The Net Promoter Score (NPS) was developed in the early 2000s by Reichheld and Markey who were working for an American consultancy (Bain and Co). At that time, the problem they thought they had to solve was how to boil down lengthy customer satisfaction surveys to one figure, the NPS. The question asked of a customer was ‘How likely are you to recommend us?’ This is indeed a really interesting question to ask, because it will give some insight into the happiness of customers and therefore the organisation’s reputation. NPS is now near ubiquitous throughout the public and private sectors. However, from my point of view, something crucial is being missed here: How would you go about improving your scores?

Obviously, NPS was not designed to help you to do that. I call NPS a lagging or consequential measure. A consequence of changing the service may become apparent if the NPS goes up, down or stays the same. In that way it might be helpful to track if changes you make show signs of an improved NPS. To me, it is a bit like staring at your weight on a set of scales when you want to lose a couple of pounds. That, in itself, that won´t help you to lose weight will it?

Now Reichheld and Markey suggest to ‘close the loop’ with a client, e.g. contacting the client immediately afterwards and solving the root cause of their problem. Solving the root cause is great, but my experience has shown that the root cause lies in management thinking about the way work is designed and managed. These are fundamental changes that require a concerted organisational effort.

Working with service organisations, they tend to go for conventional approaches to making a change: They might differentiate NPS in products or functions and prioritise. They might train people to be politer to ensure a better customer experience. They might define standard procedures for staff to follow to get a standard level of service quality. And beyond that they might set up quality teams to check if staff are working to these standards. They might build customer loyalty programmes or might send staff or management on training programmes to be more customer centric, etc, etc. To be honest though, in my view, this is doing the wrong thing righter and therefore a waste of time and money. It is designed around assumptions about how to improve NPS, not on proper knowledge and understanding of why the organisation is not performing. When I study service systems I hardly find unfriendly staff, but very often unhappy customers who have to contact organisations again and again to get their problem solved. Do we want to ensure that customers get served in a friendly manner to a standard? Or do we want to understand why customers have to get in touch with us multiple times in the first place?

There is a better, more systemic method to understand where it goes wrong and what needs to be changed: study the system from a customer’s point of view and establish better measures to learn and improve.

Definition of Failure Demand by John Seddon, Freedom from Command and Control (2003)

In any transactional service, the very first thing to study is the nature of the demand being placed on that service. There are two demand types: value demand and failure demand. Value demand is demand that an organisation wants – ‘Ì want to buy your product,’ ‘I need car insurance’, ‘Please fix my boiler’, and so on. Failure demand is demand that is caused by a failure to do something or do something right for the customer – ‘I am still waiting for my insurance policy’; ‘I don´t understand your form’, ‘I wanted a green one, you sent me a blue one’, etc.

So that includes all the times where customers need to contact you again, or maybe even at all, because your digital channel did not work or was not clear. If you understand type and frequency of failure demand and the cause, you can reduce it and improve customer experience. Consequently, NPS ratings usually go through the roof. Failure demand is therefore a measure that helps to learn and improve, which I call a leading measure. If you pay attention to these leading measures, the lagging measures like NPS will take care of themselves.

Now the icing on the cake: you will not only increase customer satisfaction, but you will also reduce costs. Vanguard studied service systems around the world and we experienced failure demand rates between 40 and 90 percent. In other words, the organisation wasted between 40 and 90 percent of their capacity with failure demand, demand that we do not want in the first place. Imagine what happens to costs when failure demand is reduced.

The opportunity to reduce costs and increase NPS, reputation and sales in the long run is massive. But it will take some rethinking of current management practice, some determination to study your own system before defining better measures that help you to learn and improve. So, let’s stop staring at our scales in the hope of magically losing weight!

Hendrik Ascheberg

Additional resources on Failure Demand and NPS

New article by Simon Caulkin: Failure demand: what’s the big secret?

New Vanguard Event: Better digital from better method, London 16th May 2018

Listen to Rob Brown talking about saving £100 million and improving Aviva’s NPS from -40 to +50.