• Andreas Jörbeck

Measured service performance – the next piece to the NPS puzzle

Updated: Sep 29


NPS (Net promoter score) has become the new standard for a CSP to measure their overall service performance on the market. It has turned customer service from a cost center to a strategic tool to increase customer satisfaction. The reason for this is that CSP with higher NPS generally are more profitable and they have higher growth than their peers. This has been shown in several scientific studies. This is one of them. So it’s no surprise that NPS has become an integral part of measuring general performance of CSP.


It is still early days measuring NPS, most CSPs have only done this for a couple of years, and already there are some dark clouds on the horizon. You try to break down your NPS metrics into “touch points”. For instance, after a subscriber has finished a call to customer service regarding a billing issue, they are asked to score the call. With this method the CSP can see if they improve over time serving their customer in this particular touch point. The issues with only using this method to do NPS are the following:

  1. Only a fraction of the subscribers can be bothered completing the surveys, generally around 20%. And this figure is declining. So how good picture are you really getting?

  2. Secondly, some touch points, aren’t measured at all. Service quality, which you could argue is the main touch point for a CSP, is generally a black box. A good provisioning process, a friendly customer service and a neat and tidy bill can improve your NPS, but if your service isn’t up to standard, you will be left with an unhappy, possibly churning customers.

Generally service quality KPIs are measured on a network level. For instance you measure call drop rate on a weekly basis and try to improve this figure over time. This figure is used to measure the network department’s performance and gives an indication if they are doing a good job. In addition to this, your CEM function will once in a while do an NPS survey around service quality. If you are lucky there might be some correlation between the two metrics, but it’s highly doubtful.

Having the service quality KPIs on subscriber level means that you can tie a single customer NPS score to real KPI metrics. By doing this you gain:

  1. Tighter relationship between NPS and your KPI metrics

  2. You can find out if there is correlation between your internal performance KPIs and NPS. If there is no correlation, that KPI should be scrapped. If there is, you have something you can industrialize.

  3. You can see at what level NPS score starts to fall. Maybe 3% drop rate won’t affect a customer’s perception of the service, but if it goes over, you can see the effect in the NPS measure.

  4. Now you have something to shoot for, your network department should try to limit the number of subscribers with more than 3% drop rate. If you improve these customer’s quality, you will see a real increase in NPS

  5. Since you can measure these KPIs for your whole customer base, the issue that maybe only 20% answers a NPS survey becomes insignificant. Since you can correlate NPS with the KPI, you can estimate the general NPS for your whole customer base on this touch point with high certainty.

By tying “industrialized” service metrics with NPS score you will gain access to touch points that were very hard to measure before and because you industrialize the measurement process on a per customer level you can be more surgical in your improvement processes. I believe it is time to take NPS to the next level by creating the links between NPS and industrialized KPIs measured on a per customer level for all your customers.

Subscribe to our newsletter

© Subtonomy AB 2020. All rights reserved. | GDPR policy