Move over, skills-based call routing.

Technologies come and go and, for the most part, skills-based routing is a technology that has changed little over the decades. The concept is simple--match the type of caller to the person with the needed skills to service that caller, such as language, product knowledge or a knack for salvaging troubled customer relationships or soothing irate callers.

It may be that the time has come for skills-based routing to share the spotlight. Over the last half decade, the concept of using analytics in call centers and CRM systems has become more popular and prevalent. Companies that deal with massive numbers of customers have so much data to mine through, it seems ridiculous not to mine that information while looking for trends that can help companies sell more effectively.

Recently, I met with Assurant Solutions (www.assurantsolutions.com/targsols.html), a company that has developed a new way of connecting customers and agents. The company, a division of Assurant, a multibillion-dollar insurance company, is well known in both the financial call center business and the payment protection market. Payment protection programs are associated with credit card accounts, installment loans, lines of credit and mortgages. Under the terms and conditions of a debt protection agreement, the monthly interest due from a customer may be waived or the monthly payments may be paid for a covered life event, such as disability, unemployment or family leave. Most often, in the case of the death of a covered account holder, the debt is extinguished.

As time went on, the company started to notice they were able to conduct more effective campaigns if they matched customers carefully with customer service agents or sales people.

Using analytics, the company is able to determine the best agent for the call--not only on basic skills, but on a host of other factors that increase the likelihood of the customer buying from the agent. I know it sounds a bit like magic, but bear with me--this could become a call center breakthrough if it can be duplicated often enough. In fact, the company's product, aptly named Targeted Solution, may end up making a far greater impact than you might think. Assurant Solutions told me that when the system is in operation, agent retention increases. Matched with the kinds of customers with whom they are more likely to experience success, agents' performance levels improve, and thus their job satisfaction.

The company has scores of mathematicians and scientists working on this project, and some of their findings are counter to general industry practices in most call centers. For example, they have found the learning curve for new agents does not increase as much depending on time as it depends on how successful the agent is. Of course, we all know confidence inspires salespeople and sales build confidence, but most centers look at time on the phone as the primary deterministic factor in agent experience.

Targeted Solution ties into customer databases and the workforce management system and continually adjusts, allowing, for example, an agent who is having a bad week to service a type of customer who may be different than that agent's normal ideal customer.

Stop and think for a moment and ask yourself what type of data would be useful in matching callers and agents. The answer I received was enlightening. Assurant Solutions develops a demographic and psychographic profile of each customer with data points like when the customer last paid and details from the customer's purchase history. The company develops a profile based on 27 "dimensions" of a customer. Targeted Solutions goes even further with agents, using 150 "dimensions" or data factors. It was at this point in the briefing that I thought about the matchmaking company eHarmony, which claims to use a sort of personality factor matrix to find compatible couples. The process, if not similar, is somewhat analogous.

In determining some data points, I surmised that a compulsive person may always pay his or her bill immediately, while a procrastinator would most likely pay at the last minute. Other factors include things such as ZIP code to determine what type of environment a customer lives in, as well as details such as the building floor a customer lives in a high rise. Where people live certainly gives clues as to their financial status--at least in general. I imagine some agents are perhaps more effective when speaking to customers of higher income; another agent may be more effective at communicating with lower-income callers. These are just a few examples.

According to the company, while "perfect matching" is important, more critical are the business factors such as revenue and persistency. They look at driving more revenue to a business in a more efficient matter. All factors are weighted and they have the ability to "turn the knobs," so to speak, to adjust. Some reps are better at selling $500 customers while some are better at selling $50 customers. Their success has been more focused on retaining/selling/collecting the most revenue and achieving the most revenue out of campaigns by aligning customers to agents who are best at retaining, selling or collecting specific revenue targets.

Targeted Solutions offers three modules: Retention, Sales and Collection. The system is configurable, allowing a user to decide they are more interested in customer retention, for example, than maximizing initial revenue generation.

Assurant Solutions informed me that within a 12 month Targeted Solutions campaign, a client achieved 67 percent "lift" over baseline in revenue retained, which translates to a projected $25 million lift in revenue over a 24 month period. And the best part is that Assurant Solutions offers a variety of payment options allowing customers--assuming the situation warrants--to pay based on the success of the solution in the company's organization. Assurant Solutions refers to it as "a percentage of the lift," or the success achieved over baseline rates of customer retention or account salvages, for example. It's essentially pay-for-performance.

Here is another thought I had. If using analytics for more effective call routing is the future of the call center, then we need an all-encompassing term for the industry. I discussed some terms with company execs, but we didn't come up with anything 100 percent definitive. I'll keep you posted!

For me, it seems to make sense that analytics should be the overriding factor in connecting agents and callers. In fact, analytics can be considered an overriding term that encompasses "skills-based." In my opinion, at the present time, the best term for this market seems to be "analytics-based routing," as it renders the concept simple to understand. It should be noted that in addition to helping companies with inbound business, Assurant Solutions successfully offers their solution for collections business and other outbound campaigns.

Obviously, the larger the number of agents and the more customer data there are, the more successful the program can be. Assurant Solutions' executives provided us with a number of examples of enormous savings from a variety of companies and campaigns. If this technique can be applied more widely, we can expect call centers to become much more efficient and able to generate more revenue, becoming even more strategic to their organizations.

This concept also has incredible ramifications for e-commerce. Imagine if Circuit City, for example, was able to send out targeted e-mail based on this technology. They could determine what I am likely to buy before I even know what I'm looking for. For example, I might be a chronic procrastinator who begins buying holiday gifts on December 20th. Based on that information, Circuit City might determine that I am therefore more likely to respond to gift cards as an incentive than a discount that must be used before the end of the holiday buying season. Using this, in combination with other data, the store could create the perfect offer for me, served up into my mailbox each time I log on.

But it gets more exciting than that. Imagine personalized Web sites from Amazon, for example. The company could have millions of customized pages--they do this now to some degree--but imagine e-commerce companies using reams of data to be better able to serve up the best offers and customer interactions possible. Customers and companies will be better off if analytics become more widely deployed in all aspects of commerce. Ultimately, these sorts of innovations benefit the entire value chain from customer to manufacturer.