What do big data and carrier pigeons have in common?

August 5, 2014

Not a lot, but they both do 'help' deliver accurate personalised messages…

Big data is certainly big. In fact, it's revolutionised the way brands create their strategies and predictions. However, not every big data story is a successful one. Despite many marketers' efforts, despite the dashboards, the predictive analytics, the data providers and the various other insightful tools, sometimes it just doesn't work. But why not? I came across an interesting blog post on the iMediaConnection website which discussed just this.

When big data is accurate, it's great. The article draws comparisons to meteorologists: they have been able to increasingly make accurate predictions over the last 20 years or so because they understand not only the collected data, but also the context behind the data. Marketers are in the same position.

We have access to a never-ending stream of digital channels to explore - email, social media, display advertising - to name but a few. There are numerous touch points to analyse.

Using cross-channel analytics, we can even understand how these ecosystems relate to an integrated marketing strategy and have in-depth knowledge of our audiences thanks to first and third party data. With this plethora of data, we have a pretty good idea of how our campaigns will play out.

The article then points out what happens when big data fails; even when it comes to weather forecasting, earthquake prediction has made little progress over the years, says iMediaConnection.

In relation to marketing, we face a similar challenge when it comes to channels such as referral traffic and organic search - data collected on these channels only represents the end product. The factors that determine these outcomes are not easy to work out.

We may be control freaks, but there are some things that are just completely out of our control - for example, how can we predict how a publication will react to a press release? How do we know about other news being released at the same time? Accurate prediction in these kinds of situations is verging on impossible.

We need to understand that, while big data can take our marketing efforts to a whole new level, it is not the be-all-and-end-all. It's great for exposing patterns in areas that are easily measured and well understood, such as email and digital advertising. But there are some areas where it won't - or can't - have a meaningful impact. It's all about finding what will have the most meaningful impact on your brand and working within big data's limitations.