The Strategy of Push Upgrades

This is the set of share curves for Chrome browsers.  Beautifully ordered, right?  Pretty much the idealized portrait of release early, release often.  So how does Google do it?

From Royal Pingdom (awesome analytics, btw):

Google Chrome handles its upgrades in a completely automated fashion, even for a completely new version of the browser. While other browsers will ask the user for approval to move from version 1 to version 2 (just an example), Chrome will just handle that in the background and move ahead with the upgrade.

If you think about it, this mirrors the way web apps work, i.e. updates go through to all users so everyone is using the same version of the software.

Genius.  The post focuses on the mechanics of upgrade push and the correlation between “pushy” browsers and people who love web analytics.  These charts could seed a million great thoughts, but I want to focus on the idea that changing the defaults can change the game. 

Check out IE’s share curve set:

I generally believe that the mark of a mature company is dominant legacy costs (real or abstract), and that’s exactly what we see here.  Despite prolonged effort, MSFT hasn’t been able to get its old (brand destroying) software off of the world’s computers.  IE6 is three generations and nearly a decade old, but killing it depends on every IE6 user out there independently choosing to go through the upgrade process.  Who wants to take that bet?  And don’t forget, many of the users running IE6 are doing so because they can’t risk an upgrade on their pirated copy of Windows (there’s another post in there about effective methods of deterring software piracy).  Meanwhile, as MSFT slowly watches its past starve its present, Google chose a different default upgrading behavior for Chrome, and with that single tweak ensured that it’ll never have to deal with the legacy costs of an intractable, negative brand equity browser product.  

Now that’s a strategic advantage.

Dealing with Unexpected Barriers

When purchasing decisions involve heterodox interests, sales difficulties increase exponentially. A perfect example is enterprise Gmail. Gmail is undeniably a better-designed email product than most companies use, but Google butted up against unanticipated resistance when selling Gmail to enterprises. It turns out that enterprise purchasing decisions are far more complex than a single users.

The IT guy at my work and I talk often about what it would take to move our firm from Lotus to Gmail (consensus: an act of god). He pointed me to this article from CIO: Why Enterprises Are Moving to Google Apps, Gmail. Here’s what I found most interesting:

Perhaps most significantly, at a Google Apps CIO roundtable event in San Francisco this week, Google announced that enterprise users of Google Apps could access Gmail through an Outlook client. The company hopes it will quell the protests by users who have become tethered to the desktop app and who, as a result, have sometimes hindered enterprise adoption of Google Apps.

“For me, it eliminates the last hurdle or mindset for letting go of [Microsoft] Exchange or the Exchange mentality,” said Bob Rudy, vice president and CIO of Avago, a semiconductor company that moved its employees over to Google Apps, during the event. “This will help with adoption.”

It turns out that people don’t necessarily want better email (at least not in the sense that Google initially thought).  Sometimes they just want a better version of their current user experience. Sometimes they even want something you never even imagined could be important.  From the consistently excellent IgnoretheCode (aka, Lukas Mathis):

If your users reject an improved user interface, you need to start out by figuring out exactly what motivates them to prefer the more complex solution.

Of course, there’s more at issue here than UI, but the general point remains. When people reject an objectively better solution (at least in your eyes), you better find out why.

Why not a ‘secure’ tag?

It occurs to me that I can store a lot of non-public things in gmail because I believe in the security of my password.  But anyone with access to my cellphone has access to my email, without a password.  So does that mean I believe a password secures my email digitally at least as much as possession secures my phone physically?  Given that I’m not a ninja, but also that I use secure passwords, I have to think my digital life is the more secure.  But I access my phone too often to make a PIN worthwhile—I’d get fed up with the barrier before one day was through. (I know this because I’ve tried. More than once)

So how can I keep my phone open and keep personal stuff in the cloud?  One solution would be the ability to tag a message “secure.”  The tag would require me to re-enter my password before viewing that message (like how my ATM re-asks for my PIN when I go to get cash after I deposit).