Customer analytics is becoming an important tool in gauging the performance of contact centers, and will be a frequent topic of discussion in my postings here. I’ll start by saying that this is a multi-faceted and complex discipline, but offers significant value when applied effectively. There are many ways to do this, and the logical response is to define your processes from the inside-out. By that, I mean to build – or outsource – the contact center around operational considerations, and only once that is in place, then focus on managing customer interactions. Look after the internal needs first and then look outwards to the customers.

This is an enterprise-centric model that can be characterized as a cradle-to-grave view of CRM. However, to get a deeper understanding from customer analytics, a reverse approach may be needed. In other words, take a customer-centric view and work backwards from there into the organization to determine which processes are getting effective results and which ones are not. This is fraught with complexity since there are multiple processes that must be managed both individually as well as interdependently.

For example, branches of the IVR tree may keep leading to dead ends, or escalating a support call to a more qualified agent involves starting all over again, or handling inquiries that involve multiple lines of business require the customer to make separate calls for each one, etc. All of these are separate processes, but each has an impact on customer satisfaction, and it is this end result that should determine which processes need fixing. This is a very different mindset than just making adjustments based on operational inefficiencies.

Thinking this way, we now have the grave-to-cradle approach, where every outcome from a customer interaction can be mapped back to a specific action or process. A key challenge, however, is the fact that each process has distinct metrics and parameters that need mapping. As noted earlier, each link in the chain needs to be defined and measured, but aside from fixing the weakest link, the solution must also make overall chain stronger. That said, whether the outcome is positive or negative, these analytics serve the customer first, and from that provide direction on how processes and workflows can be improved.

This may sound like an enlightened approach to contact center management, but it runs contrary to conventional business thinking, which is driven more by cost and profit considerations. Given the complexities of a truly customer-centric model, it is easier to see why enterprises would adopt a hosted contact center. This is hardly the only reason to go hosted, but be forewarned that simply outsourcing a non-core operation does not translate into a well-oiled grave-to-cradle solution.

There are many variations on hosted contact centers, and care must be taken to partner with the right one. In short, this means having both the tools and analytical expertise to manage the processes that apply specifically to your customers. The discipline of customer analytics is complex, and outside the comfort level of most enterprises to manage internally. As such, when it becomes evident that incremental improvements in customer satisfaction yield substantial benefits to the enterprise as a whole, the rationale to find the right hosted partner carries more weight.

The first thing to consider with customer analytics is deciding what to measure. That very much depends on who makes the decisions, and can be the root cause as to why these metrics have limited value. If the decisions are coming from the contact center, the metrics will be driven by operations – call completion rates, calls per hour, cost per interaction, etc. These can be very useful to showcase performance efficiencies for both individual agents and the overall operations, but with little connection to the broader business.

Conversely, when these decisions are tied directly to organization-level drivers, the value of customer analytics becomes much greater. Strategic objectives that serve the enterprise and their underlying business will always trump those of the contact center, but with the right integration, can complement each other very well. In this context, operational metrics for the contact center have more impact when they are tied to broader business objectives such as gaining market share or increasing share of wallet with customers.

Tying these two sets of metrics together is complex, but when the benefits are understood, the rationale is self evident. This brings us back to the grave-to-cradle concept. If I take the term “grave” literally, consider contact center scenarios that end badly. These are all too familiar for enterprises, and whether they translate into lost business immediately or down the road, contact centers need to be accountable. Since it is unrealistic to expect perfect performance from any contact center, enterprises must view these outcomes as events that need to be managed and mitigated as best as possible.

This is not an easy fact of life to accept, but once this mental hurdle is passed, the action item is to establish processes that allow the enterprise to learn from these setbacks. We all know that you cannot improve what you cannot measure, and that brings us to customer analytics. In this scenario, customer analytics are used to identify not just where problems occur, but how they impact strategic business objectives. Along the way, they will identify breakdowns in the service delivery process, but this is not their sole purpose.

The overriding focus here is to identify patterns from the analytics that pinpoint weak links in the service delivery process. Conventional examples would be useless IVR options, dropped calls, circular routing back to the initial agent, inability to reach a live agent, etc. These weak links become even more problematic in today’s multimedia environment, where customers expect to use multiple channels to resolve problems, several of which are new for many contact centers. At one level, this analytical data is prescriptive in that it helps the contact center address these problems. However, there is also a strategic value that resonates with business-level decision makers. This is where complexity becomes a challenge, but when the metrics can identify both the root causes of problems and the specific impact on the business, these decision makers will be all ears.

There are actually two contrasting scenarios that make these customer analytics compelling. The first scenario was just presented – cases where contact centers have come up short, and as a result, business is at risk. Quantifying this is important for all kinds of reasons, but above all the business must use the data to learn from their mistakes and make the corresponding improvements.

This is important at a general level, but has a deeper impact at a more granular level. Customer analytics can segregate patterns by different types of customers, which allows the contact center to further refine their improvements. This is becoming increasingly important as the modes of communication keep increasing. Some customers prefer direct contact – such as by telephone – whereas others prefer indirect channels such as email. Even if the problem sets are similar here, the workflows are different and must be addressed as such.

At the other end of the spectrum, customer analytics can identify patterns around positive outcomes. Of course, businesses must fix the problems, but this only helps maintain the status quo. To drive true competitive advantage, it’s equally important for any business to set the bar high for best practices in customer satisfaction. This means learning from interactions that meet or exceed customer expectations, especially when analytics can single out the high value customers. Not only are they the most profitable, but they present the best opportunities for upselling.

To sum up, in terms of serving business-level objectives, customer analytics are valuable not just for redressing contact center shortcomings, but also for maximizing revenue potential among their most valuable customers. A case can be made for either scenario to support the grave-to-cradle proposition, but taken together I think they tell a pretty strong story. Actually, I think they tell a few strong stories, and I’ll get to each of them in future articles.