How Big is Big Data?

The universe of “Big Data” seems to be expanding as rapidly as well, the actual universe. Data is now measured in zettabytes (ZB) which will eventually become yottabytes (YB), where just a few years ago, the gigabyte seemed enormous. A comparative scale of these Bytes can be seen here.

But how does big data actually help us and why do we need it? Everything we do adds to our digital footprint, data is created and stored from almost everything we do, from where we buy our coffee to the places we’ve travelled. This type of data can lead to exceptional results within marketing, but it has to be first transformed into useful information.

Having big data doesn’t automatically lead to great marketing. A common practice in today’s technology driven age is to gather huge amounts of data and find the patterns within it. But correlation seen in pattern is not always the same as cause-and-effect.

So, what’s the difference? In terms of big data and marketing, correlations between two variables can be seemingly related, and while the results may look good, you may not find the actual reason for changes in the data you’re studying. Even through the clearest numbers in sales data or website traffic, correlations still may not give you all the answers to which factors are influencing buyers.

So, is causation more important than correlation? Causation is the proven fact that one event is the single reason for another event occurring. But this can’t be seen through correlating data. For that, we have to dig deeper.

For example, a store may notice that their sales of cereal is high when it rains, and while it may be easy for marketers to make the assumption that weather patterns affect cereal sales, when we take a closer look, it’s more likely that there are other factors influencing cereal sales, such as a new type of cereal being released, or a special price for cereal that happened to be in a rainy week.

Though the correlation may be easy to make, weather patterns usually wont influence the sales of cereal, and it’s important to find causality for the spikes in sales before implementing a marketing plan based on the forecast of the rain.

While digging deeper to find causalities is incredibly important in determining factors which influence consumer behaviour, finding correlations in big data to base these cause-and-effect studies on is extremely useful.

Correlation or causation? Leave your vote in the comments.

References:

Testyotta (2018). Comparative Sizes of Bytes. [image] Available at: https://www.waterfordtechnologies.com/just-big-big-data/ [Accessed 6 Sep. 2018].

Ben, S. (2018). Correlation vs. Causation: Why marketers should stop saving for a rainy day – Fourth Source. [online] Fourth Source. Available at: http://www.fourthsource.com/general/correlation-vs-causation-marketers-stop-saving-rainy-day-19411 [Accessed 6 Sep. 2018].