Net Promoter Score: working on the 5%

One of the more recent fads to hit our service organisations is something called Net Promoter Score (NPS). As recently I bought a service from my telephone provider, I got hit by it. I’m sure many of you are familiar with the routine. It goes like this:

You get a text on your mobile phone: ‘Hello, thanks for contacting us. We’ll text you shortly for some feedback about your experience’, quickly followed by a second text: ‘Question 1 of 3: How satisfied were you with the service provided by our agent, Lewis – from 1 (not at all) to 10 (extremely).

I was, initially, stumped. For Lewis had been OK, in fact a nice fellow, but before I got to Lewis I’d had to struggle through a voice-recognition system that kept asking me to repeat myself (‘I want to buy a static IP address’ – and the machine, like me, didn’t understand what a static IP address was) and when, after much frustration, I eventually got through to Lewis, the first thing he did was tell me I could have got what I wanted on their web site.

I’d already been to the web site and hadn’t been able to find what I wanted. I explained to Lewis that I had tried, and as ‘buy a static IP address’ was not the first choice on the web page, and I didn’t even know what a static IP address was, so wouldn’t know where best to click, I had given up and called, only to experience the labyrinthine computer system. I also said I assumed his managers had told him to tell all customers they could get their services on-line; Lewis acknowledged that was the case. I imagine his manager will have that on the check-list for inspecting poor Lewis’s calls.

Anyway, back to the first NPS question. I replied: 10 for Lewis (he’d solved my problem), and 1 for the company. This got an immediate reply: ‘Sorry to hear that. Question 2 of 3: Based on this call, how likely would you be to recommend [the company] to a friend or colleague? From 1 (definitely not) to 10 (definitely)?’

I replied: 1.

This again got an instant reply: ‘Sorry you would not recommend [the company]. Question 3 of 3: Did the Agent recap what had been agreed with you? Yes or no?’

I replied: ‘You missed the point’.

Which got another instant reply: ‘Thanks for letting us know. Finally, is there anything else you’d like to tell us about your call or overall experience?’

I replied: ‘You missed the point’.

Another instant reply: ‘Thanks for taking part in our survey. Your feedback will help us continually improve.’

I replied: ‘I doubt that’.

Half an hour later I got a call from Karen, Lewis’s boss. She was concerned that Lewis had received a ‘1’ rating for his call. Clearly the computer had only picked up one number from my text. I explained to Karen that Lewis got 10 but the company got 1. Karen was relieved because Lewis’s NPS ratings go towards his bonus. She told me she’d change his rating to a 10. (As though I should care?)

I explained to Karen why my experience was lousy, wasting lots of my time, and how this had nothing to do with Lewis and everything to do with the company’s management. Karen was nice, polite, but couldn’t do anything about anything. She hesitated but felt obliged to tell me I’d be getting another text following her call. And this is what came:

‘Hello, now we’ve called you back about your enquiry, how likely would you be to recommend our company to a friend?’

You couldn’t, as they say, make it up.

NPS is a lagging measure. You should never manage with a lagging measure. Look at what happens when you do. Managers fail to realise that it’s the system that governs performance, not poor old Lewis. But poor old Lewis and his colleagues are assumed to be the people who can be held responsible for customer service. NPS is used to manage the people; in the words of its architects, it is used to hold agents to account. It is, at best, to work on the 5% (see later). The architects of NPS believe that controlling the likes of Lewis and ringing up customers who give a lousy rating (they call this ‘closing the loop’!) will improve their score for ‘would you recommend us to a friend?’

Recently NPS changed its name: from Net Promoter Score, to Net Promoter System. Systemic it is, but a systemic disease, not a source of systemic health and vitality.

The ‘net’ score is calculated by subtracting those who (say they) won’t recommend from those who (say they) would. It is not unusual for us to find Net Promoter Scores to be negative in clients that are new to us. We don’t bother to educate managers on the folly of NPS at the start; we just wait for the first intervention to make huge difference to the score, as it always does. It causes leaders to get curious. Eventually they learn that the measures used to understand and improve performance had nothing to do with NPS.

So why did we buy that then?

It’s the system stupid

The thing that attracts managers to NPS is their belief that it’s the people who deliver service, so NPS is attractive as a device for controlling them. It is, as regular readers know, a way of controlling that sends the system out of control. The counterintuitive truth is that 95% of performance is governed by the system. My favourite video, illustrating how leaders react to this idea and how they learn it is true will be familiar to some readers; I have referred to it in the past as the ‘Hayley video’. Hayley now appears on the new site, embedded in a video talking about the process of changing thinking. If you haven’t seen it you can watch it here:


From lagging to leading

The first time I got hit by NPS, I thought ‘terrific’, because I assumed they were using a Vanguard technique – the first question is the same (how would you rate the service out of ten?), except of course, we make no mention of the agent. But no, what followed was anything but what we’d teach clients to do.

Vanguard’s second question: For those who rate the service (say) less than 8, ask: ‘What would we have had to do for you to rate it ten?’ That question elicits the things that matter to customers – their ‘nominal values’ – and is crucial for understanding how to improve. This knowledge can be turned into ‘leading’ measures, vital for tracking improvement; secure in the knowledge that what’s improved is what needs to be improved.

Vanguard’s two questions are asked by the people who serve customers; no expensive computer paraphernalia, no reports driven down the hierarchy; no corporate blather, simply an elegant and powerful means to get to appreciate how to work on the 95%.

Does Mr Hunt agree?

Jeremy Hunt, the new health secretary, is bringing NPS into the health service. Aside from the folly I describe above, I haven’t met anybody who thinks we should be thinking about recommending hospitals to our friends – a bizarre idea. Mr Hunt has a first from Oxford, so we can assume he is smart. But knowledge is not the same as intelligence.

I met Mr Hunt at a think tank on the day he was announcing his decision that all NHS organisations will be required to have electronic patient records by the end of the year. Mr Hunt makes the same argument as Blair and Hewitt did in New Labour days: If someone from Bradford fell over in Birmingham access to the patient record would ensure the right things are done and the wrong things avoided, except Mr Hunt’s version is it would help ambulance medics. If Mr Hunt got out and asked ‘how often does the lack of a patient record impede doing the right thing?’ he would learn that the answer is never – as a matter of fact I did ask an ambulance medic, as in New Labour days I asked GPs. When you ask this question everyone sees the plausibility of the argument for records, but it is not borne out by the facts. People speculate that a patient record might be critical in special cases, but can’t think of when such cases occurred.

I didn’t challenge Mr Hunt on that, instead I made two points. Firstly on that very day, the same scheme, promising savings of bn in the USA, was announced as having failed. His civil servants made a note to read the report. You can read it here:

The spurious arguments for savings are the same in the UK as they were for the USA: Here PWC have supplied the arguments, important to say because Mr Hunt hid behind their report.

Secondly, I politely pointed out he could be solving the wrong problem: the fragmentation of services being the right problem to solve. No joy there. Mr Hunt thinks the patient record will solve the fragmentation problem.

Mr Hunt also thinks spending millions on a children’s database will help prevent child abuse. Same problem: the current system is fragmented and not child-focused. A database is the last thing we need; instead we need to design a service that builds a relationship and provides continuity.

What was clear from meeting Mr Hunt was that policy would be a matter of what he agreed with. That’ll do it.

See more on this on our health blog:

Other ministerial madness

Just a few of the other highlights (or lowlights) of ministers being driven by ideology rather than evidence:

Frances ‘mad’ Maude is pressing on with ‘digital by default’. Mr Maude crows about the low cost of digital services like Statutory off-road Notices and assumes all services could be made as cheap. He fails to appreciate that what can be done with predictable, low-variety demands cannot be done with high-variety demands.

Turning to a place that has high-variety demands, HMRC has been hauled over the coals yet again for terrible performance (primary cause being industrialisation – can’t deal with variety). Aside from the failure to pick up 50,000 calls every day, the most striking revelation is that HMRC and its private-sector supplier make money out of failure demand, using premium lines. See:

Making money out of failure demand is, actually, common practice amongst the worst private-sector service organisations and ought to be outlawed – something useful for a minister to do?

The Universal Credit programme has racked up over £600m in IT costs alone and copious reports point to problems with the ‘pilots’ which are doomed to succeed.

And finally, two items on the horrors of out-sourcing: Managers of an out-sourced out-of-hours GP service tell their staff to stop making so many 999 calls: http://www.guardian.co.uk/society/2013/jan/23/private-health-contractor-999-calls

And the private-sector provider making millions out of Birmingham City for handling failure demand put in a voice navigation system to cut costs. But, unfortunately the computer doesn’t understand Brummie accents:

There is much more, but enough for now.

Is anyone listening?

Maybe some are. The popular blog Conservative Home picked up on an article I wrote about the folly of back offices:


Locality, whose voice is listened to in Whitehall, has published our joint paper on diseconomies of scale:

Click to access Public-Services-Civil-Society-and-Diseconomies-of-Scale.pdf

I was asked to write an article on the impending doom of Universal Credit. The journal goes out to senior leaders in social services world-wide:


Call for beta-testers

We are still accepting beta-testers to help us develop the subscriber area of our new web site (01handshake01.com). You have to be a group of people in housing, local authorities of the voluntary sector and you have to agree to work with the site to study and redesign a specified service. The object of this phase of development is to ensure the new site can do what it says on the tin.

Kittens are evil

A whacky title for a special event on the perils of payment by results and outcomes-based commissioning (saying these things won’t work is like saying kittens are evil); 6th March, Manchester. More info here: https://www.01handshake01.com/news/?story=9

One of the speakers at Kittens are evil is Toby Lowe, an academic who has written a powerful piece for the Guardian describing how Payment for results is a ‘dangerous idiocy’. Get off the fence. You can read it here:


And there is a web site were people who are being forced to cheat to make ‘outcomes’ can fess up:

I told a porkie

Action-learning in Hull

The ninth iteration of our popular action-learning programme, in conjunction with Hull University Business School, is open for applications. These are the dates: Weds 20th March, Thu 25th April, Weds 22 May, Weds 26th June. More info here: