Reality Check: the iPhone 4 launch in perspective
July 1st, 2010
Daniel Eran Dilger
It’s not just the Gizmodo bloggers under investigation in the iPhone prototype case who are painting the electronic world’s biggest launch as a fair to middling failure. The mainstream media is also in on the act. How delusional are they? Here’s a look.
Selling iPhone 4
Apple’s online pre-order system collapsed immediately after AT&T’s server for handling subsidy authorizations failed. It turned out that for most people, the only way to order a phone was to schedule a reservation at an Apple retail store, something that itself was only possible after Apple skipped the authorization step for people using the Apple Store app.
Sure, not very cool for those “gotta be first” people who wanted to have iPhone 4 delivered to their home, and even worse for those who discovered that in-store pickup would take as much as six hours to endure. It was certainly no fun for those who lined up, but they still shelled out $200-300 to get it.
The reality is Apple was under the crushing load of getting 1.7 million units to those bleeding edge users in just the first three days. That’s 70% more phones than Apple shipped in the first two and a half months when it launched the first iPhone just three years ago. It’s approaching double the record-setting launch Apple pulled off last year.
What’s a good launch? Remember when Motorola was talking about how its Verizon Droid was doing “extremely well” back in January? It took 74 days for Droid to reach 1.05 million units sold. Apple sold over 60% more iPhone 4 units in a fraction of the time. About 14,000 Droids sold per day over a month and a half during the holiday season, compared to the 567,000 iPhone 4 units sold per day over a single summer weekend. What’s “extremely well” times 40?
Now look at Google’s own highly publicized (and non-demonized) launch of the Nexus One: Flurry Analytics says that only 135,000 sold in its first 74 days. Apple sold four times as many iPhone 4s on its first day as Google sold Nexus Ones in total over the next two and a half months; Apple continued to sell four times as many over and over again over the next two days. How many iPhone 4 units will ship 74 days after launch? A lot.
The Nexus One didn’t just fail because of a lack of interest; Google insisted that all orders would happen over the web, and didn’t allow T-Mobile to stock the phone in its retail stores. That’s failure of the self imposed, bumbling sort of variety. Yet there was no highly publicized complaint about how badly Google flubbed its first, and perhaps last, launch of a consumer product.
Apple delayed the iPhone 4 launch at AT&T stores, but only because it was too difficult to manage a launch that large across multiple vendors. When AT&T began offering iPhone 4, the lines began afresh. Apple is still selling a huge number of devices, and some people are apparently still waiting for a white model. Anyone who tries to say the iPhone 4 launch went badly from a sales perspective is simply not telling the truth.
Building iPhone 4
Sure, critics claim, Apple sold a lot of iPhone 4 units. But there are major manufacturing flaws! Spots on the screen, proximity sensors that reportedly don’t always shut the screen off properly, and manufacturing issues that delayed the availability of the white model. Turns out the screen spots were a temporary thing caused by rushing devices from the plant before the glue dried, a problem that solved itself literally overnight.
Now let’s frame some context for the word failure in terms of manufacturing. HTC’s Android phones, including the Nexus One, Droid Incredible and EVO 4G, all shipped with significant touchscreen sensor problems. The EVO’s screen has a “detachment” problem that appears to be caused by an adhesive failure of its glass screen. The phones have major battery life issues. HTC can’t source enough OLED screens so it’s having to replace them with conventional displays. Android users in general report lots of software crashes, random Bluetooth failures, and problems taking pictures (on hardware which despite high megapixel specs, is just not very good at taking photos).
What’s the acceptable failure/return rate among consumer devices? Something like 3-5%. For the 1.7 million iPhone 4 units that shipped in the first three days, that means we should see about 51,000 that don’t work. What about the rest of the industry?
A source within Microsoft told me that Verizon was suffering from a 25% return rate on Windows Mobile phones due to user dissatisfaction. Of course, part of that might be because of the software, but recall that HTC built 80% of the Windows Mobile phones. It now builds most Android phones. That’s why they’re suffering so many problems, even if the tech media is completely silent about it all. Readers don’t know what an “HTC” is, so it doesn’t make for a lurid headline the way writing about the iPhone does.
Recall that Palm also suffered poor build quality with its Treo line, which was also assembled by HTC. Despite finding a new supplier for the Palm Pre, that company continued to experience 10% failure rates for its webOS phone, more than three times the acceptable failure/return rate.
Remember too that Microsoft’s Xbox has suffered a failure rate greater than 50%. Some sources say the failure rate is approaching 100%, making service returns not a matter of if but when. Other game console makers report failure rates of 6.8% (Nintendo Wii) to 10.6% (Sony PlayStation 3). If Apple had that bad of hardware manufacturing problems on its state-of-the-art iPhone 4, it would be handling returns on between 115,000 to 180,000 new iPhones just from its first three days of sales. That’s not happening.
The iPhone 4 launch was exceptionally smooth; it was just covered by axe grinding bloggers with a fine toothed comb looking for minor flaws to turn into fanboy fodder.
Designing iPhone 4
Keep in mind too that iPhone 4 uses all sorts of new parts (including that Retina Display with an entirely new screen fusion process, a new gyroscope, and Apple’s new A4 chip), an entirely new industrial design, and packs it all into a tiny, battery operated device you hold in your hand. The only real complaint users can find is that it obeys the laws of physics, making it susceptible to breaking if you drop it without a protective cover. And you can attenuate the antenna’s signal with your hands (just not as seriously as the Nexus One).
Now let’s calibrate design failure by looking at recent competitors to the iPhone. The Palm Pre appeared last year with terrible hardware and an atrocious keyboard. It also didn’t help the SDK wasn’t ready, the software was still green, and the company’s cloud services were failing.
Motorola’s Droid keyboard is also terrible. The design of Android software, including the UI layers that radically change from vendor to vendor, is notoriously bad. Phones don’t get software updates until months after the updates are available. The Android Market is a mess. The array of buttons in the Android spec is also awful, and Microsoft is copying it in Windows Phone 7.
Microsoft’s first leg of Windows Phone 7, the now aborted KIN (previously known as Pink), started out as the fairly popular Danger platform. After acquiring Danger, Microsoft insisted on using a Windows CE kernel rather than Danger’s already functional Java-based system. In order to make the device profitable, Microsoft skimped on local storage and forced users to store all their data on the cloud (even photos they take with its camera). To keep service cheap, Microsoft limited message updates to once every 15 minutes.
It then launched the once cheap, kid-oriented platform on Verizon’s network, which tried to charge users $70 a month for voice and data service, despite the device lacking any support for basic smartphone services such as calendar sync, instant messages, or even any email accounts other than Microsoft’s own.
Microsoft then tried to sell users another $15 per month a Zune Pass music subscription, hoping to salvage its Zune business by tying it to smartphones. Unsurprisingly, after two years of development, KIN was pulled off the market in just 48 days after only selling a reported 500 units.
Clearly, there’s a lot of hyperbole begin directed at Apple’s iPhone 4 launch by the company’s naysaying enemies. Meanwhile, JP Morgan has revised its estimates to indicate an expectation that Apple will sell ten million iPhones in the third calendar quarter ending in September, a milestone it did not expect Apple to cross until the first calendar quarter of 2011. Considering that just three years ago, critics doubted that Apple could sell ten million phones in year, that indicates a well oiled machine is in operation.