CAL L FO R TO P I C S
Big data dreams,
small data reality
Because of this abundance
of data, many “best
practices” rely on making
data-based decisions.
Yet there are still many
situations where we
unfortunately do not have
sufficient data to make
such decisions.
There is no doubt that we live in an era of dig data.
We seem to have mountains of data about everything
from business operations to customer behaviors,
from personal health to global disease outbreaks. Because of this abundance of data, many “best practices” rely on making data-based decisions. Yet there
are still many situations where we unfortunately do
not have sufficient data to make such decisions. So
what do you do when you have big data dreams
but a small data reality? This is the focus of a new
Analytics magazine column, debuting in the March/
April 2014 issue. Through this column we will explore
how to deal with situations where we need data, but
it’s limited or nonexistent.
AT SCALE: BIG DATA
BY BRIAN LEWIS
22
|
I’ll give a specific example from my current work
as chief data scientist at Fractal Sciences, a marketing
automation software company that optimizes digital advertising and engagement (think Facebook and Twitter
ads, but a lot more). Without giving too much away and
without getting too technical, Fractal’s ad optimization
algorithm is based in part on a proprietary feedback loop
that uses prior ad campaign data to automatically predict, recommend, create and target new ads in order to
maximize a customer’s ROI for their advertising spend.
As a result, our customers’ ad campaign results get better and better the more they use our product.
A N A LY T I C S - M A G A Z I N E . O R G
W W W. I N F O R M S . O R G