Saturday, February 18, 2012

Fixing The Failure that is G+

Don't believe Google's hype:  Google Plus is a huge flop.

You need to look no further than my own timeline to see it.  I have 287 people in my circles that I follow. Most are extremely active on other networks - Facebook, Twitter, etc..  A large number of them have Android phones, so are presumably automatically uploading their photos and videos for sharing as I do.

And yet, with those 287 people I follow, there is exactly one update in my timeline this afternoon that's not from myself sharing with other people.

On Facebook, I have 500 friends and have received dozens of updates during that time.  On Twitter, I follow 187 people--less people than on G+--and have two dozen updates.

The stat that Google keeps giving is "number of people on G+".  Sorry, number of people on G+ doesn't matter.  Everyone with a Gmail or Picasa account is basically being opted in, and no one uses it.  Most of my IRL Google friends don't even use it.

G+ is a huge failure.  The biggest social network no one uses.  It's so obvious it's not even funny.  They've been trying to do what Twitter has done and what Facebook has done -- i.e. being feature equivalent.  G+ is proof that that plan doesn't work.  How many times has Microsoft tried this strategy and failed? (Bing?  Zune?)

And the saddest part is, I actually wish people would use it.  I prefer G+ to Facebook for the most part because I'm a Google-fiend.  I prefer the mobile app by far (the Facebook app is all kinds of jacked up on my Galaxy Nexus).  But the question is, how does Google make it right?

Two answers:

First, stop taking away the services we like and folding them into G+.  We liked PicasaWeb.  But now that's it's gone, my family entirely switched to using Smugmug.  Making this change had the opposite effect than what Google intended.

The second answer is focus on the things you have that are novel.  The most novel feature of G+ so far is Hangouts.  It's what draws people to G+ more than anything:  Hang out with Obama, etc.  Figure out how to rally around this with the audience you have.  What about tying it into Google Voice?  What about making the quality crazy-high and taking on Cisco?  What about doing presentations with it?  THINK.  Think of things you can do to rally around this, because no one is looking at people's non-updates on the thing and this is the ONLY novel thing G+ has.

The other thing Google shouldn't be doing is putting G+ in everyone's face every day -- at least for now.  It's almost counter-productive when people are sign up and then never use something.   In their minds, they tried it, didn't like it, and getting them to come back is even harder than having them come the first time.

In the mean time, I'll have to suffer through Facebook's completely broken Ice Cream Sandwich experience.

[Update]:  As if you need any more evidence, Google has now added "What's hot on G+" entries to my timeline that I didn't subscribe to.  Because there's such a dearth of updates, they have to throw in stuff.

Tuesday, February 14, 2012

Big Data / Small Data

"Big Data" is in the New York Times.  Run for your lives.

The first thing I ask a candidate who says they want to work on "Big Data" is "What Constitutes Big Data?"

They'll throw out a number like "10GB" or "5PB".  Both are ridiculous answers.  I'll let you in on a secret:  there is no right answer to this.   People answer this question based on personal experience.  They'll have just been at a company with a  300GB SQL Server installation is creaking under its weight, so 1TB becomes "Big Data" in their mind.

There are really two axes to look at:

  • Size of the data in terms of disk or memory
  • Complexity of analyzing that data [EDIT:  the complexity of things you want to know]
Most problems don't challenge both of these axes.  Furthermore, most people confuse the two.

Size is not usually a problem.  As Ted Dziuba points out eloquently as usual, most processing is not complex, even if the data is large in size.  If you need to know if treatment A or treatment B of your Facebook game sold more virtual junk, it's just not that hard to figure that out.  You can use grep, cut, sort and uniq to figure that out.

I jokingly posted to Twitter today that I'm working on my Small Data skills for this reason.  I'm helping out with some analytics for some content optimizations, but am using pure unix and a simple Python script to pull it together.  No Hadoop.  No Map Reduce.  Just not needed here.

Complexity is really the major problem to tackle.  Try making some sensible decision based on the information at hand.  The size of the data collected is a crutch in order to avoid requiring real intelligence.  A "machine learning" algorithm will require 10,000 keyword searches to deduce what kind of person you are, but a human brain might require just one keyword.  Or even just by lookin' at you. 

Bottom line:  learn to differentiate "big" data from just data.  Chances are that you're probably working with regular, boring, small data.  Embrace traditional data marts if you have to for historical analytics.  Then just use the Taco Bell techniques that Ted Dziuba describes above.