Topic: Achieving top level CRM by using an Advanced Forms of Knowledge Management with Paul Tedesco of Cognitor at the 14 October KMPro Chicago meeting.
I had hoped this presentation was going to show me something new - a KM application that builds upon artificial intelligence techniques to provide enhanced functionality. Unfortunately, what I heard at this session of KMPro had more to do with designing KM systems (mainly Customer Relationship Management - CRM) with the customer in mind. This was not a bad thing in itself, it was just that I expected to hear more about the AI components than we did.
The approach that Paul Tedesco described works from the assumption that KM within the CRM world has to be used to improve customer relations which has the (assumed) effect of helping the bottom line of the organization. The heart of the approach is the aggregation of knowledge, experience and expertise to provide better customer interactions either in the sales call or in the panicked call to the support center.
The meat of the discussion had us talking about the basics of CRM and how systems with just a little more smarts would help both the customer and the company. The CRM system must be built to satisfy both the customer and the company. Ideally, CRM provides a coordinated approach to the customer that will leave them happy and ensure continued business. Through the discussion, Tedesco presented a number of examples to describe how this needs to work, from the reactive call-center to proactive data mining on your customer database to determine what new goods and services they might want, based on their history with you.
Along with the baseline automation that CRM systems provide, advanced CRM should also help bring "middle performing" customer service representatives up to a higher level. According to Tedesco, representatives can be broken into three buckets: 20% are high performers, 20% are low performers, and the remaining 60% are mediocre performers. CRM systems should be able to serve up a much better picture of the customer, including preferences, past problems, buying history, and other information that will help the representative both solve their immediate problem and potentially add new sales, if appropriate. Another advantage to building a good solution-finding tool within the CRM is a reduction in the number of on-site service calls, which will save both the customer and company time and money.
Within the business, a CRM system has to help the company discover repeated problems and move on to conduct root cause analysis and feed the solutions back into the CRM system to support the call center staff. This was an area where I expected more interesting discussion about automated discovery of problem patterns, but Tedesco simply glossed over this by saying that the representatives have to manually note repeated problems themselves. This could be much faster, particularly in a large customer service center, if the CRM would be able to identify these things automatically.
On the martketing and sales end of the equation, CRM systems should be able to help coordinate information about customers so that the company knows when and where to offer certain promotions and who are the likely customers to take advantage of such promotions.
The Cognitor product, Intelligent Answer, appears to include natural language search of issues databases, so that call centers (or users) can enter full-text questions and get matches to problems and solutions within the database. (I suspect this is where the AI comes into play - their web information isn't much clearer.) Tedesco claimed that 92-95% of the incoming problems are solved this way. (Assuming the problems are in the database to begin with.)