The Lean Six Sigma snake oil show

I watched two Lean Six Sigma ‘experts’ give their show to a group of senior managers. What a plausible and compelling show it was. It followed the normal pattern; there must be a manual on how to bamboozle managers with this stuff.

We heard how Six Sigma had been the saviour for many well-known companies (implicit message: you’d be dumb if you didn’t follow suit). We heard how Six Sigma would put the customer at the heart of operations (who could argue with that?). This needed ‘process ownership’ so things could be paid attention to across silos (another plausible idea that does not threaten anyone’s patch). We were told DMAIC is a structured methodology for doing this work (again, such an attractive idea; we managers get to define our problems and then our people wearing coloured belts would do our bidding and improve stuff… we’d better make sure we get enough of the training so our people know what they are doing).

As we got further into the nitty-gritty we were told variation is the problem and the removal of it the solution. We were told minimal variation (Six Sigma) would produce amazing results, and how being 90, 95 or even 99% good was rubbish from the customers’ point of view. We were told that Value Stream Mapping would enable work to be designed as standardised flow that would get rid of the waste. And just to wrap it up and make us feel completely compelled to buy we were treated to a few obscure Japanese words. Yes, we knew that this must be the font of truth from which we were drinking today; how honoured we were to be in the presence of these Master Black Belts.

The managers were impressed. It took only a couple of minutes to burst the bubble. First I asked what the consequences of standardised work would be on the system’s ability to absorb variety. Yes, when you standardise a service costs go up. Standardisation appeals because managers love standard cost data for planning and management of work; when they learn to take a systems view they iscover that managing with these data drives costs up. We then got into the business of variation. Using probability data (as the Six Sigma ‘experts’ do) has two problems: they start with management’s current pre-occupations (for example turning around post in five days) as the ‘standard’. Measuring around an arbitrary standard will put you at risk of treating those data points above the standard (over five days) as coming from a different system to those within the standard. When you take time-series data (of far greater utility) you often discover that the results above and below the ‘standard’ are just as probable. If you treat data as though they are different when in fact they are from the same system, you tamper. That means variation increases. Brilliant.

If all of that has gone over your head, what it boils down to is these are two fundamentally wrong-headed ideas; they actually make performance worse. These people are tool heads. They don’t know the limits of their tools because they have never learned how to think, the very thing that Ohno taught as essential.

We didn’t get on to the waste associated with the training (never use most of it), or the problem with starting at ‘Define’ (the current conception of problems is the problem) versus ‘check’ (changes you view of the problem). It was clear that what was on offer would make no change to the system.

The Lean Six Sigma snake oil show: the plausible being offered to the gullible. This was one show that won and then lost its audience. If you want to invite me to your forthcoming Lean Six Sigma snake oil show, get in touch, its educational and fun. And it could stop you wasting loads of money. Bargain.

Tool heads at work in the NHS

People in the NHS who are following Vanguard’s work have been in touch to get my views about lean manufacturing tools being applied to the health service. Just as with the above nonsense they tell me standardisation is offered as the route. There is talk of ‘job families’. As I understand it this is something manufacturers do to reduce variation – if things they make have similar parts it makes sense for them to travel down the same flows. It might seem plausible for health services to follow the same idea. But the thing you have to work on first is demand and you should develop a type classification empirically, without reference to current ‘treatments’. I spent two days on an ambulance and saw for myself the waste caused by complete insensitivity to the nature of demand.

And the tool heads say nothing about measures, they are not attuned to the way the current targets etc drive waste into the system. But that’s because they are tool heads not systems thinkers. If you have not read it already you can find “Watch out for the tool heads” at:

Adult Social Care

On March 22nd in London, a Vanguard expert will be presenting our work on Adult Social Care to NHS people. If you work for the NHS and want to attend, please contact Sean Barnett:

Expensive misguidance from the ‘big boys’

A reader sent an article from one of the ‘big’ consulting firms. The authors claim that services are more difficult to measure and monitor than manufacturing. But they say executives should rein in variance (note ‘variance’, not variation) and boost productivity by implementing rigorous metrics.

The authors tell us variance is difficult to measure and the discussion on variance reveals they mean differences in performance between operations or sites. So now we know, they have started with a command and control pre-occupation, getting data to the centre to give the management factory something to ‘do’. So the hunt is on for ‘meaningful’ comparisons. What follows is advice on service-level agreements, work volume measurement, measuring cost-drivers and, finally, standardising service design.

All of these things drive up costs. If only they knew. And you can bet their ‘help’ will be expensive.

“Buy us, the big boys, we’ll drive up your costs in two ways”.

The Audit Commission gets it wrong again

The Audit Commission is telling local authorities they should use private sector contractors to handle planning work; their idea is that this will speed up planning applications. Classic command-and-control thinking, they assume the only way to do more work is to use more resource. Wrong. The problems with planning start with the waste attributable to the current targets, there are associated problems with design, lots of bureaucratic waste.

Vanguard clients have designed the waste out by acting on the causes. No longer do their customers get an unhelpful and slow system that might cause them to pay more than once for the same thing (getting a service). Instead customers get help with putting in a ‘clean application’ (clean input means fast flow) and get to know the result quickly.

Removing waste is the way to go if you want to create more capacity, something the Audit Commission does not understand. Instead the Audit Commission is a cause of waste; it is at the heart of the specifications and inspection regime. Counting the wrong things not only wastes resource, but also causes managers to manage their systems in the wrong way, driving up their costs. Another example of driving up costs in two ways with one intervention.

In fact in three ways. If private sector contractors get lots of planning work, they’ll hire in expertise from the public sector, exacerbating the current problem with lack of experienced planners. Costs of all sorts will ensue.

Will ID cards protect us?

Apparently ministers have learned that ID cards won’t help in the fight against terrorism, so they have moved their grounds for promoting the idea. We are now told we need identity cards to protect our identities; identity theft is on the rise etc.

In New Zealand they introduced photo-driving licences. The consequence is an increase in identity theft. They are learning that the ‘better’ an ID system is, the greater the advantage to the fraudster. Apparently you can now buy high quality (modern electronic) IDs for under $20,000.

In a related way, unemployment benefit fraud has become a greater problem because the new ‘Gershon flagship’ factory design (they have call centres and back-office processing units) keeps people out of the process. The best way to detect fraud is to be talking to people face to face, something that has been designed out.

Is the ID card going to cut a swathe into our economy? Should lack of knowledge about the consequences worry us? I think so.