Back to news & views

In a world of distracting big data, how do you get anyone to pay attention?

Big Data analytics promises to unlock the value of your information. However, Big Data itself creates a much simpler, but just as challenging, problem. In a world of information overload, how do keep your stakeholders and customers engaged long enough to land your message or sell your product? Today, your ability to grab and retain attention defines your success. New research has coined the term the Attention Economy and tools are starting to uncover what matters.

Firstly, we will cover two parts of the problem.

1. The Internet is huge and still growing

One prediction suggests there will be more bits by 2020 than there are stars in the Universe. That is 1022, or as Brian Cox might loquaciously state, 10 thousand million million million. YouTube has more video uploaded monthly than the combined US TV networks have created in their 6o year history. And this is from a company that outsources all its content creation and pays nothing for it. The web therefore is a paradox. You have a microphone (a blog, Twitter, Facebook, YouTube, etc. account) and an amplifier (a network to post to and then “likes/reposts/retweets”) but so does everyone else. This means you can shout as loudly as you like, but it’s difficult to hear above all the noise.

2. The brain has a limited capacity for information processing

Susan Greenfield and Nicholas Carr hold controversial views. They believe the volume of Internet content is damaging our brains by reducing our ability to concentrate. However, what is uncontroversial is that the human mind has a finite attention span. This is not just the time we have in a day to think. Our brains come with a limited built-in amount of working memory, a little like RAM. If it becomes overloaded, it stops working as well. In a world where there is so much competing for our attention, this is therefore likely to happen more frequently. The figures show that you have around 10 seconds to grab someone’s attention. Perhaps because of information overload, people are ruthless at weeding out irrelevant or boring content.

What does the new science say about allocation of attention?

Bernardo Huberman, a former physicist and now a director of HP’s Social Media Lab, has published some interesting research. On the Internet, measuring attention is relatively easy. Initially, he examined two angles: novelty (newness of content) and popularity (followers, clicks, etc.). The researchers discovered a predictable log normal distribution to describe how social networks allocate attention.  A further rule shows how novelty of content falls off over time (all content has a half-life). Most web sites have a simple rule for listing their articles. Either, they place the newest at the top or the most popular. However, a combination of the two maximises the time spent on a web site. This change increased the number of page views by 17%.

Huberman’s team also examined a factor called “influence”. Some social networkers have large direct networks, but their network is not active. The followers were not “liking” or “retweeting” their content and so the originator was not taking advantage of their followers’ networks. Therefore, someone with an active smaller network can actually have more influence.

As well as showing how to prioritise content, Huberman’s team has created software that predicts future trends based on social media activity. In one application, they tracked and analysed Twitter activity ahead of film launches. From this and looking at historical data, the work accurately predicted box office revenues for new films. Their model has also predicted the size of demonstrations for police forces based on activity ahead of a march.

What does this mean for your business?  There are four golden rules.

  • You have 10 seconds to grab attention. So your initial message must be easy to read and simple to grasp.
  • Personalisation improves engagement by 20-30%. Gather relevant customer data and target your audience based on this information.
  • Interest in messages falls off quickly, all communications has a half-life. Therefore, the frequency of your communication must be linked to this half-life.
  • Active networks are better than large networks. Some organisations actually buy followers, but this is a fallacy. It is better to target messages to the interests of your active members. You then benefit from the multiplier effect of their networks.

The latest thinking therefore highlights that big data solutions are not enough. They may be able to suggest what people might want to buy next but if your content does not grab and hold attention, nobody will buy anything. What is even more interesting is that you can use previews of new products, communications or events to test their potential. You can then analyse the initial social media response to predict the future response to that message or product.

This article was inspired by DMW’s experience of:

  • Project managing and designing a large-scale data analytics capability for a government client.
  • Rebuilding an interactive marketing web site for a UK software company that is a leader in cross platform personalised content.
  • And finally by a Bernardo Huberman lecture.