Real-time customer experience is a vital driver of growth. Acting in real-time, armed with the most up-to-date information about your customer, can hugely improve customer experience. Many of SAS’s customers have generated significant competitive advantage from trying to align closer with the real-life experiences of their customers. But how many organisations recognise this, or even understand what the term truly means?
Defining real-time customer experience
Real-time customer experience could just be about reacting to real-time triggers with a pre-planned message, or following people around on the website and reacting to click-throughs and open rates. And, in fairness, that is how some organisations seem to see it.
It should however, be much more about harnessing relevant data to understand more about customer behaviours, and then using that understanding to take meaningful decisions to optimise customer experience. More simply, it enables organisations to respond to their customers in a more human way, bringing them closer together.
To truly “do” real-time properly, you need to be able to listen to real-time signals from your customers, understand what these signals mean in the larger context of their relationship with you, and then identify and execute the right action to meet these needs.
Real-time may not always be essential. Some suggest that making the right decision is be more important than trying to force an instant response, but there is no question that real-time analytics is changing how organisations interact with customers. Life was simpler when customised coupons were sent by snail mail, but then customers did not have so many competing demands for their attention. Organisations now need to adapt or die, and real-time data allows them to be more relevant, and therefore grab their customers’ attention more effectively.
To give you an example, a bank or retailer may wish to offer additional credit to a credit card or store card customer. They may periodically decide, based on current credit scores, balances and spending patterns, how much extra credit to offer to each customer, and then make that offer through their regular communication channels.
However, the need for credit is often felt most strongly at particular moments in a customer’s life (such as when on a Christmas shopping spree, and reaching the limit of their store card). Additionally, the amount of credit needed may change significantly from day to day, or even minute to minute. These organisations can use real-time triggers and analytics to determine the right amount of credit to offer (in order to maximize the utilization rate, maximize overall profit and keep risk parameters in check), and also deliver in real-time to the customer if a particular pattern is identified (such as a customer on 23rd December making three purchases within 30 minutes).
Real time therefore enables right time by supporting adaptive behaviour that changes based on the preferences shown by customers as they happen. It is, of course, important not to be creepy. Knowing ‘everything’ about your customers is weird, and actually quite threatening. There is, therefore, a fine balance between ‘relevant’ and ‘too much’: that balance lies in knowing what will be useful and acceptable to your customers, and add value to their lives.
Getting value from real-time data
The first question is how organisations should obtain value from real-time data. Rules-based real-time systems offer potential in simple situations, and may well be ‘good enough’ quite a lot of the time. For more complex scenarios, however, such as to decide (in real-time) between a range of available marketing decisions for a specific customer, a more analytics-based approach will be necessary. This type of approach is allows the customer experience to be adapted continuously, and through machine-learning techniques, the analytically-powered system could come up with new approaches to customer experience that the business hadn’t considered.
Ultimately, when considering implementing a real-time analytical model, it’s important to consider how the model result will change your decision, and whether real-time information will significantly change the result of the model. Some examples, such as a real-time balance check, are very relevant when applied in real-time, whereas data such as income level, life stage and segment are unlikely to change in real-time (and so a batch approach is appropriate). Mobile operators live in a very real-time enabled world, and so they can do this well by analyzing the up-to-the-minute data usage patterns, and combine this with longer term parameters such as monthly averages and device type, in order to make a truly contextually relevant decision.
There are, however, barriers to integrating real-time analytics into customer interactions. Perhaps the biggest of these is perceptions: many people believe that it is too difficult and time-consuming, or are not aware of the potential. However, with the right solution, and by starting with a high-value, low-complexity use case, you can be in production within weeks, and generating a big return on investment. One SAS customer made their first sale within 7 minutes of going live, and within weeks, found a 5x increase in response rates when converting a previous batch offer to a real-time, contextually-sensitive one.
Other barriers include the need for coordinated decision-making. Too many uncoordinated customer touchpoints reduces the ability to embed real-time analytics effectively and consistently. Success requires each touchpoint being able to leverage a single view of the customer, and that means a lot of work goes into getting the data right. As most marketers now realise, a disjointed or inconsistent customer experience across channels leads to poor customer satisfaction, and central coordination of marketing and service decisions is vital. But by embedded analytics in this centralized decision making process, real positive benefits in customer experience and revenue can be found.
Black box models
Finally, there is a question about whether marketers should embrace ‘black box’ AI models to drive real-time decisions to customers, as opposed to a more “transparent” AI platform. Some commentators have predicted that AI may replace the need for marketers, but I think this is risky. AI can be used to automate many processes, improve time-to-market and identify the best option from a range of available marketing decisions, but it should deliver relatively predictable results. Where AI supplements human decision-making, to me this is the sweet spot for AI applications right now.
Black box models are also unhelpful if you are looking for transparency and traceability in decision-making. In the banking and financial sector, these two components are key, and only likely to become more so in regulatory terms. The way round this may be to manage the inputs, but let machine learning optimise the outputs. If an algorithm can help us to understand the market better, why not use it to improve customer experience?
The whole business could benefit from real-time
The bottom line, perhaps, is that it is hard to see any part of the business that might not benefit from enabling real-time decisioning amongst the business processes. The goal, or maybe the grail, is to let decision-making become more adaptive, and less rules-based, in real time. This, in turn, will personalise and improve customer experience—and better customer experience, as we know, drives customer retention and improved sales. For example, a bank who is offering credit in real time (when the customer feels the strongest need) will immediately see significant increase in uptake, and the same applies for a mobile operator providing a personalized data bundle offer at the exact moment that the customer runs out.
And beyond marketing, customer service call centres are all expected to be real-time enabled. Whether staffed by chatbots or humans, being able to make decisions in real time will have a major effect on customer satisfaction. It’s vital that real-time decisions are made taking ALL available customer data into account, however, as opposed to just the siloed data available within the call centre.
There is, in other words, no question that real-time data and analytics can support decision-making that will improve the customer experience. With customer experience rapidly becoming the main battleground for competitive advantage, embracing real-time is no longer optional.