Jerry Silva of Tower Group recently published a report on "Customer Self-Service and Retail Banking in the US: Rising Expectations, Challenges, Opportunities". Jerry does a nice job of describing the various options banks are using to deliver self-service to their customers and of showing how this is likely to grow and become more sophisticated over time.
Jerry identified three main channels for self-service - call centers with IVR (Interactive Voice Response), Web and ATM. It seems to me that all these channels have their own challenges but they also have a need for a strong decisioning backbone. Let's take them one at a time but breaking the first one down a little.
- Call Centers
In a call center you have issues around routing, script generation issues, escalation, staff training and more. However you still need to make sure that the folks in the call center are making good decisions - how much credit to offer, which customers to offer payment holidays to etc.
- IVR and voice recognition
There are many challenges in getting computers to be able to parse voice and integrate it with IVR systems to empower customers to ask relatively complex questions. However, once you have done this and know what the customer's request is, what next? If you don't have automated decision-making systems available then all you are going to be able to do is refer them to a person. That's not really going to get it done.
- The Web
The Web has problems in terms of making the "conversation" with the customer effective and easy enough to complete not to make customers feel like it would be easier just to call. This involves "smart forms" of some kind. But if you can get the customer to fill out the form you need then you still have the problem of how to treat them - will the website make a good decision? will it make one consistent with the call center or the branch? will you be able to change it when you figure out a better way to decide? Again, you really need a decisioning backbone.
ATMs are increasingly seen as a way not just to deliver simple transactions to customers, but also a way to let a bank market to its customers. Clearly there are issues with updating the hardware and network to let more than just the basics to flow through the ATM. But once you have done that what decisions are you going to pumping through it? Are they going to be good ones, targeted to the customer or are they going to be vanilla defaults?
It seems to me that an approach to get more customers to do more for themselves, regardless of channels, requires a ruthless focus on automating customer treatment decisions. If you can automate this, then you can use analytics to improve the decisions you make and make rapid responses when you find out when treatments work or don't work the way you expected. If you focus on these decisions, you can also deliver them consistently across your channels no matter which channel a customer picks.