There’s a fine line that separates analytics from BI. I don’t sure it even exists.
I remember this course in the university. One that had our professor trying to explain the quadratic numbers field. It had something to do with having a number in between every two other numbers. And then complaining that somehow this doesn’t work on his bathtub water dial – he could never get the temperature of the water just right.
This to me, is like finding the delineation point between analytics and BI.
BI – Business Intelligence. Put simply, this is the ability to show dashboards and reports to managers about how the business operates. Collect information across the enterprise, aggregate it and then display it on a graph.
Analytics starts in much the same manner, but somehow it does things that are more sophisticated. What that sophistication is – there’s a question for you.
Here are a few examples. Can you guess in each one if it relates to analytics or BI?
Google Analytics
Is it analytics or pure BI?
We let Google collect visitor counts of people coming to our website. And then we just view it. We look at the history. We track specific parameters. We slice and dice the data from different angles.
But is there really any analytics going on in there? Something that merits its name?
Hard to say.
Autocomplete of Google search
When Google starts suggesting search term from the first characters I click – is it BI or analytics?
It is predicting what I most probably want. And it does that from analyzing huge quantities of information. Easy to call it analytics.
Autocomplete on Facebook
What about autocomplete on Facebook?
When you search for something, Facebook starts showing possible results – names of people in your circle, pages you follow.
Is this analytics? It seems like a simple search. Much simpler than the autocomplete of a Google search. But is this the difference then? The amount of data being searched to bring a useful suggestion?
Autocomplete on a phone directory
And the autocomplete on my phone’s dialer?
It searches for contacts based on the number I click. It looks for prefixes first. It may show me numbers I dial more frequently first in such searches.
But can we call it analytics? And if we can’t, then what’s the real difference between this and the autocomplete on Facebook?
And what if I decide to mark better results based on the time of day the search is made correlated to previous calls the user made? Or based on his location? Does it make it into analytics?
–
Somewhere along this line, I lost track of what’s analytics and what isn’t. What do you think? Which of the above can be defined as analytics?