Vanguard’s advice on NI 14

A view from the originators

As the originators of the concept of 'failure demand', the Vanguard team thought public sector managers might value advice from the experts; that is the purpose of this note. It might at least help you avoid the waste that will ensue if you slavishly follow DCLG guidance and it will help you learn what others have learned about the power of understanding demand in order to improve services and reduce costs.

Muddling along without understanding

In its first version, NI 14 was concerned with the removal of 'failure demand'. It was pleasing to see how many systems thinkers objected to the idea of having failure demand as a target - pointing out not just the problems of operational definition thrown up by DCLG guidance but also questioning the value of using the concept as a 'target' (systems thinkers know targets make performance worse). In the revised version there is a new definition ('poor use of customer time'). This only serves to confuse the issue, it will increase the problem of operational definitions and thus measurement and, as a consequence, it will lead to high levels of waste associated with recording, measuring and reporting.

Failure demand is a systems concept

To start at the start: Failure demand is a systems concept, first articulated by John Seddon in a book published in 1990 ("I Want you to Cheat". In this book we used the labels: 'demand we want' and 'demand we don't want'). As that book explained it is a counter-intuitive idea; managers usually assume that all demand is work to be done, because their operating paradigm (reinforced in public sector 'best practice' guidance and standards) is:

How much work comes in?
How many people do I have?
How long do people take to do things?

The failure to appreciate the nature of demand - as 'value' and 'failure' demands means command-and-control thinkers miss the greatest lever for improvement. Because command-and-control thinkers treat all demand as 'work to be done', they fail to realise how much demand is created by poor service design (following the regime's advice, exemplified in 'beacons'). Studying demand is just part of appreciating the organisation as a system, and this appreciation causes managers to re-think the way they design and manage services. It also opens the door to understanding the folly of the 'reform' regime.


Failure demand is defined as:

Demand caused by a failure to do something or do something right for the customer

Vanguard people spent many years discussing the definition-in-use and the definition above has proved robust, it works. What is vital about this definition is that it ensures we think about failure in customer terms - something the current DCLG guidance fails to do. For example 'poor use of customer time' (current guidance) ignores the fact that customers are often happy to wait if they know they are going to get the service they want.

Failure demand is a temporary measure

To a systems thinker, understanding demand is a 'temporary' measure, something you use to get an understanding of what is going on in order to make a change. It is not a permanent measure; something useful for monitoring improvement of a service.

Only the predictable is preventable

When you study failure demand you are concerned to learn what demands are predictable; these will be preventable. The point is things will always go wrong (a central idea in the theory of variation), you study demand to learn what is going wrong predictably (again, a perspective based in the theory of variation). If demand is not predictable it won't be easily preventable.

[NB The current DCLG and Cabinet Office definition - 'avoidable demand' - is the wrong concept; 'preventable' is the right concept]

It is not unusual to discover as much as 80% failure demand in local authority services; caused by following the regime's guidance.

Understanding causes

To understand the causes of failure demand you need to study flow. This is where the regime will get a shock (if it is listening). To study flow you have to follow (predictable) 'value demands' through the service as it is currently designed. Doing this you learn that the primary causes of service failure are the targets being used to manage and report performance. So services can look good on targets but be experiencing high levels of failure demand. Targets (and all arbitrary measures mandated by the regime) are 'systems conditions' causing waste.

Back to muddling through - the blind leading the blind

The regime has dropped the phrase 'failure demand' from NI 14. I have little doubt it was because it was becoming labelled as 'the Vanguard target' - a source of amusement amongst systems thinkers. When the target was first published I wrote to DCLG pointing out their mistake. I received a snotty, curt reply to the effect that I was suggesting DCLG had offended the Vanguard 'religion'. Vanguard people are interested in what works and what doesn't work, not defending a particular point of view at all costs.

You might wonder why the specifiers did not seek Vanguard's advice. It is because their method for developing targets is to have the blind (in this case a classics scholar) run consensus-building exercises with the blind (local authority managers who have no knowledge of systems thinking). The regime is not concerned with knowledge, merely consensus-building and compliance. Deming taught us opinion is no substitute for knowledge.

Serving the minister

The specifiers are just doing their job. It is easy to see why failure demand became a hot topic for the regime; removing it would reduce costs. Ministers are obsessed with cost reduction. They are unaware of the costs they create.

The specifiers say: "we need to be able to give the minister a number, we need to be able to compare, everyone needs the same definition, and everyone needs to collect the same data".

Temporary measures have use in learning, not comparability. Indeed to mandate comparability will reduce their value in learning. People need to know when to collect data about demand, how to do that (classifying demand from the customers' point of view), how to determine the predictability of demand, how to understand what matters to customers and, most importantly, how to design a service so that it works (delivers the value as specified by the customer).

If the regime could understand that managing value (the other type of demand you should study) is the key to removing failure and that in managing value you need measures that relate to purpose from the customers' point of view (not targets) they would appreciate that the removal of targets is the fastest way to release efficiencies (for at least you have stopped doing the wrong thing and there is more chance people can do the right thing).

The latest version of NI 14 has significant problems of operational definition and thus will lead to confused and confusing data collection and reporting. It is indeed ironic that the regime's guidance suggests CRM systems can be used to track the customer's experience; CRM systems have institutionalised waste (creating 'work objects' from failure demand, making end-to-end flow for customers worse through 'front- and back-office' designs); CRM systems are a primary cause of failure demand! The more effort the regime expends on defining the measures and defining how the CRM systems will be used to study demand, the more it will take peoples' attention away from where it needs to be.

But that, as we have learned, is the de-facto purpose of the regime: make services worse.

Vanguard's advice

Remember that studying demand is a temporary measure. You only measure it when you think there is a problem and you turn the measure off when you have resolved the problem. Studying demand is a central part of the Vanguard Model for 'check' - understanding the 'what and why' of performance as a system. The next step in this Model - measuring achievement of purpose from the customer's point of view - is the step that creates useful permanent measures.

Study demand into your services in customer terms, studying both value and failure demands. It is essential to develop a customer typology of demand.

Understand the predictability of each type of demand (you will only act on the predictable).

For predictable failure demand, study the causes (they will include working to the regime's targets and other requirements).

Report your demand analysis, the proportion of failure demand and the primary causes of failure - the things the regime made you do that made your services worse.

This will enable you to make a report to your inspectors that is robust and, more importantly, it will enable you to understand how to improve performance. Your report will be limited to the things that are useful to you; you won't have wasted any of your time doing things that have no value to you in understanding and improving the work. If the regime insists you do more, insist that they prove that doing more will be of benefit in understanding and improving your work. If they cannot do so and continue to seek your compliance or threaten to 'down-grade' you, take them to court.

Return to top