Because customers are constantly connected, they leave data footprints everywhere. A streaming flow of digital clues about theirs activities, behavior and needs.

 

If you could make all these clues speak and tell the story that is behind data, then it will be easier to deliver the right message, propose the right offer or deliver the good level of service. Once you really understand your customers, all become easier.

 

The old sort of thinking about demographics using age and profession is necessary but not sufficient. It’s more efficient to use the digital behavior because it is much more predictive of who you are and what you will actually need and buy.

 

 

These Big Data streams come from all types of customers devices :

  • Web with search, visits, transactions and next product they might be buying.
  • Mobile devices with app liked, user location and the next place they might be going.
  • Phone with the main reason for call and emotion.
  • Social with main interests and level of influence.
  • The internet of things telling a lot about the customer context.

 

The challenge is to collect and turn all these streams of customer data into valuable decisions and actions for each of your customer segment :

  • A website visitor interested in a product and about to make the big decision.
  • A pedestrian near to your agency where a checkbook is available.
  • A motorist in trouble expecting support to get back on the road asap.
  • A passenger not happy and willing to share that attitude if nothing is done quickly.

 

  Customer-centric data model

Alone each piece of data isn’t so valuable. But if you are able to aggregate all these datasets, you will see the whole customer experience.

 

Once you recognise the patterns in the data, you will be able to act quickly with the Next Best Action and even proactively making a special offer before the visitor leaves, sending an SMS as a reminder, to make a call before being in trouble, sharing a direct message on what to do when your luggage goes missing…

NBA

Five years ago, there was a lot of focus on feature construction and which algorithm to use. Now, we are in a mature, open source world and it's more of an attitude than a method: define your business objectives and then we‘ll find the technology to solve it. It costs almost nothing to be curious but might kill your business if you don’t.

 

So it’s now a business use-case problem, specific to each organisation with one question. What is your major pain point or challenge? Is it fraud? Is it risk? Need to improve conversion rate on digital marketing or save money on the CustServ budget? Is it customer churn? For all of these use-cases, there’s a technology stack that can solve it.

 

Lets listen to all type of customer signals, find meaning for patterns and proactively anticipate the customers needs. And that use-case is pretty descriptive of what I’ve done for the last three years...