The Magic Trackpad is Just a Better Mouse
For all the hyperbole about the mouse being dead, no one has called out the Magic Trackpad for being what it really is—just a better mouse. Though it is touch based and enables certain gestures, the power of a real touch based OS like iOS is in direct manipulation via a touch screen. Here’s how I put it in my chapter on iOS UI in iPhone User Interface Design Projects:
“One of the keys to creating great UI on the iPhone is taking a step back and thinking a bit about how users actually interact with the iPhone—with their fingers. Yes, that’s incredibly obvious, but something so obvious generally caries significance that few people take the time to explore.
The finger is an incredibly efficient pointing device, far more efficient than the mouse. When the mouse was first introduced, it revolutionized human interaction with computers. I would argue that Apple’s multitouch interface [on a touch screen] will, in time, prove to be even more revolutionary.
A mouse [and the mouse-like Magic Trackpad] manifests an unnatural disconnect between the motion of the user’s hand and the action on screen. Most people these days have used a computer enough to be somewhat accustomed to the mouse, but if you watch someone use a computer for the first time, you’ll see a definite learning curve to using a mouse! I’ve even noticed a bit of a learning curve when using a new mouse or tracking settings with which I’m unfamiliar.
But the human finger, that’s something just about every human is quite accustomed to using. We’ve learned since birth how to control our fingers with an amazing level of precision and speed. Imagine if you were to attempt playing the piano with a mouse. Your tune would be choppy and unmusical at best. A touch-screen piano, however, would be playable, even if it didn’t match a real piano for feel and accuracy. The more I thought about the finger as a means of interaction the more intrigued I was by how drastic the shift was from the mouse to multitouch [on a touch screen]. That’s when it finally hit me. Taps are cheap!
If the appropriate action is obvious to the user, the time actually required for that user to tap the proper spot on the screen is miniscule. Confusion about where to tap wastes far more time than an extra tap.
Again, this conclusion may seem quite obvious. After all, ambiguity has been a challenge in all human computer interfaces, and reducing ambiguity has been one of the pillars of good interface design. But the iPhone is the first graphical computer interface where the speed and precision of the pointing device makes the physical action of pointing almost irrelevant when considering the time it takes to accomplish a specific result. Let that sink in for a minute—taps are cheap.”
Using a mouse (and mouse-like Magic Trackpad) is expensive. Moving the cursor around a screen and precisely aligning it with buttons and performing other actions takes significant time and skill compared to direct manipulation with fingers on a touch screen. The gestures of the Magic Trackpad do make certain actions faster and more natural, but it’s still just indirect manipulation of objects on a remote screen.
To better understand what I’m saying, think about a Magic Trackpad piano where different parts of the trackpad triggered different keys. Sure you could use more fingers and eventually do better than a mouse based piano, but there would be a HUGE learning curve as you memorized the position and spacing of the keys on the blank surface. And if the piano keys were just buttons on a computer screen, the Magic Trackpad gives no advantage over a traditional mouse, you’d still be moving a single cursor around a remote screen.
The Magic Trackpad is a small evolutionary step in human computer interaction, touch screen multitouch is revolutionary and will live on long after the mouse really dies.
Anti-Competitive AND Potentially Creepy
I’m at WWDC and don’t have time to fully polish my thoughts, but I thought this was important enough to post a rough draft… please excuse the rough edges and rambling. I was quoted in the Wall Street Journal yesterday, but the story just didn’t go deep enough into my thoughts about this fascinating turn of events.
The new terms in section 3.3.9 of the iOS developer agreement are all about Apple taking back control of how developers and third parties access and use sensitive user data on their iOS platform. The amount and detail of data that can be gleaned from a person’s mobile device is absolutely staggering and I honestly think that Apple has been a bit naive about how developers have been collecting, processing, and using data from iOS devices.
Last fall I was pitched an advertising platform for displaying the Honeywell ads in my Gas Cubby by FRAM app. The goal of this ad platform was to build a network of apps using their ad analytics so that they could cross pollinate data to deliver laser focused ads. Let’s say one app collects a user’s UDID, and because the app has sync, the person’s email address, or maybe the GPS coordinates of all the gas stations they’ve visited in the last 6 months (because an ad is served every time the user launches the app at a gas station, and the ad requests the user’s location to provide a targeted ad). Then that same person plays a game that asks their age to customize the experience (and that data is also packaged with the UDID as all analytics are). Then a 3rd party browser app sends a users search terms to better serve ads. And other games just log gaming sessions, sending the user’s UDID and detailed stats on when and how long they play certain games.
So, as the ad network grows, it’s absolutely incredible the depth of information that developers can collect and send to analytics/ad platforms over time. And that data is all quite easy to correlate based on the unique device identifier (UDID). Do you remember what a big fuss everyone made when Facebook used Beacon to start delivering “more relevant” ads by scraping personal information. Well, that’s what’s been slowly building right under the noses of mobile users and generally without their consent.
When you use Google search and other Google products, they collect a tremendous amount of information and use that information to customize and better serve the ads that are the core of their business. Many users don’t even realize this is happening, others are comfortable with it and have some level of trust for Google’s intent in using that data.
Well, Apple doesn’t trust the benevolence of Google, developers, and other third parties involved in the iOS platform. Apple wants to control the flow of user information. They may use more detail in targeting iAds than they are going to allow others to use for their own ads and other analysis, and that’s a competitive advantage, but it’s a fair competitive advantage for them to maintain on their own platform. Apple hasn’t said AdMob can’t advertise on iOS, just that they must get written permission from Apple for ANY user and/or device information that is sent back to AdMob servers.
For reference, here’s a story about Amazon not allowing third parties access to user’s purchase history to prevent them from usurping Amazon’s incredibly valuable recommendation engine: http://bits.blogs.nytimes.com/2010/04/22/what-danger-do-blippy-and-swipely-pose-to-amazon/
But this goes way deeper than just screwing Ad Mob/Google and having a competitive advantage for iAds, it’s about Apple taking back control of how user and device information is accessed on their platform.
At the end of the day, when someone buys an iPhone, they are putting a certain amount of trust in Apple. And Apple is positioning the App Store as a place where users can trust the apps that they buy. See Neven Mrgan’s excellent post about the benefits of Apple’s walled garden approach: http://mrgan.tumblr.com/post/653708588/the-walled-garden
If Apple allows 3rd party apps unmitigated access to user data, they’ve essentially passed that trust and responsibility down a level to developers and other 3rd parties. If Apple is going to position the App Store as a walled garden of apps that are safe to buy and use, they MUST control these aspects of what’s going on under the hood.
If Apple didn’t do this, a year from now a self-conscious woman would look down at her phone and see an ad promoting weighloss products to overweight 41 year old women with thinning blond hair who live in a blue house and drive a black Ford Tauras. And that’s going to scare the crap out of her. Who’s she going to blame? What product is going to trashed in the press for enabling this kind of eerily specific advertising?
Many people think that this kind of targeting is the future of all advertising (it has been slowly and subversively taking over the web), and it might be, but mature companies realize that privacy is a very delicate thing and must be treated with extreme care. See this excellent Wired article about Thefind.com deciding it’s still just too “creepy”: http://www.wired.com/epicenter/2010/06/facebook-thefind/all/1
By controlling the flow of information and how targeted the iAd platform becomes, Apple is taking back control so that it can decide what is appropriate. And I trust Apple in that regard a hell of a lot more than I trust Google, Facebook, etc. The thing is, Apple is a hardware company, that’s where they have and will continue to make their money. Google, Facebook, and others trade in information. The more detailed and specific, the more valuable that information. For Apple, the better the overall experience of the device, the more valuable that device becomes. They can throttle ad targeting and the specificity of 3rd party analytics according to the taste of users. Trusting 3rd parties to do so would be incredibly foolish, and Apple seems to have just recently figured that out.
Then there’s Google’s incredible competitive advantage in being able to track copious amounts of device and user demographic data, usage patterns, and other data on a competitor’s platform. This deserves another full blog post, but I’ll quickly preview my thoughts…
If you think about it, Google (via Ad Mob) being able to collect specific usage data about iOS would be a HUGE competitive advantage in shaping their own Android strategy.
This is a completely reasonable attempt by Apple to prevent their direct competitor from gaining a competitive advantage. In fact, Google may have spent the $700 million (a ridiculous sum of money for what it appears they were buying) for just this reason. It’s essentially like the Greeks having been able to just buy the Trojan Horse, with warriors already positioned inside, from someone who had established trust with the Trojans.
Oh, and there’s also AdMob’s incredibly flakey “Mobile Metrics” reports. They do such a hatchet job on those it must drive Apple nuts. Why do you think Apple cited so many “more reliable” sources of statistics on the mobile market during the keynote on Monday? Don’t even get me started on how poorly AdMob has been handling that data. Here’s a really smart take on it: http://www.cultofmac.com/admob-owned-by-google-shows-android-overtaking-iphone-in-web-traffic/40491
That’s pretty rough, but I’m still processing…
david
Made, Is Making, or Will Make?
Properly conjugating a verb helps distinguish among past, present, and future events. Mobile developers and the tech press seem to have an awful time with conjugation, often making bad decisions or writing misleading headlines on account of that apparent confusion.
Gas Cubby *made* $32,446.54 in the 30 days following March 25th, 2009—the day it was selected by Apple as the Staff Pick of the Week and featured prominently throughout Apple’s online ecosystem. I’m not actually sure what Gas Cubby *is making* because Apple doesn’t allow real time sales tracking. I do know that Gas Cubby *made* $266.34 yesterday and I hope it *will make* the same or more today, but using daily or monthly sales as a predictor of future earnings is perilous in the rapidly changing, cutthroat mobile app space.
After nearly 2 years in the App Store and having crafted countless spreadsheet models, I still have trouble accurately estimating future sales. Sales skyrocket when an app is covered by the press or featured by Apple, then seem to settle into a predictable pattern, only to plunge 50% or more for no apparent reason.
Unfortunately, that doesn’t stop people from busting out a calculator for a little presumptive math. Let’s see, $266.34 a day x 30 means Gas Cubby *is making* $7990.20/month ($97,214.10/year). Interestingly, if you use those same calculations for March 25th, 2009, you’d get $93,215.10/month ($1.1 million/year). Try February 11th, 2010 and you’d get $5,381.70/month ($65,477.35/year).
The problem is, when a present tense verb is used with a specific duration of time that extends into the future, it turns into a sort of presumptive present tense (Wikipedia tells me that’s a modal tense, as opposed to a pure tense). Take for example the sentence “Bob is working today.” If you left it at “Bob is working” it would indicate his present state with no indication of what might happen in the future. Saying “Bob is working today” indicates his present state, but also leads to certain presumptions. Depending on the time of day the statement was made and Bob’s occupation you might assume that he’ll be working until 6PM, but there’s really no way to predict Bob’s future. So, if it’s March 2nd and someone says “Gas Cubby *is making* $7k a month” most people interpret that sentence as “Gas Cubby *will make* $7k in the month of March.”
Hearing that an Android developer *made* $13k in one month is encouraging for the platform, but to say that this particular developer *is making* $13k a month is subtly misleading. With all the press surrounding his blog post, I wouldn’t be surprised if Edward makes more than $13k in March, but as the press hype starts to fade and Google inevitably removes the app from its featured position, I doubt Car Locator *will make* anywhere near $13k/month. I’m not saying he can’t build a business around this app and innovate his way back into favor with Google and the press; but, if I were him, I wouldn’t quit my day job just yet.
When I started App Cubby in April of 2008 I jumped head first into the fray without any kind of backup plan; no day job, no savings. Then a month later I found out my wife was pregnant. I wouldn’t say I thrived under that pressure, but I survived and somehow managed to build a few great apps and learn a lot along the way.
The stress and countless 100 hour work weeks make me a bit more touchy when I see misleading headlines and presumptive math in the mobile space. Sure, a few developers have struck it rich—some deservedly, some not so much—but it’s a complex and ever changing market. I wish the tech press would take a more responsible approach to covering the space rather than sensationalizing the “mobile app gold rush.” The distinction between *made* and *is making* may seem subtle, but it’s important.
Greener Pastures in the App Store
Monday night I finally decided to pull the trigger on a decision I’ve been debating for quite some time. I moved Gas Cubby from the Utilities category to Productivity.
As the App Store grows, the “Top Paid” list (the top 100 apps by sales volume) in each category has become crowded with various levels of crap, soft porn, and cheap apps. I don’t know the exact number, but when Gas Cubby first launched in November of 2008, the Utilities category had less than 1k apps. Now it’s overrun with almost 10k.
The problem with the Utilities category is that most of the apps in the category, especially the popular apps, are priced at $1.99 or $0.99. That makes it difficult for an app like Gas Cubby ($6.99) to stay in the top 100 of the category even though it’s one of the top 20 grossing apps in the category.
Apple doesn’t display category specific Top Grossing lists (the top 100 apps by revenue) publicly in the App Store, but it’s easy to see them using an app like MajicRank, or with a bit of URL hackery. Here are several links to category specific Top Grossing lists (links work on the iPhone and desktop):
Productivity: http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewTop?id=25244&popId=38
Utilities: http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewTop?id=25284&popId=38
Finance: http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewTop?id=25172&popId=38
Now that Gas Cubby is in Productivity and showing up in both the Top Paid and Top Grossing lists of the new category, I have a few interesting data points to analyze. Monday night Gas Cubby was #75 on the Top Paid list and #19 on the Top Grossing list in Utilities. This morning, Gas Cubby debuted at #32 on the Top Paid list and #15 on the Top Grossing list in Productivity.
I had assumed that the sales volume in the Productivity category was going to be lower, but I was just hoping to break the top 50 and didn’t expect Gas Cubby to land so high on the Top Paid list. It’s also interesting to see the disparity between the Top Paid and Top Grossing lists for each category. Having only moved up a few spots from one Top Grossing list to the other, it appears that the top apps in both categories generate roughly the same revenue even though the Utilities category does significantly more volume.
Being able to analyze the Top Grossing list and compare relative position with my competitors has had a huge impact on my decision making process. In the past, I would freak out and drop the price or take some other drastic action any time Gas Cubby started sliding in the charts. But since Apple first introduced the Top Grossing lists in September 2009, Gas Cubby has been grossing significantly more than any of its competitors and has spent most of its time as a top 20 grossing app in Utilties. That extra bit of data has given me the confidence to ride out some of the scarier days aboard the App Store roller coaster. Even so, the App Store is all about visibility and I’ve been wondering what a new category might do for Gas Cubby.
Moving from #75 to #32 on a Top Paid list should give Gas Cubby more visibility, but it’s hard to tell without knowing much about App Store shopping patterns. Do more shoppers view the Utilities category than the Productivity category? The overall Top Paid list has been proven to drive significant sales, do shoppers even spend much time looking through the Top Paid list in each category? How much does being “above the fold” (first 25 apps on the iPhone App Store, depends on window sizing for the iTunes App Store) impact sales?
It’s obvious why Apple can’t allow developers access to raw App Store traffic and sales logs, but I do wish they would give us a glimpse into our own presence on the App Store (like how many users view our App Store listing each day and whether it was on the iPhone or in iTunes). As a developer I’m doing everything I can to “read the tea leaves” and optimize my position in the App Store. With access to so much more data, I sure hope Apple is working hard to improve the shopping experience for users AND optimize business opportunities for developers. Every time there is an Apple event looming, I get my hopes up that the App Store (and iTunes Connect) will get some love, but progress with the App Store has been slow… slow, but steady.
Touch Screen Edge Cases
While fiddling with the Nexus One last week, I was incredibly annoyed at how many times I accidentally tapped the Home button while trying to tap the space bar on the software keyboard. I posted this little note on Twitter about the issue: “DON’T EVER PUT BUTTONS FLUSH WITH THE BOTTOM OF A SOFTWARE KEYBOARD. PEOPLE WILL ACCIDENTALLY TAP THEM AND BE PISSED” (http://bit.ly/8MUQL2).
At the time, I felt something was fundamentally wrong with the “hardware buttons” (really just an extension of the touch sensitive area) being flush with the edge of the screen, but it didn’t dawn on me until today the particular problem it presents.
The unused edges of a touch screen serve to expand the “target” of any button positioned against them.
While typing on the iPhone I’ve been subconsciously tapping slightly below the bottom row of the software keyboard because it requires less precision. My fingers were using that muscle memory while typing on the Nexus One keyboard. By putting the “hardware buttons” flush with the bottom of the screen, the Google/HTC engineers made the bottom row of the software keyboard (or any other button at the bottom of the screen) inherently more difficult to use.
Here’s a video example of Palm having made the exact same mistake on the PRE: http://bit.ly/8gCqEO (starts around 00:40)
There are so many subtle things that Apple got right with the iPhone. Whether or not there was ever an iPhone prototype with more “hardware buttons” or a “gesture area” we’ll never know, but I do know that being able to visually define and re-define all aspects of an app’s UI within that 320x480 screen has let to some incredible fun and innovation.
Quick thoughts on the Nexus One
Overall I don’t think Apple has much to worry about with the Nexus One, but Google sure is trying. The thing is, my 55 year old mother LOVES her iPhone. I could never recommend something fiddly like the Nexus One to my parents (who took to the iPhone like a duck to water). The logarithmic growth in smart-phones is happening in the non-techie crowd and Apple clearly has the best device for that market and a huge lead in mindshare.
