Ready for Prescriptive Analytics? (A: Probably Not!)

Yesterday I joined a webinar hosted by Laura Ramos of Forrester Research in conjunction with Idio on the topic of Predictive Analytics. Predictive Analytics is not a particularly new term – I first bumped into it several years ago with a case study where the social media analysis tools were being used to predict the outbreak of ‘flu epidemics across the US (turns out that before people retire their sick bed the last thing they do is to tweet their friends to tell them all that they’re feeling a bit snotty). The business benefit of this was that the analysis allowed you to predict an outbreak several days before people made an appointment with their doctor. For a drugs company or a medical supplies company that could provide vital intelligence to ensure that supplies were prioritised to the most-needed locations.

 

Yesterday’s presentation included several examples of how new technology innovations can be used to deliver useful predictions – from internet-connected tractors able to maximise crop yields, to  Marketing Automation systems that help suggest next actions via lead scoring.

But something was bothering me. While some of these newer insights are exciting, most of my clients are struggling with more fundamental issues of getting basic customer information that they can rely on. Indeed Gartner suggest that only 13% of companies make extensive use of predictive analytics. And in their 2015 data quality benchmark, Experian found that the level of inaccurate data in our databases is rising. 92% of those who deem their contact data to be be essential to marketing success admit to inaccuracies. On average they identify that 23% of their data is at fault.

So while these newer shiny methodologies seem alluring, my recommendation is that we take a stepwise approach and focus on some of the basics first. Unless we build and maintain some sanity into our basic contact data, we are going to miss out on opportunities to fully leverage the more sophisticated approach.

Start with the Basics

Ready for Prescriptive Analytics? (A: Probably Not!) Purple Salix

A typical client data maturity journey would start with getting a grip on our basic customer contact data. A regular data audit is the obvious place to start here (as I’ve written previously). Experian’s research found that  63% of companies lack a coherent, centralised approach to their data quality strategy. And even when there is some centralisation, many departments have their own data quality strategy.

Beyond client contact data comes Firmographic data – data that provides insight into the industry within which the company operates, where it is located, how well it is performing, etc. This is an essential element to be able to segment for more targeted industry messages.

What customers have bought from you (or from your competitors) can provide powerful insights about additional offers that you should be making (complimentary offerings, end of warranty services etc) but are frequently not well integrated into the data set available to marketing. 

Most of us have at least primitive engagement insight for our email campaigns (who opened, clicked on links etc). With the increasing adoption of Marketing Automation we are able to connect this insight with a clients web and social behaviour, this providing a much greater insight into both their behaviour and their interest areas.

The road to Insight Driven Marketing

Ready for Prescriptive Analytics? (A: Probably Not!) Purple Salix

As one travels along this path to a more customer insight-driven approach, we see our focus evolving from trying to answer simple “rear view mirror” analytics – who were they, and why were they engaging? – to a more forward looking predictive approach. The ultimately we can even move beyond prediction to a more prescriptive approach. An example might be the following: if we know that clients who don’t use our service extensively in the second half of their contract tend not to renew, how can we encourage greater usage during that critical period so that renewal rates increase?

So, as ever, my message is simple. Accept that your client data is something of a mess. What’s important is the actions you take now to address that, so that one day you might be able to take full advantage of these more sophisticated techniques. But unless you have a coherent strategy for all you client data your risk creating more confusion when what you’re seeking is clarity and insight.

Are you one of the happy few exploiting predictive analytics in your marketing? How did you get there?

Leave a Reply

Your email address will not be published. Required fields are marked *