CSat, AHT, ASA, DCs, ER, NPS, FCR, FTR, TT and TTR*. We could build a new a language with all the metrics that we have for measuring customer service performance! My colleagues at Impact Learning have addressed the why and how of customer service metrics in earlier blog posts. While measuring customer service is necessary and good, customer service managers would benefit themselves and their customers if they reviewed their metrics against two common flaws.


An over-reliance on productivity (efficiency) measures, and an underweighting of quality (effectiveness) measures will drive your performance in the wrong direction. It is important to track how many tickets are handled; how many cases are closed – valid quantitative measures. However… revenue growth and customer retention are triggered by customer satisfaction – a quality measure. As a rule, most productivity measures are internally focused, helping your organization to manage expenses. And most quality measures are externally focused, to build customer loyalty and revenue growth. Both types of measures are important.

One story relates how a retail bank network invested heavily in product training, expecting their bank tellers to cross-sell additional services. For example, a customer withdrawing a large savings to purchase a car may be offered a car loan. A new parent may want to look at a college fund or investment programs. However, this bank’s main customer service metric was transactions per hour (a productivity measure) and this goal was amplified with monthly teller bonuses. It would obviously take time with customers to explain the additional products. In spite of a considerable investment in product training, the program failed. The root cause was not in the training; the flaw was in the metrics.


What gets measured, gets done. Of course, we’re not the first to cite this. What can happen is that we measure OLD behaviors, or worse, the WRONG behaviors. We put posters up about the importance of customer service, we launch internal marketing campaigns, we set “customer service” as a corporate value, and we even have customer satisfaction contests and customer service training. However, all can be a waste of time and effort if we continue to measure and recognize the wrong behavior.

Another more recent story is from a computer company that implemented a multi-million dollar knowledge base system in their technical support division. The rationale was that with a searchable database of technical articles, customers could access this information and not call customer service. Moving some of the cases to a self-help solution had tremendous potential savings. For this to work, the assumption was that the agents would gladly populate this new knowledge base that could be accessed by the customer. However, the main metric remained case productivity – number of cases closed per agent. The new behavior (writing articles) would actually be punished with lower case productivity scores. The metrics were not adjusted to support the desired behavior.

As a trainer and consultant with Impact Learning Systems, we do use metrics to track our performance. Fortunately, we underweight on productivity (number of classes conducted) and overweigh on quality (satisfaction ratings from our clients). To see some of the metrics about our customer service training, go to Impact Learning Systems and search on results.

(* CSat = Customer Satisfaction, AHT = Average Handle Time, ASA = Average Speed of Answer, DC = Dropped Calls, ER = Escalation Rate, NPS = Net Promoter Score, FCR = First Call Resolution, FTR = First Time Resolution, TT = Talk Time, TTR = Time To Resolution)