Stage 1: "Everyone is asking this NPS question, I better ask it too!"

In this stage, businesses measure their NPS for the sake of measuring. They send out an invitation to an NPS survey, get the results back from the survey, and they stop there. Companies don't actively try to improve their score in stage 1.

Stage 2: Alerts for unsatisfied customers

The second stage adds a flow of automated alerts when respondents give a low score. Usually, businesses choose to cover the entire group of detractors with a score from 0 to 6, but you can also choose to narrow that window (e.g., score 0-4) if you experience large volumes of alerts. These alerts are sent to your support team or account manager including some open text about why people were unhappy with the provided service. Then your team tries to get that detractor back online.

All these alerts are coming back. Your support team and account managers are getting notifications and contacting detractors. At the management level, however, there is a link missing. Management has no idea what the effectiveness is of the NPS feedback program because they aren't tracking the results. Most businesses are in this stage right now. This stage is definitely better than stage one since you are actively trying to improve your NPS score ,but this stage lacks the structure of getting the data back to the management level:

Stage 3: Manage automated workflow

In the flowchart ,you will see that the alerts transformed into a dedicated feedback loop. The biggest hurdle from building your automated NPS program is going from stage 2 to stage 3. To have a successful stage 3, you have to create a dedicated NPS follow-up team inside your company. When there is a detractor ,a follow-up form is sent to your follow-up team instead of the basic alert. The team will fill in their own survey (with all the data from the clients NPS survey pumped into it) to answer questions that helps break down all the detractors' feedback into categories for a structured reporting.

Firstly, they will be tracking their calls. This allows your follow-up team to share the workload so nobody is calling the same person twice. You can exactly track how many open and closed follow-ups still to be done.

Secondly, they will try to identify the underlying reasons of the detractor and divide the open feedback into categories. It's crucial to let your follow-up team categorize the feedback. When asking "What can we do better?" or "What did we do wrong?", you don't want to put words into the respondent's mouth by giving them a fixed list of possible issues like "bad service" or "late delivery". At the end of your clients, you want an open response. At the end of your follow-up team, however, you want to be as specific as possible to get closed answers. In the follow-up survey, you want to be as specific as possible paired with an 'other' response option. Over time, we recommend to analyze the responses in the 'other' category. You will be able to identify recurring items and put those items as fixed categories in the response list.

With the structured data coming back from your detractors, you can also do A/B testing of your incentive with your detractor strategy. You can, for example, give your detractors 1,000 text messages for free or you can make your detractor commit to a two year deal where they get $ 10 off each month. In the results of the follow-up survey you can determine which incentive works best for which detractor category.

Structured data from your detractors will flow back into your company in the format of dashboards, report forms or custom integrations. Management can then analyze the data at least monthly to take actions to improve the categories where the most complaints are coming from. It's important you get to this stage as fast as possible. You will notice the structured data from the follow-up team really improves the quality of data you receive.

Stage 4: Integrate with Business Intelligence Software (BIS)

The final stage is going to add a layer at the top of the company in the flowchart. It's probably some kind of BIS (e.g., SAP, Microsoft Dynamics) where you will be crunching all the different data streams. Another telecom example explains the dynamic between your business intelligence software and the NPS feedback program well. Person X (churns and) switches to a different telecom provider. The business intelligence systems will determine the influence of person X on other people of his personal network. What influence will this person X have on his personal network and are therefore others in his network at risk of switching to a different provider? BIS will use various data streams to identify the strength of the social connection between person X and person Y. Frequent and long calls between person X and Y will make their social connection more important.

Now the NPS feedback program comes into play. There could be two sources here to use the NPS survey: as another data stream or as confirmation of your churn identification. The NPS survey as a data stream contributes to the churn risk coefficient as one of many factors. The other case where the survey serves as a confirmation NPS plays a much bigger role. When your BIS identified an at-risk client (based on other data streams), you can send out a survey as a trigger to confirm your churn identification is working. If the client then indeed gives a low score, you will know that client is indeed at risk of churning. In short, the NPS feedback program can be a positive feedback mechanism for your churn identification. The score itself can alter the churn risk coefficient to improve the business intelligence software.

Based on this flow, you can set up an automated system with incentives to win back the at-risk customer when you are dealing with large volumes because you might not be screening every single detractor with your follow-up team. Based on the client's feedback and other data streams, you can send them that deal with 1.000 monthly free text messages or that deal to lock them down into a two year package with a great offer to stop them from detracting. To stop one client from detracting is a bigger deal than it might look at the surface. One detractor is like a snowball rolling down the slope taking others in his personal network with them; those people from his network will do the same and that starts a vicious cycle. If you can put a stick in front of that one client to prevent that person from detracting, you are not only saving this particular detractor but many others in his network at the same time with the same effort.