Big Data, Little Picture?

We can use Big Data in creative and impactful ways. Even classic industries can benefit from what people search on the internet. Be careful though. Incredible tools connected to gargantuan amounts of data can be misleading if the wrong metric or paramaters are looked at. Conclusions might be drawn that don’t paint an accurate picture if only one metric or possibility is considered. It’s like if an artist tried painting a landscape in only one brushstroke or color of paint. Adding search metrics in Google Trends and choosing relevant parameters was imperative in the article linked above. Having an incredible amount of data and customization at our fingertips but not utilizing it would be like driving a Corvette but never shifting out of first gear.  Let’s look at how we could have gotten it wrong in the first article of this series.

The data clearly shows when spikes in bookstore searches occur. If I owned a bookstore I could have quickly decided big data was great and put all of my advertising into August and December. I’d have been making a huge mistake, unless I’m the rare private owner of a college bookstore. It isn’t until looking at the related search metric that I see those spikes are tied to college bookstores and the timeframes match the start of new semesters. Metrics analyzed in unison created a greater understanding of what was going on. I was then able to move on to more relevant data.

I started by adjusting my timeframe for looking at the data. As a consumer I don't think about buying books in 12 month increments so it probably doesn't make sense for looking at the millions of Google users. I narrowed it down to 90 days, then 30 days, but it wasn’t until I looked at a seven day period that a trend appeared. There were consistent times that people were looking for books to read. I then began changing my search terms to see if I could find a more significant dataset. I ended with the terms and timeframe shown below that answered my question about when to connect with people already looking for books.

Now that we've walked though some of the process I went through to gain the insight in my first article you can see the potential for wrong turns. Looking at the big picture from the wrong angle can distort the image. It will take practice and work to learn how to shift big data into the right gear for you. I hope I’ve demonstrated for you that the work is worth it.

by Dexter Nordhues