms of distinct releases. You build a version, do heavy QA on it because it’s not easily updated once out, wait for Apple’s (or Microsoft’s, or someone’s) approval, then release. And then wait for your customers to update. Which some will never do.

Compare this to the web, where leading practitioners push new code to production every day, and it’s no wonder that most people aren’t conditioned to think about app development as a continuous process.

Second, with apps, a monthly release cycle is about as fast as you can reasonably go. A month is a short time in the grand scheme of things, but it can feel terribly long if your app’s current live version is buggy or unpolished. This means that Minimum Viable Apps can be small in scope, but they can’t be lacking in polish. In this sense the bar is set higher in apps than on the web — with web sites, the ability to address customer problems and user experience rough spots in what’s essentially real time means that much less up-front polish is required.

Recommendation two: Trim down your 1.0 by cutting features, not by skimping on polish.

Putting it all together, my rule of thumb is that a 1.0 iOS application should almost never take longer than three months to build.

With this approach, an example project with a six-month scope would be re-cast as as a three-month 1.0 project and three consequent monthly point releases.

Let’s say this app needs to satisfy a dozen distinct requirements. You’d prioritize them and expect to fulfill the most important six in the initial release. The remaining six requirements would then be tackled, two at a time, in the three point releases.

But the fact is that you are unlikely to ever get to those six lower-priority items. By the time you’d be ready to start on them, you will have shipped the 1.0, learned a lot, and have a whole new set of high-priority work to think about.

The beauty of this approach, and the whole point of this post, is that the expected value of that new post-1.0 work is much, much higher than anything you did for 1.0. You simply know so much more about what it is that you’re trying to do.

So get your app past the 1.0. My iPhone is waiting.

Bad news, everyone: it seems Apple stopped selling iPads in the U.S.

Making the rounds this week is a Pew survey that puts iPad market share at only 52% — down from 81% a year ago.

Back then, 8.1% of the survey respondents had an iPad. Now, one year later, 22% say they have a tablet but only 52% of those are iPad owners — making iPad’s share just 11.3%. This modest gain equals only 34 new iPad owners in the surveyed group of 1069 and is within the study’s margin of error. For all we know, Apple could have sold zero iPads in the U.S. over the past year.

What’s going on here?

The year-ago study was conducted over the phone and was delightfully simple. All respondents were simply asked whether they had a tablet, and if they did, the make and model. But crucially, if the respondent said that they had a Kindle or a Nook, the interviewer clarified that e-readers didn’t count as tablet computers. The new survey, conducted over the web, offered no such clarification. Kindle Fire and Nook Color appear in the multiple choice question without explanation.

In the 2012 survey, like in the old one, respondents were immediately asked if they had a tablet, but the question about tablet brands was at the very end: it’s question number 105. Only about half of the surveyed tablet users got far enough in the questionnaire to answer it.

Because of selection procedures, non-smartphone owners were significantly overrepresented in the group of respondents that did finally get asked about their tablet brand. In the overall survey, about two thirds of tablet owners had both a tablet and a smartphone, but at the point where respondents were asked to name their tablets, 55% of smartphone owners had already been eliminated from the survey. Based on that, it seems that there were potentially 2.2 times as many non-smartphone owners in this group as there were in the overall survey.

Do these things explain away the iPad’s plummeting share? I don’t know. It’s certainly believable that Kindle Fire now has significant market share. But I also think that the distinctions between e-readers, tablet computers and smartphones can still be too tricky for the general public, and I’m sure that some owners of e-ink Kindles ended up being counted as Kindle Fire owners with this methodology. Another likely source of bias is the overrepresentation of people without smartphones. I would assume that, as a group, they’d be oriented towards the low end of the tablet market.

Then again, I’m just a shill for my iPad-only magazine platform.

Magazine subscriptions are too cheap, yet feel too expensive

Imagine a cable TV company demanding $1032 upfront to set you up with HBO and ESPN for the year. Or AT&T expecting a check for $960 to activate 12 months of service to your smartphone (phone not included). Or a broadband connection? A steep but somewhat manageable $488.¹

That’s life in a world where services are sold by prepaid annual subscription. Seems kind of expensive, right? That’s why very few businesses choose to price their offerings this way.

Magazine subscriptions are a notable exception. There are a couple of reasons for this, like the general desire to dampen seasonality and the usefulness of a predictable circulation in the capital-intensive business of printing and distributing physical issues of a magazine.

Aside from all that, though, most magazine subscriptions are simply too cheap to be billed for more often than annually.

The average annual subscription cost of a U.S. magazine available at newsstands was $24.87 in 2011.² If that cost was billed for each month, the charge would be $2.07. Given the transaction costs involved, that’s simply not worth charging by conventional means.

But if you could charge monthly, it is almost certain that you could charge more. Here’s what David Ball, then the vice president for consumer marketing at Meredith, said to the New York Times in 2009³:

“It’s amazing how price-sensitive people are. Honestly, we’ve tested raising [the subscription price] 50 cents and we see a drop-off — sometimes startlingly high.”

Making the sticker price 1/12th of the current annual cost would help with this price sensitivity. When comparing the cost of magazine subscriptions to other communications bills, monthly charges would seem to make a huge difference. Like this:

Circulation data shows that families started canceling magazine subscriptions in 2008. Based on that timing, it seems that this was mostly to cut frivolous spending amidst the financial crisis. An annual bill for a magazine only comes once a year, but when it does, it is evaluated in the context of the other bills on the table at that time. A smaller sticker price would simply lead to fewer cost-based cancellations⁴.

iTunes to the rescue

Cost-effective monthly subscription pricing is one of the key features of Apple’s Newsstand. A typical magazine subscription will translate to the iTunes price points of either $1.99 or $2.99 a month. That’s right within App Store impulse purchase territory (not to mention the half-a-latte territory).

iTunes subscriptions can now also have a free trial period. For example, you can offer one month free and start charging $2.99 from the second month on.

In public, publishers’ reactions to Newsstand have always been dominated by talk of Apple’s 30% revenue cut and the company’s refusal to hand subscriber data over to publishers. But if you look at the specific things Apple has solved — discoverability, ease of purchase, monthly billing — the prescription starts to look appropriate for the disease.

After all, 70% of something is a lot more than a 100% of nothing.

The problem with magazines is in this chart

According to the MPA, single copy retail sales of magazines have been in a free fall for the past decade. During the ten years from 2001 to 2011, newsstand sales dropped a whopping 47%.

But for much of the same period, magazine subscriptions actually made gains. A decline only started with the financial crisis in 2008 as families cut back on frivolous fixed expenses. By 2011, magazines had only lost 7% of their 2001 subscription circulation. With newsstand sales — the primary way of discovering new magazines — decimated, that is an amazing feat of resilience.

So we simply stopped going to the newsstands

It’s no surprise, really. We also stopped going to Borders bookstores and to Blockbuster Video. Yet the consensus is that magazines are in trouble while books and movies are doing just fine. Why?

With books and movies, it was easy to recognize that the Internet presented a problem not for the products, but for their archaic forms of delivery. Bring the delivery (kicking and screaming, but still) to the Internet era, and you will have solved the problem. No fiddling with the product necessary.

With magazines, most observers have jumped to the opposite conclusion. That the product itself is to blame. And no wonder: it’s easy to look at the time we spend on Facebook, Twitter and on our iPhones and conclude that magazines, as products, are over.

But all of that is just a manifestation of the underlying delivery issue: the magazines we know and love are simply not present online.

Fixing the wrong problem

Wait, what? Almost all magazines have an online presence these days. On Twitter, on their web sites, everywhere.

Yes, but a web site for a magazine is not a magazine. It’s a web site. A Twitter feed for a magazine is a Twitter feed, not a magazine. The same is true of the multimedia experiences many magazines are building on the iPad. They are many things, but they’re not magazines.

Now all this may sound like an absurdly prescriptivist stance on what may or may not be called a magazine. Absurd or not, I believe that when readers and magazine fans think of a magazine, they have a very specific thing in mind. And that thing is something that most magazine publishers have failed to bring online.

Publishers are in a tough spot, and their urge to fiddle with the product is understandable. Social media experts, builders of digital publishing tools, even the professional ambitions of editorial staff are all driving publishers to try and reinvent the magazine. The holy grail in this line of thinking is the embodiment of a magazine as a social multimedia experience, using all of the capabilities of our new devices and networks.

Before going all-in on such a pursuit, however, magazine publishers would do well to make sure that their readership is willing to tag along.

Circulation data points to an alternative path forward. Readers of magazines are happy with the product. They just no longer frequent the points of sale.

Full disclosure: I have a horse in this race. I make an iPad magazine platform.

Apple continues to err on the side of insecurity

News about Mat Honan’s hacked Apple ID spread all over the web this weekend, but something about the story didn’t seem right to me. Mat was guessing that the hacker had found his password with a brute force search, but that sounded impractical: Apple is pretty aggressive about locking down accounts that get too many failed login attempts.

Now we know the real reason: Apple tech support gave the attacker access to the account. That may sound astonishing, but I have no trouble believing it. And that’s because the same thing happened to me in 2008.

The fact that the same attack vector works in 2012 proves that Apple still doesn’t take Apple ID account security seriously. Simply put, the front line tech support reps should never be able to perform a password reset like this. The fact that they still can means that Apple continues to err on the side of insecurity. What was crazy in 2008 is completely, utterly insane in 2012. Between then and now, here are some things that have happened:

• The App Store. The growth of the iOS platform has driven iTunes to 400 million credit-card enabled accounts. That means that the average monetary value of an Apple ID is a lot greater than it was in 2008. It also means that there are crazy outliers: Imagine a tech support rep giving out the password to a Rovio employee’s Apple ID. The bank account information for the multimillion dollar wire transfers Apple sends them for Angry Birds revenue could be reset in seconds.
• OS X. The security model of OS X is now based on the assumption that Apple IDs are more reliable than a Mac’s local accounts. This means that a social engineering attack on a tech support rep, quite possibly on a different continent, can be used to reset the password to the admin account on your MacBook. Or, like Mat experienced, to wipe all your data.
• iCloud. Back in 2008, the information I had on my .Mac account was largely stuff that I had knowingly stored there. Today, sharing and syncing of content is more seamless, and it would be exceedingly hard to know all the information that would be exposed in a security breach.

If Apple’s ecosystem is to succeed, Apple IDs can only grow in importance. The impact of an Apple ID breach is now much greater than it was in 2008. I can’t even imagine what it could be four years from now.

My Apple ID episode from 2008

On App Store Payment Policies

We all know that Apple has some pretty fascist payment policies on the App Store:

• Apple demands that all iOS content purchases be made using Apple’s In-App Purchase
• Processing payments within the app by credit cards or other means is forbidden
• Using existing subscriptions to unlock content will become verboten on April 1st.

Especially after that last item of news, one key question among iOS app publishers – including many of the ones we’re working with – is:

What’s with this greedy, overreaching, control-freak bullshit, Apple?

The most obvious theory is that Apple is simply a greedy and overreaching control-freak of a company – and one that would like 30% of all your content revenue, thank you very much. Apple’s take, in turn, goes like this (this is verbatim off the App Store Review Guidelines):

If it sounds like we’re control freaks, well, maybe it’s because we’re so committed to our users and making sure they have a quality experience with our products.

Let’s try taking Apple’s explanation at face value, accepting that the behaviors Apple is banning are less than user-friendly. It’s not a huge stretch: I for one wouldn’t want to be taking out the credit card whenever buying an iPhone app or content for one. I also don’t want to navigate and fill out web-based forms for subscribing to content if the alternative is a simple In-App Purchase.

Apple’s payment rules are primarily designed to make iOS a better platform. The 30% cut is icing on the cake – I don’t think it’s even a factor in Apple’s decisions here. Here’s why.

Since the launch of the App Store two and a half years ago, Apple’s 30% take has earned the company about $850 million in revenue. Meanwhile, iOS hardware sales earn Apple that much revenue every week. Percentage-wise, less of the hardware revenue is likely to be profit, but in follow-the-money terms Apple’s focus is still obvious. Above all else, their interest is in protecting and expanding the iOS hardware business.

iOS has two trump cards in the game against Android. The first is the App Store, with its 300 000 apps. The second is the fact that people prefer the overall user experience of iOS over that of Android. Now this “overall user experience” obviously encompasses third party apps as well. It’s crucial to Apple that people prefer iOS apps over apps on competing platforms.

It would be prohibitively hard and labor-intensive to try and affect the user experience of hundreds of thousands of third party apps directly. It’s much simpler to figure out a set of user-hostile behaviors and ban them. Importantly, Google is limited in their ability to do this on Android by their self-imposed commitment to “openness”. That, combined with the preference many publishers have for user-hostile behavior (profits first, user experience second), leads me to this conclusion:

If publishers prefer certain user-hostile behaviors, and Apple is the only one to ban them, then apps on Apple’s platforms will be more user-friendly than apps on competing platforms.

This helps Apple maintain their platform lead and pays off handsomely in hardware profits.