Posted by & filed under Uncategorized.

Surviving the Microsoft Interview

Microsoft’s corporate interview process involves the use of logic puzzles to figure out how you think. It’s intended to be like the Kobayashi Maru test, but with cheap index cards instead of billion-credit battle simulators.

The Microsoft method, unlike the Starfleet version, is very easy to hack.

In my first round of interviews for a full-time position at Microsoft, I was given the old chestnut “You have unlimited time and resources. I want you to move Mount Rainier 6 feet to the north”.

My first proposal was that, given that we had infinite time, we should wait for plate tectonics to do the job for us. That was rejected, but it did lead to a discussion of usable project constraints.

I then harried the interviewer to provide best-estimate constraints.

He responded with one year and $60 billion dollars. As an incentive, I could keep whatever funds were left over.

I asked how he would measure my success. He said he’d stand on the summit with an accurate GPS tracker, and would accept any value within a radius of 2 inches of the desired location.

I proceeded with an analysis of the costs of the project, but I was stalling for time, because I’d had an flash of an idea and I needed to set the stage.

The interviewer kept suggesting dumptrucks, but I knocked that one down, based on the difficulty of slicing Mount Rainier into dump truck-sized cubes, indexing the cubes, and storing them until we reassembled them. Why, the setup alone would take a minimum of 6 months to create the required tools, let alone deploy them!

I then announced my fool-proof plan: a $30 billion campaign of bribery, extortion and assassination at the USGS and Pentagon mapping offices, designed to insure that all GPS devices would, at the end of the project, report that the summit of Mt. Rainier was in the desired position.

It wouldn’t be an abrupt transition, just an accumulating error as one got closer to the summit. All we really needed to do was upload a fiddled software update to the entire GPS satellite fleet, and backed by a Bond-villain budget, we could get ‘er done on time and under budget.

I spent the next ten minutes on the general minion-hiring process and an outline of how to identify and handle the Bribable, the Extortionable, and the (most unfortunately) Expendable target groups, using the USGS bureaucratic structure as an example. I did this one standing up, turning the plan into an actionable flowchart.

The interviewer was somewhat flabbergasted, and took a few minutes to respond. I used the time to ponder my next curveball.

There were the usual objections about legality; I countered with the far thornier legal and practical issues involved in getting permission to move Mount Rainier.

He objected that the GPS and survey maps wouldn’t match. I allowed his objection, and added $1 billion to the project cost for deploying additional assassins and assorted minions to adjust the master cartographical databases at the USGS and Pentagon. All we needed were tiny tweaks in the survey marker data, really. Probably wouldn’t have to get into wet work at all.

He brought up the European and Russian GPS-clone mapping systems. At the time, the Euro-GPS system was years away, so I ruled out that objection immediately. I politely suggested that he, as chief Mount Rainier project manager, needed to broaden his technology reading.

I then went into an analysis of the Russian system. I acknowledged the difficulty of hiring reliable minions in the former Soviet Union and the problem of identifying persuadable targets under a secretive regime; I balanced these problems against the likelihood that anyone on the project could obtain Russian mapping data.

I concluded that, if need be, we could cover the Russian contingency by using at most $2 billion to provoke a serious international incident with Russia, guaranteeing that no Russian data would be available.

I then proposed an alternative.

Would the interviewer like to be $20 billion richer at the end of the project?

P.S. Thanks to my good friend Skry for triggering this post by sending me an excellent article on how to get a job in user-experience design.

Posted by & filed under Uncategorized.

From my inbox:

Job Description

The this program is a transformational, strategic supply digitization initiative at , binding software to hardware in an effort to reduce friction in ‘s product key supply chain and reduce piracy on an upcoming release of one of ‘s major revenue drivers.

Problem Uno

Some — in fact, MOST — job inquiries I’ve been getting recently have been from people who obviously have no knowledge of the skills required to fulfill the jobs they’re flogging.

One tell-tale sign of a clueless recruiter is the hurried copy/paste/mangle process they follow when creating their come-on emails.

Problem Dué

Do I want to work for someone who can turn “We’re building an App Store clone” into a boggy swamp of buzzwords?

No, I think not.

Posted by & filed under Baggage Check.

I’m instituting a new feature here at Channeling Design – the baggage check*: calling out people who make over-reaching statements about business or technology.

Intel Senior VP Renee James, Intel Developers Conference, May 2011:

“On ARM, there’ll be the new experience, which is very specifically around the mobile experience, specifically around tablet and some limited clamshell, with no legacy OS … Our competitors will not be running legacy applications. Not now, not ever [emphasis mine],” said James, referring to rival chip manufacturers that use ARM’s SoC design.”

[via Information Week]

I’m sure that similar statements were made by VPs of Roman marble quarries during the early stages of the Concrete Revolution.


* Yes, this is my version of John Gruber’s “claim chowder” list. I bow to the master.

Posted by & filed under Uncategorized.

The first phase of any serious design project has to be a serious brainstorming session. If you don’t understand the problem space before you start, you might as well hand off the project to an evolutionary algorithm. Of course, that could take years and cost thousands of lives.

My favorite way to get a handle on a problem is to spend a couple of days in a mind map. I ask myself and the rest of my team a LOT of questions.

When the humans involved start shunning me, I start cross-examining my mindmap by throwing bits of it at Google to see what comes out the other side.

For business problems, I’m particularly interested in seeing how my client’s competitors tried to solve the same types of problems. (I generally take the traditional method as either a great idea or an insufficiently-examined problem.)

Here’s one in a series of mindmaps I made on the first day I joined a site redesign project for a regional auto repair chain.

Click image for full size

Posted by & filed under Apple Watching, Meme Scams.

Paul Suarez at PCWorld sees the response to the rock-bottom remainder sale of discontinued and never-to-be-updated HP TouchPad…

TouchPads Lesson: Tablets Cost Too Much | PCWorld.

Here’s the problem with the current system: many entry-level tablets cost somewhere around $500 and that’s the same price as the iPad. I’m guessing most consumers that decide to spend a $500 on a tablet will opt to get an iPad.

HP offering its discontinued tablet for a one-fifth the cost of Apple’s tablet seems to have registered with many price-conscious and deal-hunting consumers.

…and sees a market opportunity for crappier, buggier, less-functional tablets.

It’s the perfect device for cheap and unwary users and high-volume, ultra-low margin manufacturers. It’s a godsend for OEMs looking for one last chance to repackage the New Hotness of 2004.

Meet the Netpad, the successor to the wildly popular Netbook.

All Hail The Netpad!

Posted by & filed under Developer Relations.

I received an email from HP webOS Developer Relations tonight. It’s a masterful exercise in saying … really not much at all.

Dear webOS developer:

We have opened the next chapter for webOS, and we understand that you must have many questions. Yesterday we announced that we will focus on the future of webOS as a software platform but we will no longer be producing webOS devices. While this was a difficult decision, it’s one that will strengthen our ability to focus on further innovating with webOS as we forge our path forward. Throughout this journey, our developers will continue to be a vital part of the future of webOS.

We will continue to support, innovate and develop the webOS App Catalog. Our intent is to enhance our merchandising and presentation of your great products and to continue to build our webOS app ecosystem.

As many of you are aware, we are currently scheduled to hold many developer events around the world. We are planning to continue with these events, however, due to the recent announcements; the nature of them will change. These updates will be posted on our events registration site this coming week. We are eager to present to you the updated strategy for webOS and to hear your feedback.

Lastly, I wish to express our sincere appreciation for your ongoing support for webOS and the many teams responsible for it here at HP. This is a particularly dynamic time in the mobile industry and sometimes tough decisions need to be made about not only what to do, but also what not to do. This has been one of those times. Together with our great webOS developer community, we are confident that we will meet the challenges ahead and build momentum for optimal success.

We will be communicating with you frequently over the next few weeks and we look forward to hearing from you throughout this process.

Thanks for your support

Richard Kerris
VP webOS Developer Relations

And then there’s this, discovered at a hot local dev company yesterday…

Posted by & filed under Apple Watching, User Experience.

Bloomberg News (via Ars Technica & Daring Fireball):

Intel Opens $300 Million Fund to Back ‘Ultrabook’ Laptop Push.

Intel Corp., the world’s largest chipmaker, said it will invest as much as $300 million to back the development of technology to speed up the introduction of notebook computers it’s calling “ultrabooks”…

…Intel is trying to fire up demand by adding some of the features that have made Apple’s iPad tablet and its competitors a threat to personal-computer sales. In May Intel said the new computers will be less than an inch thick, have days of battery life on standby, start up in seconds and sell for less than $1,000. Intel is aiming to convert 40 percent of consumer laptops to the new category by the end of 2012.

Here’s how I see this playing out:

Hardware Vendors

Dell, Sony, Acer, and Lenovo will plan big announcements about their ultrabook plans (but no shipping product) in November, with product launches (still no shipping product) at CES 2012 (January 10th-13th). Products will ship between February and April 2012.

At least one of the announced products will feature a touch-screen interface and a physical keyboard, will run unmodified Windows apps, and will fall over every time someone touches the screen.

At least one of the announced products will be a touch-screen tablet, running unmodified Windows 7, bundled with a dockable keyboard that you have to find every time you want to type.


Microsoft’s PR machine will start making faulty ignition noises in September, get into first gear in November, and get ready to make a big splash at at the CES keynote on the 10th.

Steve Ballmer will choose the tablet with dockable keyboard as his hero exemplar of What This All Means. [Five bucks says he will manage to lose the keyboard during his presentation.]

…And Apple

On Monday, January 9th, 2012, Steve Jobs will announce that Apple’s entire laptop line has now been redesigned along the lines of the MacBook Air, and new slimline MacBook Pros are shipping in early February. Surprise: Users will be able to swap out the built-in SSD drives.

This will officially mark the end of non-Flash storage media built into Apple’s laptop line. Users may purchase inexpensive external drives for old-fashioned rotating media. Thanks for coming…

Oh, and one more thing: Blu-Ray is still a bag of hurt.

Posted by & filed under Uncategorized.

One of the worst mistakes any company can make is fracturing their product line in the vain hope of attracting and capturing new audience segments. Doing so almost inevitably causes both the company and the company’s customers to lose sight of the company’s strengths.

One of the ways you can fracture a product line is to treat every SKU in your corporate database as an individual product, give each one a unique name, and use those unique names in the design of your advertising.


Example #1: Apple, Before the Jobsian Restoration

Between 1992 and 1996, Apple fielded no less than seventy-five different computers across fourteen product lines. Some of these allegedly different computers differed only in the bundled software. Others, in different product lines, differed only in the nameplate on the front.

Confusion reigned supreme. It nearly killed the company.

Upon his return, Steve Jobs started pruning product SKUs as fast as possible, and by 1998, he’d famously pared Apple down to a 2×2 product matrix. Laptop versus desktop on one axis, consumer vs pro on the other. Four products. (Three actually, because they weren’t ready to unveil the iBook yet.)

Now, the real magic behind this maneuver wasn’t that Apple settled on four products. It was that Apple stopped attaching dozens of meaningless product names to different configurations of the same computer.

Customers could still order different configurations, which in turn would be recorded under different SKUs, but nobody bothered to apply a grandiose name to each configuration.


Example 2: An American Mobile Carrier, Après le Déluge

One and only one of our ‘Big 4’ mobile phone carriers presents its phone plans as monolithic and utterly confusing conglomerations of features. In order to get something close to your desired plan features, you have to compare 19 named plans (including a new one that beats the other 18 like a drum…until you read the fine print on its limited ‘Unlimited’ data cap).

Some plans offering NO high-speed data cost the same as those with a low amount of data. Some two-line family plans with shared minutes cost more than two individual unlimited-minute plans with the same ‘unlimited’ limited data allowance. (I’m tempted to think that this kind of insanity is intended to drive shoppers to choose costlier plans that don’t have these fairly obvious inequities.)

Meanwhile, all the other mobile carriers sell three to five plans differentiated only by minutes of talk time, treating text messaging and data plans as a la carte add-ons. No bundling, no unwanted features, and only mild confusion.

And soon, the US line of national mobile carriers will be reduced from 4 to 3, thereby solving the problem.

 The Lesson

Sell the razor. The blades will take care of themselves.

Posted by & filed under Uncategorized.

A basic design for unobtrusive mobile way finding

Back in 2008, I was following Rachel Hinman’s grand mobile design experiment, 90 Mobiles in 90 Days [images also available on Rachel’s Flickr page].

One of her designs involving waving a mobile phone back and forth like a radio tracking device, with the phone signaling when it was pointing along the azimuth to the desired location.

I thought the idea might be improved by a less overt scanning behavior. After all, standing in the street and waving a shiny phone back and forth is Not Recommended Behavior in a number of situations.

This sketch was my first iteration: moving one’s thumb back and forth over the device screen, with a quiet vibration signaling the relative direction of the desired location.

On further reflection, in any built environment (roads, streets, cities), turn-by-turn walking directions would be preferable to mere compass directions.

This breaks down into to two sub-problems: outdoor and indoor navigation. Indoor navigation is by far the harder problem.


This part is fairly simple in a world of detailed city maps. All we need is a five-position arc of pointers in front of the user, plus haptic signals for New Turn, Behind You, At Destination and Not Found.

After that, we just orient the user to a map service’s walking path, and alert them when they have to discover a new direction at a waypoint.

This will get them to the building they want. But how do we help them find their way to an office in a 40-story building?


We’re already seeing mobile apps that provide detailed, multi-floor maps of malls across the US. These provide obvious benefits for all tenants of the malls and their current and potential customers. Mall businesses are almost entirely intended as walk-in businesses, so the more traffic the better.

Can the same business mappings be applied to large public and private buildings?

Would the same balance of discoverability benefits apply to non-retail business tenants of office buildings?

Are there countervailing disadvantages in publicizing certain business locations?

[ Haptic Wayfinding on Flickr ]

[ Flickr tags ‘wayfinding‘, ‘haptic‘ ]