Score one for contact centres

It is true that a good customer experience will lead to more business, or better "word-of-mouth marketing". Certainly, in the era of social media, a single negative customer interaction can lead to a public relations nightmare.

All of the contact centre metrics used to measure 'services' are proxies for the most-important-of-all contact centre scores. Service level and average speed of answer are maintained because long wait times lead to customer dissatisfaction.

Abandons are a great proxy for customer satisfaction, because customers who hang up are almost always, by definition, not happy with their wait time. Agent quality scores are maintained because contact centres would like to maintain a consistently excellent interaction with their customers, and the agent quality score is the mechanism used to ensure consistent excellence.

There are as many 'best' customer experience metrics as there are customer experience consultants. Different types of metrics can include customer satisfaction, first-call resolution, net promoter score, agent quality score, and others.

Internally, companies will focus on experience scores that can vary from other business units that focus on customer scoring.

But, even if the scores are called the same thing, they will almost always be calculated using different algorithms. This, of course, makes perfect sense, as different customers - calling the same company - are contacting the contact centres for different purposes. The experience should therefore be attuned to the purpose of the contact.

Customer experience scores exhibit seasonality, trends, and differences across contact centres. What does this mean to planning analysts?

Data streams that exhibit this sort of behaviour are similar to many of the other time-series data that analysts typically work with, like contact volumes, handle times, attrition, and shrinkage. Analysts cut their teeth developing forecasts of items that look just like experience data. This means analysts should be able to forecast experience data streams.

This adds yet another dimension of planning. If customer experience scores are forecast by centre and staff groups, these new forecasts can be used in a host of ways.

Forecasts, and the resulting expectations, serve to soothe executives.

First, week-over-week customer experience trends can be drawn out, simply to view where the company is heading. These forecasts then act to set executive-level expectations. If the trends are favourable, that means actual expectations are met. If they are trending in the wrong direction, it will show the given path needs to be adjusted. In effect, this time-series experience data will act as an early warning device.

Similarly, forecasts, and the resulting expectations, serve to soothe executives, too. If there is a traditional seasonal dip in customer experience scores, the company shouldn't be too alarmed when it comes to pass this year as well. But also, if there is an expectation of a seasonal dip in experience scores, the company may be able to head it off this year, by developing an agent training programme in time.

Another great use of a customer experience forecast is as a point of comparison. The best companies view all of their forecasts (volumes, handle times, attrition, shrink, etc) as a baseline for variance analyses. As weekly performance data is tallied, it can be compared to the forecast. Any differences between forecasted and actual performance implies that something has changed. If customer experience scores are being forecast and tracked, any deviation should be noted, explained, and potentially acted upon. In order for this sort of analysis to have any meaning, it must be compared to seasonally adjusted customer experience forecasts.

The final, most interesting use of forecasts of customer experience metrics is as an input into the staff planning process. Several companies use customer experience scores to help allocate their calls among their competing call centre vendors. These companies are actively attempting to improve their customer satisfaction by sending more contacts to those vendors who score best.

Who can blame them? But, there is also no reason why a company couldn't increase staff levels in their centres that also score well. If improving customer satisfaction is important to the company - and execs all think that it is - then it makes perfect sense to include the customer experience forecasts in the staff planning process and decision-making.

It is simple. By developing customer experience time-series data, using this data to forecast expected performance, and applying this forecast to the company's variance analyses and staff planning, it can greatly improve its customers' experience.

* Deon Scheepers is regional business development manager at Interactive Intelligence Africa.