When we evaluate a contact center , we are mainly interested in availability indicators %LCR, ASA, % SL , which characterize how quickly subscribers receive an agent's response, sales performance indicators and quality indicators (CEA, FCR / FLR). Sometimes CSI subscriber satisfaction indicators or a pair of CSAT / CDSAT are added to this, but in practice they are rarely measured by OCC – customers save. Also, of course, economics is always important: Occupancy (operator load factor, CSS UTZ (personnel utilization rate), Cost/Revenue per contact. Everything else, in fact, is secondary and serves to explain why the values listed above turn out the way they are.
Actually, it's pretty funny. The industry standard COPC, which I rely on in this article, provides quality management, but not the quality of the result. But there is not one (from the word “generally ""), an indicator that would reflect the quality of the management itself. There, all KPIs are only about one form or another of the result. I will try to explain with an example. Let's say you outsourced an outgoing telemarketing project (you can also incoming, you can do any, I illustrate the principle) AKC provides a report on the status of today and attaches historical data from the very beginning of the project.You look: everything seems to be normal, within tolerance.All satisfied. Operators on chairs are spinning, money is muddied.
It 's really not good there. At least it will be soon. Because if we measure the period from the admission of operators to independent work to their full achievement of planned indicators, it has doubled. The effect is not felt simply because there are few beginners, and the guys who have already more or less got used to it continue to “gain momentum”. The coach quit, who taught basic sales skills, the new one came in worse, not so literate, she herself is not very good at it. It is urgent to intervene in this process, but no one sees the need, the orchestra is still playing on the Titanic. The obvious conclusion is that in order to assess the contact center (management quality), it is necessary to track the dynamics of the timing of achieving KPI target values in any area of work. As soon as it swung down even a little, that's all, air raid alert.
Another story, all the same, only in profile, static. There is a sales department of 10 people, the plan is systematically implemented, the management is satisfied with everything. The performance of the worst and best sellers differs by 2.5 times, that is, a minority works hard “for themselves and for that guy”, and the implementation of the plan depends on three “star” employees. The problem is clear, firstly, you can earn more if you need it (it happens that you don’t need it, for example, when production does not have time to ship what is sold), but it is simply necessary to ensure additional interchangeability of sales. And according to sales statistics, this is not visible, only if you look at each employee individually. Conclusion, it is necessary to track the dispersion (scatter) of indicator values within the objects of interest, for example, operator groups. Without this, it will not be clear at all what is going on in the “kitchen” of the outsourcer. By the way, a good project manager on his part should monitor the timing of reaching the target values and the variance on his own initiative. And experience shows that with such a PM, operators very quickly get used to the project, although they, due to the difference in their positions, almost do not communicate directly with him.
An additional view of the contact center evaluation, what to measure, may be related to time. Let's say ACC serves the incoming line. Does the project manager know the actual client tolerance threshold? Tolerance is the willingness to wait for an operator's response, and the tolerance threshold is the time that has elapsed since the start of waiting, after which subscribers begin to hang up en masse. The Service Level indicator tells little about the situation, it says that there are more lost calls because the waiting time has increased. But it can also increase for internal reasons, for example, when one of the operators was mowed down by covid-19 and could not get on the line, and a replacement could not be found. That is, a change in %SL may not be associated with a change in client behavior over time, but it is important to monitor the behavior. In addition, most often the target for % SL is set not even incorrectly, but according to the principle “poke a finger in the sky”. And, of course, it turns out to be either underestimated or overestimated, and this happens exactly because the tolerance threshold of customers is not monitored. This, by the way, can be used as a criterion for evaluating the qualifications of a project manager: “On what project and how did you study customer tolerance?”. In response, everything is usually clear. Just in case, tolerance has a synonym for patience, in different sources they use whoever.
To summarize:
- Operational indicators of the OCC do not allow assessing the quality of project management. And in the general case, they are not predictors of problems. Although, to be precise, there is a special tool for this – Shewhart's control charts. If you are interested, then welcome to the Russian GOST 50779.42-99
- To assess the quality of management, it is necessary to monitor the dynamics of achieving the goal over time. The spread of KPI values within the object of interest, the time periods in which client behavior changes dramatically.
Dmitry Galkin,
independent consultant on creation and management of contact centers
