If it seems to you as though the marketing industry has been discussing the opportunities and threats posed by big data for several years now, you won’t be alone. It’s easy to see why - big data is the fuel for predictive analytics, and effective predictive analytics offers companies potentially highly significant strategic advantages: from greater efficiency in marketing segmentation to improved customer experience.

However, having the best predictive analytics models in place is merely the end goal – first of all, the importance of embedding the practice of predictive analytics into the wider business cannot be understated. Embedment is a multi-stage process – the five most important components of which are detailed here.

  1. First of all… Marry your IT department. It might sound humorous, but a close relationship between Marketing and IT is vital - without it, embedding predictive analytics into daily working simply will not succeed. IT have to be true partners to the Marketing team to develop the kind of agile, test-and-learn approach required to bed in predictive analytics.
  2. Create an analytics maturity model for your business. Not only do such models help to communicate the wider strategic direction for predictive analytics, but they are also key in helping to chart progress and showing that the positive changes being made to embed predictive analytics are there for a reason. Additionally, an effective model should not simply make clear the scope of the embedment of predictive analytics, but should also be instrumental in helping the business to identify practical opportunities for its deployment.
  3. Be prepared to embed in stages. It’s wise to begin experimenting on a small scale, ensuring that wins can be measured and communicated across the wider organisation. An iterative approach, allowing you to prove returns at every stage, should be maintained as predictive analytics gradually intertwines itself to a greater degree with the marketing team’s core activities and objectives.
  4. Carefully weigh up customer outcomes vs. business outcomes. It’s imperative not to overlook the end-customer’s interests throughout the embedment process. Understand early on that the results of using predictive analytics for the customers’ experience of the company must be considered, and never allow the customer viewpoint to fall out of the thought process (even if in some instances achieving organisational objectives must inevitably override those of the customer). This is where forming strong bonds with wider teams within the marketing department can pay dividends – liaise throughout with colleagues in customer experience and insight, and not only will predictive analytics projects benefit from their expertise, but the customer experience teams will gain a much more granular understanding of the wizardry the analysts may appear to be performing. Liaising with wider teams is also vital in ensuring that external colleagues see the potential of the work and can begin to generate ideas for where predictive analytics can help them.
  5. Be prepared to monitor and to maintain your models. You should look to monitor every month to assess how your models are performing, by testing actual outcomes against predictions and checking for the models’ validity. Once a model’s success rate drops below a certain level, it’s time for maintenance – creating a new model to better predict results. Even then, the new model must be checked against the old – if the challenger model’s results aren’t better than those of the former champion model, then there’s more work still to do. All of this is time-intensive and can appear offputtingly laborious, but unless your organisation is prepared to invest this degree of rigorousness into predictive analytics then it would be foolish to embark upon the initiative in the first place, as the results will be far less likely to achieve the requisite degree of accuracy on a longer-term basis.

To sum up, embedding predictive analytics into your marketing department’s core activities is very doable, but requires careful handling and consideration along the way. If successful, the dividends are potentially enormous in terms of strategic advantage and a more informed decision-making capacity – and even if your business isn’t thinking about predictive analytics yet, the chances are that the initiative is already on your competitors’ radar…

Jennifer Gill

Senior Consultant at Comet Global Consulting

Jenni is a Senior Consultant for Comet’s UK Strategy and Insights Team and holds an MSc in Creative Advertising. She has led Customer Experience and Digital Strategy projects across a diverse range of industries, from media to travel to financial services, and past clients include the BBC, First Group, Lloyds Bank, Aegon and AXA. Currently, Jenni is in charge of the implementation of decisioning in the B2B space at Standard Life up in Edinburgh, and in her spare time acts as a freelance writer and editor.