How powerful would it be for marketers to know what their customers are going to do before they actually do it? To predict how they would react to certain messages and offers, or to intervene just before a customer loses loyalty and switches brands?
The medical community has been doing this for years — predicting the best course of action, based on family history and genetics. Wall Street too has used sophisticated algorithms to predict future market trends that guide high-risk trading decisions. Even some prison systems use the power of prediction to determine which prisoners should be paroled and which should serve out their full term. And no one can forget the use of predictive analytics in baseball, as shown by Brad Pitt in the movie Moneyball.
Everywhere we look, the concept of “predictive analytics” is taking hold and is now becoming the most powerful tool in marketers’ increasingly sophisticated toolboxes. While the idea of predicting customers’ behavior may seem like something out of the Jetsons-era, it is very much a reality today and is dramatically shaping the way marketers engage with their customers. This can be seen in something as seemingly ordinary as Netflix predicting which movies I may like based on what I have watched, or as advanced as Target predicting which of its customers are likely to be pregnant.
It’s no secret that brands have been collecting heaps of data on their customers in order to create robust profiles and personalized customer experiences. User profiles allow marketers to segment their audience on the simplest level by: geography, basic demographics and purchase history. Apart from actual sales and general profile data though, marketers – and their data scientists — are now analyzing things like social behavior, the stickiness of certain emails and promotional campaigns, and website activity to build living profiles. Each piece of customer data allows for advanced customer segmentation and goes into the statistical equation that predicts how likely action or inaction is.
By using sophisticated algorithms based on the results of historical and real-time campaigns, marketers can now predict with a high percentage accuracy things like:
- Who will purchase?
- How much are they willing to spend?
- What is the predicted lifetime value (LTV) of new customers?
- What is a customer’s propensity to buy?
- What is a customer’s propensity to engage?
- Who will become a brand advocate and share about their experience on social media channels?
- Who is likely to abandon for a competitor?
- Which offer will resonate most with certain audience segments?
By understanding this, marketers can better allocate resources to target specific groups and, with the most compelling message, drive the most profitable relationship. By harnessing predictive analytics, brands can also understand which customers are worth pursuing and “courting,” and which will not pay off in the long run. As marketing budgets continue to be stretched, this power of prediction will be incredibly essential.
Mobile has changed the ‘digital touchpoint’ landscape. Predictive analytics solutions would not have been as powerful pre-mobile, as the always-addressable customer did not exist. Now, brands need to discover the winning combinations of who to target, when to do it, with what type of message, on what channel – and how to react quickly if the outcome is not as predicted.