Saturday, October 04, 2008

What Google knows about you? And why it will rule the world?

I didn't want to put the following statement at end of post, wanted to get the message out first.

"I'd complete control over the personal data storage settings in Google's products and still trust Google more than any other company to store my personal information."

With that statement out of the way, now to the main topic.

Google has released so many products that it's really tough to keep rack of all of them, even if you are an ardent Google fanboy. These products are feature rich and knowing all the features of all the products of Google is next to impossible for one individual.

This post is about one such feature: "Search History" (not sure if it's a product).

Like any Google fanboy I tried out "Search history" immediately when it was launched, played with it and after sometime completely forgot about it. The "Search history" has been renamed as "Web History".

To access /activate "Web History" you need to visit this page. Google account owners can access through 'My Account" after "Sign in". "Web History" is listed under "My Products".

Recently I was checking some other stuff in "My Accounts" and accidentally checked "Web History". I was accessing "Web history" after a period of 3 1/2 years!. Though "Web history" was released only on 4/19/2007, I'd data since 2005 as I'd previously signed up for "Search History".

Here are some of my personal data as recorded by "Web Hisory"

  • I'd done 3,040 Google searches in 3.5 years. I expected the numbers to be high as I'm hooked to internet most of the time and Google is my default search engine.

So, I started to dig into the data. I picked a month (September) and checked number of searches done over last 4 years. The following graphs shows the trend line.


What do you notice?
  • Sudden jump in 2008. In fact ~1500 out of total 3,040 searches were done in 2008 and still 3 more months go!

What is the Inference?
Does it mean I've become more clueless in 2008 and need Google to figure things out?

NOPE! My cluelessness is constant :-)

The unarguable reasons are : -
  • "Web history" can record data only if I'm logged in. Its only in 2008 I've started to use "Gmail" more regularly (logged into it most of the time).
  • Nobody will exclusively log in (Sign in) to do searches.
  • However if you are already logged into one of the Google services, cookies are set to automatically sign in users to search service.
  • Being signed in is very important. (I'll come back to this point)

The coolest and scariest feature within "Web History" is the "trend" feature. This feature gives information like : -
  • Shows My top 10 queries, top clicks, sites in last 1 year, 30 days and 7 days.
  • Tells me that I've searched more during week days, however this is changing, even during weekend my search usage is increasing.
  • I've done search during all hours of a day, from 0 to 24hrs. Semi insomniac!

Other information about me that surprised me.
  • I've just clicked only 4 advertisements (Google Ads). Certainly not a profitable customer!
  • Shows all videos I've seen
  • Lists News that I've followed in last 3 1/2 years
  • Traces Routes and Map searches I've done
  • Books I checked out.
  • Blogs that I've followed

I know about the popular Google's disclaimer "No human at Google, ever reads user data". So, I can't blame that a Google engineer might know about me. However Google's code (Algorithm) does surely know more about me than I've cared for.

With this data and advanced algorithm, Google's code will surely know : -
  • What is the best time to connect with me?
  • What is the best time to sell?
  • And what to sell to me?

User relevant Data is important, without this data, no matter how advanced the algorithm is, intelligent and useful services can't built.

Relevant user data can be collected only if the user is logged in. The personal data collected is the insurance against any new algorithm or massive data center.

The relevant user data is the 'positive network effect" similar to platform network effect that made Microsoft so powerful.

That is why Gmail is very important and central to Google's service. It enables automatic "sign in" for all its service. It is also equally important to have users always logged in so that user specific information can be collected all the time. That means user should not close Gmail, but let it run in the background.

Lets look at some of recent Google's initiative.
  • Gtalk in Gmail. It makes sense to have Gtalk built into Gmail. Always logged in users.
  • Google Chrome makes running rich application as long as needed without the browser hanging. This enables running "logged in" applications like Gmail as long as needed. If Gmail is logged in most of the time, more user specific information can be collected. (Can you guess why Chrome was built? ofcourse there can be many more reasons)

I'm not sure if there is any method to Google's product or feature release madness. All I can say is that Google power is growing greater than normal users or business can imagine.

Only one company can compete with Google - Microsoft. However the "Window" of opportunity is closing faster ;-)