How modelling improves marketing

Share Twitter Logo Facebook Logo LinkedIn Logo Google Plus Logo

We’d all like to be able to predict the future: what our kids will grow up to be, where I will retire to, next week’s lottery numbers, etc. Unfortunately, we can’t help with those predictions - sorry!

But we can help you predict who will respond to your campaigns, the times your target audience is most responsive, the products or services they are most likely to purchase and which customers are at risk of lapsing.

Right Message, Right Person, Right Time

These ultimate goals are supported by so many areas of marketing including the Single Customer View and Marketing Automation.

To achieve them, marketers need to automate their offers or messages at key stages of the customer journey, based on what they know about the particular recipient. Analysis of your target audiences and predictive models will facilitate achievement of these important objectives.

Modelling will also help you to segment more effectively and avoid wasting budget by promoting the wrong products to the wrong person at the wrong time!

Failing to target and personalise your marketing efforts is not only inefficient, but also means that you run the real risk of appearing irrelevant to your customers and thereby reducing engagement.

You don’t need ‘big data’ for modelling

The term big data is so popular there is really no escaping it; but in truth your data doesn't always have to be big to be useful.

However much data you have on your customers and prospects - modelling to some level will almost always be possible.

The risk of making bad marketing decisions can be even higher if it is forcibly based on too much or weak data, to translate into actionable insights. A key part of effective modelling projects is identifying what data should be retained and what can be ignored.

Types and advantages of modelling

There are various types of modelling that can help improve your marketing results including:

  • Propensity - finding the most appropriate people to mail, based on their likelihood to respond or behave in a certain way (such as predicting product preferences);
  • Forecasting - predicting outcomes including ROI over a defined period, responses to campaigns, and – following testing – how much of a customer base to contact again;
  • Profiling - building a profile model of your best customers to help you understand their makeup and motivations, and identifying similar prospects for targeting.


To create and replicate the most effective campaigns possible you should identify:

  • Your best customers
  • What products or services they are likely to respond to
  • When they are most responsive
  • The most cost effective groups to contact
  • Your likely return on investment

The insight provided by modelling ultimately improves marketing efficiency, meaning you and your team carry out only the most effective activities. Save budget and time being spent on areas that are not going to bring the same amount of improvement in campaign outcomes.

Focus on those that will drive leads, enquiries and sales.


  • Have you used any of these modelling techniques?
  • How do you ensure you are being effective in your communications?
  • Did you consider your lack of ‘big-data’ a barrier to modelling?