Skyfire, App Launch & The Accidental Genius of Server Crashing

(A version of this post previously appeared  in Digital Pivot)

In case you hadn't noticed, there's a whole new way to get media headlines for your software product or app, make your competitors cry, get your VCs  to wet their pants with excitement, and in the midst of technical chaos, gain a whole new crop of customers on Opening Day.

It’s simple. On opening day of your app launch, have your servers go down.

Here’s the flow logic: Consumers want your product, they can’t get it , they get angry, they tweet, the media reports your servers are crashing, and more consumers arrive to download what they can’t get.

Hot dang, your company didn't forecast the server load requirements accurately, and for this technical error, the market rewards you with valuable public relations and social networking buzz.

Is there an app company out there that doesn’t want this hot mess?

Technically speaking, server crashing on Opening Day is nothing new. It’s been going on since Web 1.0 and the high-flying days of Amazon, AOL, and eBay. While I am not in any way advocating you crash your servers purposefully, it is my perception that the phenomenon is not only increasing in frequency, but taking up more “air time” during a product roll-out day.

Think I’m exaggerating? Check out my collection of media-significant server crashes and service outages.

Crush-It-by-Crashing-It Hall of Fame 

My Hall of Fame includes Skyfire, Angry Birds/GetJar, FlipBoard and, of course, Twitter.

Two weeks ago, app maker SkyFire released a new version of their mobile browser, noteworthy as the first browser  to support Adobe Flash for iPhoners. Was there much demand? Oh yes. Were their servers unable to service all the new users? Yes.  So SkyFire pulled the app off the market the very first day, citing they were " taking time to upgrade their servers".

What's interesting is that the way they pulled it off (no pun intended) was to post signs all over the Web saying they were “sold out.” This is, of course, an impossibility for a mobile app, but it’s wonderful market repositioning (aka turning a fail into a feature).

Then there's Angry Birds...or more properly, their distributor, GetJar.

 

Angry Birds & GetJar

Much of  SkyFire’s “we’re too popular for our bandwidth” bravado seemed aimed at one-upping the wildly popular mobile game, Angry Birds.

It was only some two weeks ago that mobile game app maker Rovio released a new version of their app Angry Birds and independent app store GetJar’s servers went down.


Looking into the GetJar blog gives a feel for their  angst.

 

Eight to 10 times higher peak loads?

More than 10 million downloads a day?

At one point GetJar’s Twitter account reported it had 90,000 downloads "in seconds." Yowza. That's technical agony but marketing ecstasy, right?

FlipBoard, the iPad App

 
Our next server crash celebrity is Flipboard, the personalized social magazine that works on iPad. On opening day, with glowing write-ups from the likes of The Wall St. Journal and  Robert Scoble, consumer interest was piqued. However, as users downloaded the app and attempted to use the service’s social networking features, it became clear that the servers were overloaded.

With the technical problems continuing into the next day, the media and popular blog news headlines turned cynical.

Twitter

No list of crashnauts would be complete without Twitter. Though not an app, but a social networking platform, their numerous service outages are legendary. In some ways, Twitter legitimized the server-outage-as-popularity indicator: After all, the Fail Whale became a much beloved icon, having its own Twitter group and its own fan page.

Ironically, this month PC Magazine reports that in September Twitter experienced the best up- time in three years. However, doesn’t that make us Twitter users a little sad too? Gone are the days when we all had to fight our way into Twitter?

That’s what’s really what’s going on here -- whether intentional or not. It turns out scarcity has its virtues.

Server Crashing and the Scarcity Principle

We’re all familiar with the scarcity principle from retail store sales. (Think “while supplies last" or “this week only!”)

It seems if something is rare or only available for a limited time, humans somehow will find it more desirable.

Recognizing this, some companies overtly control  the supply of their goods to put a lever on demand. The classic case is De Beers. Buying an immense quantity of diamonds on the world market, the company releases their supply in highly limited amounts, thereby maintaining high desirability and a high price.

Are some software companies “leveraging demand” by creating  scarcity on  their Opening Day?

"Accidental" Genius?

Why do otherwise highly tech-competent companies (often venture-capital rich) fail to forecast their server load?

If you didn’t know better, you might suspect it to be a specific activity called for in the Marketing Plan. (At least one blogger suggested Microsoft was guilty of this during search engine Bing’s service issues in late 2009.)

But we do know better. Clearly, angering customers by announcing a marketing launch and not having product or service available is not a good practice. Even more clearly, if you are charging for your downloaded products (or like GetJar, if you are distributing for a popular app), it’s not survivor behavior.
 
In this age of freemium, one must wonder that when the spotlights are on, does just a little bit of downtime do just a little bit of good? Is it too nefarious to ask: Has any behavioral cyber-economist studied whether bringing a server down for one hour on opening day spurs on incremental demand over the next 10 hours?

What do you think? Did I miss any good crushing crash stories?

Have companies just become clever at repositioning an execution flaw, or is the Fail Whale sometimes winking at us?