02/15/18

Let’s talk predictive science

written by Martine Farrugia
Let’s talk predictive science

In a world where we’re bombarded with information at every angle, it’s getting harder for brands to cut through the noise. The traditional behavioural model we’re all familiar with is starting to feel a little stereotypical. Predictive science goes much deeper. It’s about why people like things, join things, vote for certain things – and, in short – behave the way they do. It deals with the subconscious part of decision-making: our instincts.

We’re really interested in all of this. So, we work in partnership with Bob Raleigh (founder of PathSight Predictive Science), to develop brands and collateral that eliminate ambivalence.

It’s all about clusters

Lengthy research has found that people fall into 18 different behavioural groups. And each of these groups belongs to one of five broader clusters. Each cluster represents different characteristics, different ‘triggers’, and different likes and dislikes on an instinctual level.

Compare this to the more common demographic and behavioural segmentation model. According to this, 18- to 35-year-olds like brunching and Netflix, while a 60-year-old could be cast as an active and adventuring grandparent. But your cluster is defined based on your inherent reactions to certain stimuli.

Cluster map
Which cluster do you think you belong to?

Traditional profiles can definitely be useful. But they don’t show the whole picture. Predictive science taps into the unnoticed, underlying data quietly residing in every complex system. And in each cause, political affiliation, social trait, and emotional response.

Based on research involving over 80,000 consumers, we’ve worked out how to tailor communications to suit the person receiving them. Gone are the days of one-size-fits-all. With slick document composition platforms readily available, you can tweak things to get the attention of different clusters, without diluting the meaning and weight of your message.

What does this mean for your business? Well, you can almost guarantee a better response rate by building deeper connections through your choice of words and images.

Predictive science flips the script

In the good old days, the first thing we’d define was the value proposition. From there came the message development, the choice of channel and then the audience. In true to outré form, we’re looking at things differently to reach that all important gut reaction.

We define the ‘who’, ‘why’ and ‘how’ first (based on the audience cluster make up). Then we choose the most appropriate channel and format. Finally, we create a message that makes customers think: “they just get me!”

Flipping the script horizontal diagram
Here’s how we turn things all the way around.

Who uses predictive science?

Brands all over the world are turning to predictive science to give them the engagement they crave. By triggering the part of the brain that makes people feel like they belong, we build robust relationships and offer reassurance. Which keeps customers coming back for more.

We’ve applied the PathSight model to a whole host of industries and channels. Take what we did for the New York Jets. As with many sports teams in the US, loyalty is fierce. But there was still room for improvement when it came to email engagement. PathSight profiled the fans to find out which clusters they fell into. We tweaked the language (to accommodate for different clusters) to position every fan as a part of the team. Since we’d created an innate sense of belonging, open rates and clickthroughs shot up by an average of 37%.

But it’s not just words that matter. Design has an important part to play too.

We produced three different creative campaigns for The Economist, all with the same end goal: increasing subscriptions. But we didn’t just send the same thing to everyone. Instead, we adapted the format, chose specific taxonomies and the right visual styles for each audience. Each comm pinpointed particular areas of differentiation in the messaging, to deliver a personalised experience for every reader.

How does it work at a project level?

For every predictive science project, you need a model and an engagement plan.

A model
We create a proprietary model of your market and use that as a guide to everything that follows. For example, we could examine your mailing list and split it up into groups. Or we could determine market drivers, product attributes, and motivators based on your audience.

An engagement plan
Once we have our model, we develop a strategic plan to put it into action. This involves building a narrative, creative, catalogue of words, images, and themes based on your needs.

We then embed personalised elements into whatever touchpoints you’re using to connect with customers. That could be something simple like emails and your telephone IVR system. Or it could be something fancier, like software, app or website. The possibilities are endless.

And the great thing about predictive science is that it’s future-proof. The traditional model of segmentation is based on past behaviours, while predictive science is future-focused. It’s constantly evolving based on new, monitored, interactions. Which means you get insights in real-time and can adapt quickly to whatever life throws at you (and your customers). So you can be sure you’re nailing the ‘gut reaction’ instead of relying on guesswork.

Project? Deadline? Coffee?

Let’s have a chat.

TALK TO US

For our address and contact details, click here.