So, the big news at the Lean conference seems to have been Agile. Which is a bit depressing when you think about it. I guess these things have to take turns. A few years ago it was Six Sigma. I imagine, in a few more years’ time, the carousel will have gone full circle and we’ll be back at Lean again. If it wasn’t so depressing, you’d probably laugh.

Someone asked the question, ‘has Lean innovated?’ I thought this was quite a good question. The answer didn’t appear to be immediately apparent. Clearly some Lean initiatives have delivered tangible benefit to the organisations that have deployed them. But is the benefit coming from step-changes to the Lean methodology or merely optimization of what’s been there from the first time Taiichi Ohno devised a better way to organise Toyota’s production systems?

It began to sound like a contradiction. Innovation or optimization? Or innovation and optimization? The more I thought about it, the more the resolution looked like a story of hierarchical separation. At the tool level, Lean has made several step changes. At the governing-principle level, I don’t think it has. The pillars of Lean have not changed over the years. Lean practitioners have, at best, optimized their understanding of these pillars. No doubt derailed along the way by diversions to investigate Six Sigma. And more recently, it now seems, via Agile.

Maybe the integration of the three is the innovation?

Sadly, if we look at the Iron Triangle of business, all Lean, SixSigma and Agile are really doing is shifting the emphasis between Cost, Quality and Speed. It’s still the same old same old: cost, quality, speed – choose any two you like… “everything was taking too long… so we went Agile… then we realised we were just doing the wrong things faster… which made us a bit confused so we brought in the Six Sigma Black Belts… which took out some of the variation… but that didn’t make the customer any happier…” and, hey presto, the carousel has come full circle again.

Should companies use Lean? Should they go Agile? Should they use SixSigma? Should they use Lean and Agile? All three? They’re all meaningless questions, because they’re all just causing the same stupid carousel to spin faster and faster. All three are fundamentally based on an erroneous assumption that life is full of inherent trade-offs. We can see it most vividly in the Agile Manifesto:

‘Over‘? What are you talking about? What these lines effectively say is, ‘there is a trade-off between X and Y, and we’re biased in the direction of X. Granted, it’s a choice, and sometimes choices are good. But rarely is this so if the initial question was a dumb one. Why can’t I have individuals and processes? Why can’t I have responsiveness and a plan?

In fairness to the Agile world, they do at least acknowledge that the Quality-Cost-Speed carousel stops being fun after a few spins. Coming from the high revolutions per minute world of software development, Agile practitioners have realised faster than most that the iron triangle actually has a fourth side: Scope.

Adding Scope to the carousel ride ought to be a good idea. But, sadly, not if – as the Agile practitioners have done – you just use it as another trade-off parameter. Here’s how the Agile community describes their four-sided carousel ride:

Essentially this picture reads: let’s stop fixing Scope and oscillating between over-budget, late and inadequate quality, and instead let’s fix those attributes and let the Scope be the thing that varies. Huh?

Again the wrong question. It’s not a trade-off. You just have to put yourself in the position of the customer to know that. The customer doesn’t want a ‘perm any three from four’ compromise. They want you to solve the contradictions and ‘get better’. Something like this:

Including ‘Scope’ on the carousel ride allows us to recognise that successful evolution has a direction. Everything evolves towards an Ideal Final Result end state in which all of the desire customer benefits get delivered ‘free, perfect and now’.

The way the world works is that we only step closer to that evolutionary end goal when we reveal and resolve contradictions. The only reliable method for doing that is TRIZ.

Lean, Agile and SixSigma have a role in life solely as optimization tools. The right blend of those tools then depends on the relative importance of Cost, Speed and Quality at any given moment in time. When people on the optimization carousel start to feel queasy, it’s probably a sign that there’s a need to solve a contradiction and make a jump to a better carousel. Something like this maybe:

I wish this could be the end of the discussion. Sadly, I suspect it won’t be, and before too long we’ll start seeing Lean-Agile-SixSigma education programmes. As if this is some new kind of panacea. And then we’ll get Certification. And then Standards. And before we know it, the whole optimization carousel will fly off its axis and spin into futile oblivion. Still, I suppose it keeps everyone from having to deal with the real issues. Like giving customers what they might actually want.

Progress happens fastest when contradictions are revealed and resolved. My job means I spend pretty much all my time looking for contradictions. Sometimes they reveal themselves more easily than others. Sometimes it’s necessary to dig quite deep before you discover things you thought were the same are actually not the same.

This is the experience I had when I attended the Lean educators conference last month. One of the governing principles of Lean is the drive for ‘Perfection’. For a long while I assumed this was the same principle as the ‘Ideality’ pillar in TRIZ. I thought, too, that both were present in order to provide those tasked with improving and evolving systems with a compass heading. Everyone knows that ‘perfect’ is a hypothetical end destination that will never occur in practice. Everyone in the TRIZ and Lean communities also knows that unless you know where you’re heading, you’ll never know if you’re moving in the right direction or not.

The fact that Lean ‘Perfect’ and TRIZ ‘Ideal Final Result’ are not the same kind of crept up on me during the conference. Lean people constantly tell me that Taiichi Ohno, the father of the Toyota Production System, made no distinction between innovation and optimization, and that in striving for perfection one would sometimes be making incremental improvements, and other times discontinuous jumps. I can kind of see this in theory. What happens in practice, however, especially if you examine the Lean toolkit, is that the incremental is precipitated much more readily than the step-change. Ohno’s start point was the elimination of waste, and specifically, what he saw as the overriding waste of ‘over-production’. All of the other 7, 11, 15 or – pick a number – ‘wastes’ were introduced merely because Ohno’s start point was deemed too abstract by those he tasked with improving Toyota’s manufacturing operations. The ‘Waste of Over-production’ in Ohno’s terms, by definition incorporated the potential waste of a lost customer, but if you look at the catalogue of Wastes being managed in most organisations, you will see how this difficult one tends to get side-lined in favour of the ‘easier’ internal wastes. This is particularly evident when we look at Lean Manufacturing initiatives. Anyone responsible for squeezing Perfection out of a production facility – like Ohno – has no ability to do anything about lost customers. ‘Perfection’ in this manufacture-oriented context effectively means building the Perfect Car. And that in turn means a car that leaves the production line having generated Zero Waste.

While this might be a good direction-providing (unattainable) target, it is very definitely not the same as the Ideal Final Result found in TRIZ. Like Lean, TRIZ focuses on the customer. The Ideal Final Result is first and foremost the customers’ Ideal Final Result. But there the similarity ends. In the Lean world, the customer gets a Perfect Car. In the TRIZ world, the customer gets Perfect Transport. In TRIZ world, there is a clear recognition that what customer’s really want is the Function. They are trying to get from A to B, rather than have a shiny new, waste-free, Prius in the garage.

Ideal Final Result in TRIZ is defined as the (unattainable) end point of delivering all of the desired Benefits, without any of the Costs or Harms. TRIZ, in other words, takes the B/(C+H) value equation and extrapolates it to its logical end point: all of the positives with none of the negatives.

The fact that Toyota still makes cars and didn’t invent Uber tells me that the Toyota Production System is very clearly not working towards the TRIZ version of Perfect. I think we can see similar dangers of having the wrong ‘Perfect’ goal in any organisation that find itself gazumped by upstart providers offering customers the function rather than the product. Whenever I see Lean being introduced into an enterprise, my first thought these days is, “here is a business heading on a ‘make the wrong thing perfectly’ road to certain oblivion”.

The TRIZ definition of Perfect is a better target, but it too – now I’m thinking about it – carries its own set of dangers. TRIZ emerged into the world at around the same time as Ohno was pulling together the various borrowed threads of the Toyota Production System. The world in the 50s and 60s – at least from a management perspective – was built very much around tangible and measurable factors. Value meant tangible-Benefits, tangible-Costs and tangible-Harm.

What 21^st Century ‘design-lead’ companies like Apple, Tesla, and Uber taught the business world was that customer Value was as much, if not more, defined by intangible factors such as trust, fairness, autonomy, belonging and competence. The definition of the (unattainable) end target needed to change. And that’s why, whenever you see the Systematic Innovation version of the TRIZ Ideality equation, you’ll see it has the word ‘Perceived’ added to it. Value in the SI world is defined as Perceived{Benefits/(Cost + Harm)}, and Perfect in this definition is when the customer perceives they have received all of the tangible and intangible benefits they desire without any of the tangible or intangible negatives.

Adding ‘intangibles’ into the definition of Perfect is, I think, a subtle but quite profound step-change in the evolution of the compass heading of any improvement initiative. It forces those tasked with improving the system to consciously build in to their efforts the need to improve customers’ sense of autonomy, belonging, competence and meaning, etc, as well as delivering all of the tangible elements written into the specifications. It thus also opens up the search for a whole new set of conflicts and contradictions – when the tangible conflicts with the intangible, those are some of the richest opportunities to step-change to a better solution.

In theory, the (unattainable) P{B/(C+H)} end-point is the point at which ‘all’ of the conflicts and contradictions have been eliminated. Every individual customer receives a solution that is Perfectly customised to their own personal needs and wishes.

But, of course, the moment we think we’re anywhere close to what we initially might think ‘Perfect’ might mean, we begin to realise that we’re merely approaching a horizon that, once we get within touching distance, we realise keeps moving further and further away from us.

What could possibly be better as a compass heading for any improvement activity than every customer getting their own personal perfect solution, complete with all of the implications that carries of how fickle we are as a species? What’s perfect for me to day, is anything but perfect tomorrow. Perfect in human terms is potentially a very ethereal and transient thing. In theory the Systematic Innovation definition of ‘Perfect’ doesn’t exclude such fickle-ness, but on the other hand, I now believe the ‘AntiFragile’ work of Nassim Nicholas Taleb provides us with a need to look further over the Perfect horizon and build the concept of anti-fragility in to our definition. The (next) true Perfect solution not only gives individual customers exactly what they want, whenever they might want it, it also self-learns to adapt and change as the tangible and intangible needs of the customer shift and evolve. The more the customer might change their definition of ‘Perfect’, the more the Perfect solution is able to compensate, adapt and anticipate future changes. It becomes, to put it another way, ‘meta-Perfect’ – perfectly working out what ‘perfect’ means to the customer.

Taken all together, I think the various worlds of Lean, TRIZ, Systematic Innovation and AntiFragile give us a story of the evolution of ‘Perfect’. The story seems to me to look something like this:

So what? you say, if the whole thing is unattainable, why should companies care that the end point has moved? If we’re nowhere near the ‘Perfect Car’, why should anyone be building Anti-Fragile Meta-Perfect definitions into their way of doing business?

For me, we just have to look at the amount of value Uber has stolen from the automotive companies to get the answer we need. Our understanding of ‘Perfect’ might be evolving, but the underlying customer understanding of what they want has always been there. Any organisation working in a high-pulse rate change environment – i.e. nearly all of us – needs to recognise that today’s customers already want useful functions our product delivers, not the stupid product itself, they already want all the intangible factors we don’t know how to manage, and, now Professor Taleb has given us the word, they already know they want the antifragility.

Any question containing the word ‘or’, I propose, is a bad question. ‘Is it nature or nurture?’ ‘Should I vote Labour or Conservative?’ Either/or questions everywhere. I blame the intelligentsia. Mostly, though, I blame Socrates. He was the one that made a science out of either/or thinking, encouraging us all – anyone that spent any time in a classroom at least – to bludgeon our way through an argument until one person emerges the victor. In reality, most people – the people stood at the front of the classroom aside, obviously – have learned enough to know that the right answer to these kinds of question is both or neither. Respectively.

Listen out for the word ‘or’ during conversation and it’s amazing to me how widespread it is. Trade-off thinking is everywhere. When you start to point it out to people, at first they don’t know what you’re talking about. Then later, if you badger them long enough, they start removing the ‘or’. So the question then becomes, ‘should we go to Blackpool for our holiday?’ Unfortunately, the shorthand hasn’t taken away the either/or thrust of the question. The word ‘or’ didn’t feature in the actual words, but the way it was meant to be interpreted was still very much, ‘should we go to Blackpool for our holiday or somewhere else?’

This is a much more insidious form of trade-off thinking. Especially if it results in me having to go to Blackpool again. Thinking about it, Blackpool is a very either/or kind of place too. I’ve been there five times in my life, and four times I found myself in a fist-fight. Usually in a pub, and usually coming shortly after the question, ‘are you looking at me?’

Granted, I probably shouldn’t have responded on the most recent occasion, that did he know he’d just used an implied or. ‘Are you looking at me or were you merely scanning your eyes around the room?’. It just goes to show how dangerous either/or thinking can turn out to be.

If you think people use the word ‘or’ too much, just wait til you start listening out for the implied or. Either/or thinking is so endemic we’ve made it invisible.

Every time you hear a closed question, you’re hearing an implied or. Most times you hear a future-tense statement, you’re also hearing an implied or. Every time you hear the word Blackpool… you get the idea.

We built a PanSensic narrative lens to pick up on ‘or’s and implied ors. Just to see how endemic endemic is. The answer turns out to be ‘very’. Especially if you also go listen out for the trade-off solving ‘and’ words and calculate the ratio of ‘or’s to ‘and’s.

I was at a conference last week. A reviewer put in their feedback form at the end of the two days that they thought it had been a very ‘and’ event. I ran all of the papers and narrative content I could find through PanSensic. The ‘or/and’ ratio was a shade over 40. If that makes it an ‘and’ event, I think the best we can say about society as a whole is that we still have a long way to go. Or…

No-one goes to work intent on doing a bad job. W. Edwards Deming taught me that. I also believe that merely arriving at work with an intention to do good things does not mean that good things will inevitably happen. Deming taught me that too. Sometimes, very well intentioned people do an incredible amount of harm.

Like, for example, the people tasked with writing Standards.

I was once on a Committee tasked with writing an International Standard. Every time I went to one of the meetings I used to get a knot in my stomach. At the time I didn’t know why. In the end, I made my apologies and left. I never really understood the reason either for the knots or the departure.

More recently, I’ve been in the position of watching other people participate in the Standard writing process. The forthcoming ISO18404 to be precise. A Standard that is designed to regulate the Lean, SixSigma and ‘Lean & SixSigma’ markets. Under normal circumstances I’d watch the emerging shambles with a wry smile on my face. There are only so many ills one person can get themselves worked up over. And, in theory at least, a Lean and/or SixSigma Standard ought to have nothing at all to do with innovation. One might go so far as to say they are the very antithesis. In practice, however, I have a sneaking suspicion ISO18404 will do even more damage to the world of innovation that it inevitably will to the Lean and SixSigma communities.

What I know about Standards in general and ISO18404 in particular is that they have been established because ‘someone’ has realised that as industries grow they invariably attract rogue elements. Encourage enough Lean practitioners into the consulting market, in other words, and sooner or later it will begin to fill up with consultants that don’t really know what they’re doing.

These rogue elements, therefore, the well-intentioned Standards Committee members conclude, need to be weeded out. Good ones will be ISO18404 ‘Certified’ and, bad ones will not meet the regulated competencies (of which there are 23 in the case of Lean) and hence won’t be certified. So much for the theory. So much for helping wary prospective clients from hiring a ‘bad’ consultant.

The first question that needs to be asked, I think, is how much harm ‘bad’ consultants actually cause? We saw the ‘bad consultant’ problem in the TRIZ world in several countries for a while a decade ago. Idiots that go on a two-day TRIZ course and then think they’re qualified to solve nuclear fission problems for clients. It was a problem for the TRIZ world for about a year, and it was a problem for prospective clients – the real customer – for about a day.

In one particularly memorable (as in ‘I still bear the emotional scars’ way) experience, we had a client that insisted they wanted to go through a ‘real’ ARIZ session. So we found them a real TRIZ Master (Certified by Altshuller himself), and helped set the workshop up. The Russian TRIZ Master told us the session would need three days. At the end of the first day, the client pulled me over on one side and said, ‘don’t worry, we’ll of course pay you for all three days, but, please, we need to end this here’. It had taken them a day to realise that TRIZ ‘Master’ or not, they were in the presence of someone that had absolutely no ability to empathise in any way with a room full of restless engineers.

In reality ‘bad’ consultants aren’t a real problem at all for clients. If the client is looking to, say, embark on a Lean journey, the most likely thing they’re going to do is put out a Request For Proposals. Lots of Lean practitioners will see these proposals, and the least busy will very likely submit proposals.

At the moment none of these proposals will show any kind of Standard accreditation. This is a little bit inconvenient for the client, because it means they are actually going to have to read the proposal. That’s probably going to cost about half an hour. If it’s a really bad consultant, though, they will very likely be found out at this stage. If it’s a more cunning bad consultant, they might get past this first down-select. They might get to a point where they are invited to turn up and talk face-to-face with the client. Another couple of hours gone, but it’s pretty difficult for incompetent people to hide their incompetence for more than an hour when someone is looking them in the eye. When someone lies about their past track record, a simple phone call to references will pick up the problem. And even if the bad consultant prevails through this evaluation and gets the job, their incompetence will be found out within a couple of days of starting it.

Put frankly, if a prospective client isn’t able to weed out bad consultants very quickly and very easily, they deserve all they get. Call it a self-organising system. Idiots deserve idiots. Its exactly the same situation as when clients choose their consultants on the basis of price and it goes wrong: the failure is annoying, but at least they learned something important: choosing on today’s ticket price is a dumb way to go about any kind of business.

Here’s what a Standard will do in the above scenario: a) clients will look for the ISO badge and as soon as they see it will tend to stop reading the proposals and – worse – stop thinking (this is called ‘plausible deniability’ – when things go wrong, they can now shrug their shoulders, point their boss to the ISO logo, and say, what else could I have done?’), b) the self-organising nature of the industry becomes progressively destroyed and everyone – clients and consultants – become more interested in the badge than thinking about the actual need, c) before too long, the whole charade descends into a ‘badge collecting’ industry. Good consultants waste time administering their ongoing Certification evidence; bad consultants spend theirs working out how to subvert the system. See ISO 9000 for a good example of what this looks like when the system is properly ‘mature’.

It’s a bad thing that this happens. Even worse that it appears to be universal.

But it’s not the worst thing. The worst thing is that the well-intentioned Standard-writing do-gooders have only ever looked at half the story. They (naively) see a ‘bad practice’ problem and decide they had to fix it. But they never look at the other side of the coin. Standards do good things, but they also do bad things. Standards are, by their very nature Contradictions: ‘We want a Standard and we don’t want a Standard’. We want Standards (in theory) to weed out the ‘bad’; we don’t want them because they impede the ‘good’.

And in the case of ISO 18404, the specific ‘good’ I think they’re inadvertently about to kill is innovation:

Standards lock in today’s practices and impede advances. They stop people thinking, and as a consequence they become a ready-made excuse to not innovate. That’s what I see. I’ve had this argument several times with Standards people. ‘Ah, yes,’ they will say to me, ‘but we review the Standards every three years’. To which my answer is why three years? What happens when the world pulses at a rate faster than three years… as is the case in most industries today.

But even this is the wrong argument to get into – we should always know that if a question contains the word ‘or’ in it, it’s the wrong question. It’s not about should we review a Standard every three years or every six weeks. The right question is ‘how can we have a Standard AND not a Standard?’

Now, I admit that I’ve never met a Standards person yet that understands what the hell I’m talking about when I use this kind of TRIZ language. Sometimes I even try to tone my language down a bit. ISO18404 – like every other Standard before it – kills innovation because no-one has done anything to cater for the down-side. Maybe I don’t need to mention ‘contradictions’, maybe instead, all we need to do is get Standards people to think about the likely negative consequences of their actions and to build a solution to these consequences into the Standard.

Its not even that its hard to do this, it simply means that people have to give themselves permission to think about the down-sides and incorporate something into the Standard to prevent them from happening.

In the case of ISO18404, for example, it would have been very, very easy to include a twenty-fourth competence that ‘Good’ Lean or SixSigma or ‘Lean & SixSigma’ practitioners would have to demonstrate their abilities against:

‘Practitioners need to be able to demonstrate that they understand the limits of the Standard, and that they do not encourage or endorse client solutions that will impede the innovation ability of the client’.

That’s called ‘solving the contradiction’.

It’s time for another Standard, I believe. A Standard for Standards writers. A Meta-Standard. A Standard that prevents Standards from turning into value-destroying industries that serve only the officials they employ. It probably only needs two clauses:

  Clause 1: the system should be designed in such a way that it emerges in a progressively self-organising manner, that will eventually eliminate the need for the Standard.

   Clause 2: whenever an either/or Contradiction emerges that hinders the achievement of Clause 1, Standards Committees should devise solutions that eliminate such Contradictions.

The only downside, I suppose, is that it will kill the Standards-writing industry. And then where would the ‘bad’ consultants go and work?

The more I do work in the healthcare sector, the more sceptical I get about ‘clinical evidence’.

I started my career in the aerospace so I understand the idea of evidence. When aerospace engineers get things wrong, aeroplanes fall out of the sky, and that’s never a good thing. Evidence is what keeps planes flying.

A couple of year’s ago, after a particularly frustrating experience with a part of the UK NHS that will remain nameless for the moment, we did a few calculations to translate the current level of safety performance of the NHS into aerospace terms. The numbers came out a bit scary. So we did them again. And then again from a different direction. Expressed in mortality rates, the NHS, we concluded, is currently equivalent to a shade under 2000 plane crashes per day. Over UK air-space.

When we shared the data with this part of the NHS, asking them to check over our numbers to see where we were going wrong, surprise, surprise, we never heard from them again.

Clinical Evidence matters came to a head again a couple of months ago. This time with a different part of the NHS. This time in a workshop setting. We were talking about Complex Systems Theory. And someone – inevitably I now see – asked the clinical evidence question. Where’s your evidence that treating healthcare as a complex system is a better way of doing things?

For a few moments, I didn’t know what to say. Fortunately, the question came just before a break, so I muttered a no-doubt inane answer and tried to move on.

Over the break a made a new slide. Here’s a copy of it:

After the break, when everyone was back in the room, I put it up on the screen and asked for a hands-up vote. Everyone sat there, paralysed. ‘Any thoughts?’ I probed. Nothing.

I’d kind of anticipated the reaction, so, after I’d let the tension build a bit more, I advanced the presentation to a new version of the slide. This is it:

Now answering the question had become easy. Within a minute we had a collective answer: 10% A, 90% C.

It’s a fine line sometimes. And it’s difficult to know which side we’re on.

Is it better to treat a system as a system or not as a system? There’s a clue in the question, right?

Is it better to treat – say – a headache as a system or with a pill?

Is it better to deal with crime as a system or by longer prison sentences?

Is it better to deal with education as a system or by re-introduction of grammar schools?

Is it better, post the Brexit vote, to have experts or not to have experts?

At which point on the line do we make the transition from tautology to ‘we really don’t know, so we need to go gather some actual evidence’?

I have some sympathy with those that, per Michael Gove’s epoch-making statement, ‘have had enough of experts’. But the alternative is not to say, ‘oh, in that case, let’s listen to dumb, stupid people instead’, it’s to ask the question, ‘experts in what?’

Wherever we all might individually draw the tautology line, given the choice of treating a system as a system or not a system, there’s really only one sensible answer. And a good part of the answer to the ‘experts in what?’ question, therefore, ought to be, ‘experts in systems’.

So why then did 90% of the people in my workshop ignore the obvious? People say and do things for two reasons; the good one and the real one. Apparent absence of clinical evidence is a good reason for denying the need to treat systems as systems. The real reason, of course, is that the 90% of people that voted ‘C’ in my workshop simply didn’t understand what a system was, and so used clinical evidence as their get-out-of-jail-free’ card.

To me, anyone that doesn’t understand systems, probably shouldn’t be working within one, but that, unfortunately, would mean no-one could go to work anymore. Everything in life is a system. Life is systems. So, assuming we have to have people working within systems that don’t understand systems, that doesn’t also mean we can or should allow managers the same privilege. And there’s the problem in a nutshell. Not just in the NHS, but in all walks of life, 90% of managers or leaders have no idea what a system is. So what we end up with are a million and one ‘fixes’ that backfire: headache medications that lead to addiction and long-term digestive tract injury; harsher prison sentences that increase crime-rates; education standards initiatives that increasingly make students into dysfunctional members of society; homeless shelters that perpetuate homelessness; food-aid programmes that increase starvation.

When it comes to politics my main rule is anyone that wants to be a politician, shouldn’t be allowed to become one. My second rule is, whoever’s left over, is only allowed to become a politician once they’ve graduated Systems Theory class. My new rule, as of two months ago, is that what applies to politicians also applies to managers and leaders.

Every complex problem has a million simple wrong answers. If you’re lucky – if you can get to the core principles of the system – you might just find a simple right answer. Most people don’t get lucky because they don’t know how to get to the core principles. Most people don’t get lucky because they listen to supposedly smart people who also don’t know how to get to the core principles.

Take Frederick F. Reichheld, the man that wrote, ‘One Number You Need To Grow’ in 2003. What manager wouldn’t want to know what that ‘one number’ was? It was a sure-fire Harvard Business Review hit, and it spawned the monster we now know as Net Promoter Score.

The one number starts from one question. Even simpler. “On a scale of 0-10, how likely is it that you would recommend our company/product/service to a friend or colleague?”

It’s a simplicity that turns out to be flawed on so many levels it’s stops being funny after about five minutes. The laughter turns to tears when Frederick F Reichheld spotted that a good Net Promoter score was correlated to company share price.

Despite the fact that Reichheld himself eventually worked out that he’d fallen into the correlation-isn’t-causation bearpit, it was too late. Every Fortune500 company in every Fortune500 listing had taken for the bait, and so a whole industry of NPS surveyors found themselves riding a gravy train that still feels like one of the greatest gravy trains in the history of gravy trains.

The rules of the Hype Cycle – fortunately – tell us no ride goes on forever. Some will die outright. Those that have some underlying merit will prevail, once everyone works out what the underlying merit is. And, more important, how to meaningfully get to it.

Given the choice of knowing or not knowing whether I’m making our customers happy or not, the responsible side of me thinks I’d rather know. That’s the ‘underlying merit’ of NPS: knowing whether my customers are actually happy. Whether or not I can meaningfully know that my customers are happy or not, and – more importantly – knowing what actions I should take to make sure everything is moving in the right direction, becomes the critical question.

Answering it requires at least three things:

1)    I need to know that my customers are telling me the truth

2)    I need to know the local context within which they gave me their answer

3)    I’d like a measure of a customer’s ‘threshold for action’

Current NPS assessment methods fail on all three counts.

The first of the three is probably the easiest one to put right. Or, it is if you are using PanSensic and are able to map where a customer’s responses are on the 5Gs model:

Where you are on the map depends on a whole bunch of (measurable) things. One of them is your level of (over-)familiarity with NPS questions. Picture, if you can, the very first time you read that ‘how likely are you to recommend…’ question. You were probably in a restaurant. And it was probably part of a chain. Your surprise and lack of familiarity with the question probably meant you were intrigued. And very probably intrigued enough to do some actual, proper thinking about your answer. You were likely somewhere in the Golden middle of the 5G graph. Which was good for the restaurant. Now, bring yourself back to the present day, where it’s quite likely you’ve been asked ‘the question’ a couple of times during the last week. Not just in restaurants now, but on trains, at the airport, in the supermarket, at the mall, in the hospital. I even got asked it at a football game last month. Now you don’t think about your response any more. You probably have no inclination to respond at all. If you do feel inclined, it’s most likely because something extreme happened. Something extremely above-and-beyond-the-call-of-duty, or something horrendously bad. In neither case, though, are we going to answer the question objectively. That’s because the NPS question has progressively numbed our senses to the point where it has become meaningless the moment we see or hear it.

Local context is more difficult to capture, but still within the realm of possibility given the current range of different PanSensic lenses. Context, in terms of my likelihood or otherwise to recommend your products or services to my friends, has everything to do with the difference between correlation and causation. When I was asked whether I would recommend my friends to come and attend a game at the football club that asked me the question last month, the most sensible answer I could’ve given is ‘I’m an away supporter, I only came because my team is playing here.’ My actual likelihood of attending that club again – the causal link – is solely about whether they are in the same Division as my crappy team next season. I’m slightly ashamed to say that my actual answer to the questioner was, ‘yes, I am very likely to recommend your football club to my friends’. I watched as the questioner ticked the relevant box on her nifty Likert Scale. We were both happy. She was happy because she had something to correlate. The main reason I was happy, however, was because my crappy team had just beaten their even crappier team. If I’d been pushed any further, I’d very likely have made the request – as many of my fellow travelling fans chanted during the game – ‘can we play you every week?’ My reason for giving the answer I did, in other words, had precisely nothing to do with the way my answer was going to be interpreted.

So much for the situational aspect of ‘local context’. If you’re analysing the narrative around the answer rather than the score on the 0-10 scale it’s relatively easy to pick up this sort of situational effect. Ditto regarding whether someone is qualified to answer the question in terms of domain knowledge. I’ve been down this rant path before. Usually with things like Trip Advisor where the ‘reviews’ I read are usually written by people who stay in a hotel once in a blue moon and consequently have no way of saying this particular hotel is any better or worse than any other one on the planet. Their ‘review’ is usually – for the sorts of hotel I tend to stay at – based on a comparison between their experience and an advert they saw on TV for a seven star hotel in Dubai. i.e. fiction piled on fiction. “The taps weren’t even gold, two stars.”

The third aspect of ‘local context’ is the context of the person I might consider recommending your products and services to. I like music. I own a building full of records, tapes and CDs that aren’t going to be digitised and disposed of any time soon. When visitors see my music collection, they usually ask me to recommend something to them. A question that I can only usefully answer provided I know something about the sort of music they currently like. And how far in or out of their comfort zone they might want to go. And how much I want them to come back and ask for more recommendations in the future. It’s rarely as straightforward as saying, ‘Blue Nile, Hats’, although, I know that particular recommendation will work more often than it won’t, and if it doesn’t work, the visitor won’t be invited back very often in the future anyway.

Last up is ‘threshold for action’. This is the most difficult of the meaningful-NPS desire foundations. To an extent we know it boils down to the strength of the adjectives that people use to describe their experiences. Or rather the relative strengths.

Back to my football match. At half-time I – unusually – decided to go and get a cup of tea. There was a queue. I was stood behind a father and son duo. The son looked like he was about eight, and sounded like he hadn’t been to an away game before. Everything, therefore, was ‘awesome’. The journey to the ground had been awesome. The sandwiches were awesome. The floodlights were awesome. The two goals we’d scored were awesome. The late-night return home was also going to be awesome.

He was basically me forty-five years ago. Now I know that most things aren’t awesome. Our second goal, as it happens, was pretty awesome, but the first one was an umissable tap-in following a bad mis-kick by one of their defenders. It was never going to win any goal-of-the-season competition any time soon. Likewise, my cup of tea, when I eventually reached the front of the queue, tasted like it had been brewed a fortnight earlier, and had long past its moment of awesomeness. It doesn’t take much for an eight-year old to see that everything is awesome. For a cynical old man, awesome doesn’t happen very often any more. When it does happen, though, it probably means more in terms of useful feedback to a company than when they hear the same adjective come out of the mouth of the eight-year-old.

You need a lot of an individual’s narrative in order to calibrate their adjective use in order to work out where they need to be on an adjective-strength spectrum before they will act – i.e. recommend to their friends or family. Getting access to sufficient of this narrative is ‘possible’ if you have access to, say, the Facebook narrative of a smiley Millennial, but very often the people that talk the most are the ones with the least to say. In which case, the current ‘best’ way to calibrate the how well a customer thinks about you is to calibrate across many customers. Even better, thinking about the PanSensic ‘Mental Gear’ lens, is to calibrate across Blue, Orange, Green and (especially) Yellow customers, and then – most crucially of all – close the loop by finding some actual customers that actually did recommend you to their friends and family and examining their actual collective PanSensic profiles.

NPS is not as far along the maturity scale as other Weapons of Mass Distraction (Balanced Scorecard, SixSigma, QFD, PRINCE2, Scrum, Agile, etc). Unlike most of them, it at least has a valid start-point in that measuring customer emotions is a fundamentally good thing to do. Whether it gets to survive in the long term is largely dependent on how quickly it can evolve to a point where the measurements it’s is used to make are meaningful – as in truthful, context-relevant and actionable. Which is hopefully where PanSensic comes in to play.

“Incontinent the void. The zenith. Evening again. When not night it will be evening. Death again of deathless day. On one hand embers. On the other ashes. Day without end won and lost. Unseen.” Samuel Beckett

We use the expression ‘successful step-change’ to define innovation because it takes us right to the heart of the world of s-curves. The ‘step-change’ in question being the jump from one curve to the next. We usually draw the s-curve-jump story to look something like this:

There are no hard and fast rules about the relative positioning of two adjacent s-curves, but we know for sure that a big part of the innovation management job is successfully managing the gap between the two. And specifically the void defined by this area:

Whenever an enterprise embarks on an innovation project, they essentially make a decision to jump off the top of their current curve into the unknown of the coming ‘next world’. However we choose to measure the vertical axis defining the s-curve, this jump almost inevitably results in things looking and feeling worse than we felt when we were safely ensconced on the top of our current cliff. The trick to our descent and then – hopefully – rise up the next s-curve is to reach the point where the height we’ve ascended to on the new curve reaches the level we’re at on the top of our current curve. This is what the shaded area is all about.

Managing the Void, then, is all about minimising the area of this shaded region: the smaller the area, the faster everyone accepts that the innovation attempt being made has been successful. Or, put another way, the smaller the void area, the less time and money we have to invest in getting the project out of the red and into the black.

Given the fact that 98% of innovation attempts still end in failure, it is probably fair to say that the majority of enterprises on the planet are uncomfortable when it comes to managing the void. Most enterprises still see the world through Operational Excellence eyes. Which means they focus on improving what they have, rather than jumping off a cliff into some mysterious, unknown ‘void’.

As far as we can see, when we look at the 2% of innovation attempts that end in success, there are five basic strategies for managing the Void. None is mutually exclusive, and if we really were ‘managing’ our s-curve jump we would no doubt look to adopt as many of the five strategies as makes sense given our available resources. Here are the five in graphical terms:

And here’s what each of the five means in practical terms:

1)    ‘Start Earlier’ – in many ways the easiest of the five strategies to engineer and manage, but, alas, in most enterprises, the strategy that gets adopted the least often. One of the best ways to manage the step-change from one s-curve to the next is to start work before we reach the top of the current s-curve. Starting to look for the new curve when we’re on the steepest climbing part of the current curve makes the most sense since that’s when our margins and cash-flows are at their strongest. The problem, however, is that most organisation leadership teams think that the best way to manage their affairs during the exhilarating climb is to put all of their attention on the job of maximising margins and revenues. Nobody, it seems, wants to be seen to be the killjoy that tells everyone that the future won’t always be so rosy, and we should start planning now for future rainy days.

2)    ‘De-Traumatise’ – when people are stood on the top of their current s-curve, it means they’ve done an awful lot of hard work to get there. Their system has been massively optimised, everything has been worked out, and everyone has settled in to their comfort zone. No matter how good the new solution that kicks-off the start of the next s-curve is in reality, it looks and feels worse to everyone who looks at it from the perspective of the beautiful position of the existing system. Managing the Void strategy 2) is thus all about managing the psychology of change and the innovation version of the Kubler-Ross Grief Cycle – first we experience the shock of the new, then we deny it, then we get angry, then we try and bargain our way out of the difficult situation, then we get depressed, then we accept a little realism by testing the new, until, finally, we accept the change, and realise, it’s not as far down as it looks. ‘De-traumatising’ is about letting people still safely positioned on the old s-curve see that getting worse to get better is sometimes just how the world works.

3)    ‘Climb Faster’ – the strategy that sits at the heart of ‘Lean-Startup’ and Design-Thinking – methods and processes that are all about making lots of rapid iterations of the new solution, exposing them to customers, learning from their reactions, and building new iterations as swiftly (and as cheaply) as possible. ‘Fail-Fast, Fail-Forward’ is the frequently heard mantra of teams comfortable operating in the Void: we can’t know everything right now, so our job is to try new things, learn from it and use the learnings to design the next iteration of the solution.

4)    ‘New Measures’- perhaps the least immediately visible of the five possible strategies, the ‘new measures’ strategy is about recognising that in a large majority of cases, when innovation happens, it comes alongside new ways of measuring ‘success’. When JCB first invented the hydraulic earth-mover, for example, it was significantly inferior to the industry-standard cable-driven earth-movers in terms of earth-moving capacity, but what JCB understood before anyone else was that ‘value’ to a customer wasn’t just about how much earth could be lifted in a single bucket-load, it was also about how easy it was to get the earth-mover on-site, how manoeuvrable it was once it arrived, and how flexible it was. It was ultimately about earth-moving productivity, and if you spent a week less time getting the earth-mover on site, that more than compensated for the inferior bucket load. None of which were measured by the cable-driven earth-mover companies. When JCB showed earth-moving contractors how to re-frame and re-define their success criteria, that was when their business really took off.

5)    ‘Unlearn’ – In some ways analogous to strategy, 2), but the focus in the ‘unlearning’ strategy is to get people to recognise that being on top of our current cliff is not as good as we think it is. It means key parts of the learning to get to the top of the current curve has become ‘waste’ and therefore needs to be thrown away. No-one likes to think of devoting lots of hard work to create what looks like irrelevant outputs. Until such times as everyone understands that this kind of ‘un-learning’ is an inherent part of the step-change process, s-curve jumps are always going to appear difficult. Managing the Void strategy 5) is therefore largely about educating people to understand this is how the world works – fundamental means fundamental – and that we periodically unlearn stuff in order to protect the future stability of the enterprise.

Operational Excellence versus Design Thinking. Compare And Contrast.

The Regional Education Board Operational Excellence Sense Radar hears a rumour that the budget for next year is going to be cut again.

The Management, as ever, moves swiftly and launches an investigation. Being a ‘budget’ problem, the investigation is handed directly to the Accounts team, with an ‘urgent’ priority code. The team jump to action and ‘run the numbers’. True to form, they do this very quickly. The message back to the top is, ‘we’re in trouble’.

Management asks for options.

The Accounts team run some more numbers, draws up an elegant cost-per-pupil distribution curve for all the schools in the region and come back with three options. The way they’ve been taught. Two of the options are visibly ridiculous – also the way they’ve been taught – and the other one involves closing the two worst schools on the distribution curve. They present their findings to Management. Management asks about the two worst schools. It turns out – no surprise – that they are the two smallest schools in the region. This is good news. Closures always mean protest, but closing the two smallest schools means the smallest amount of protest. The Management team declare themselves happy with the analysis and, being dynamic thrusting types, they announce the decision to their masters. A week later, the story goes public.

Two weeks after that the first lawsuit arrives. From one of the parents at one of the two soon-to be-closed schools. ‘How can it possibly be’, the suit charges, ‘that the school with the highest academic record in the region is going to be closed?’ Management look at the letter and do the only sensible thing. They pick up the phone, dial the Lawyers and tell them, ‘we have a problem for you to come and fix’.

So much for Operational Excellence thinking.

It’s a story based on a real situation. At this point in time it is ongoing. The lawyers are hard at work sending each other letters. On one level, we don’t as yet know what the outcome will be. On another, several things are already crystal clear:

–   The outcome will be win-lose. Either the Education Board will win, or the parents will win.

–   The teachers and pupils at the school will be caught in the middle, unwitting victims of a battle over their futures.

–   The lawyers on both sides of the argument have no incentive to make a swift resolution. The longer the fight goes on, and the more acrimonious it becomes, the more money they will make.

On too many levels, it is a depressing story. Perhaps the most depressing part is how quickly the Operational Excellence-driven legal downward-spiral took hold.

Here’s how things might’ve played out if the Operational Excellence blinkers had been removed from Management’s eyes:

As part of their ongoing search for contradiction-solving opportunities, a member of the management team looks at the latest cost-per-pupil distribution curve prepared by the Accounts Department, and realises this would make a pretty good contradiction to try and solve at the next Management design-day. She draws the contradiction up in a way that will bring some structure to the discussion:

At the next design-day, the team spend a few minutes looking at the picture and someone shouts out, ‘this would be a great one for us to get all the parents, teachers and officials together to see what win-win solutions we can come up with.

Two weeks later, forty people turn up to a Saturday morning ideation session. They quickly agree on where they all ideally want to get to. Then spend an hour writing down all the reasons that might prevent the ideal from being achieved. That list then got turned into a perception map, which revealed what the key barriers were. Now capable of seeing where they were trying to get to and what was stopping them, they spent the last hour of the session working in small teams to generate solution ideas. They filled a wall with Post-It notes, clustered them, and then gave everyone three stickers so they could vote on their favourite ideas. Some people voted on things that could be done quickly, some on things they volunteered to take away and work on for the next semester.

Back in the office the following week, the Management team was still buzzed at the excitement and passion of the parents and teachers from the Saturday session. They heard a rumour that the education budget cuts next year were going to be bigger than ever, and smiled. Taking 75% out of the Legal budget suddenly seemed like a no-brainer.  

There’s been a lot of discussion in the Big Data Analytics community recently relating to the accuracy of the analyses providers are delivering to their customers. For the providers the discussion has rapidly devolved into a ‘mine’s bigger than yours’ race to be able to claim 100%. It’s not uncommon to already see numbers in the mid 90s percent. Which sounds good. At least until we start to examine the dysfunctional nature of the industry: lots of money being spent on analyses, but almost no apparent tangible benefit being delivered.

How can it be that 95+% ‘accurate’ analysis capability produces no real impact? Does it mean that all the benefit is in the final 5%? Or that the industry has defined ‘100%’ incorrectly?

‘100% of what?’ feels like a good place to start an exploration of the subject.

The answer, as far as I can tell, is something like ‘not a lot’.

Just because a BDA algorithm can pick out keywords, and synonyms and make some kind of semantic context check, and – as some of the most advanced algorithms are now claiming – to be able to identify ‘fakes’ (e.g. false reviews planted by robots), does not mean that what we end up with is ‘100% accurate’. At least not in any meaningful way. Anyone acting on this kind of ‘100% accurate’ analysis is as likely to make the wrong decision as they would have done having acquired no data.

‘100% accurate’ in the current BDA context turns out to actually mean ‘100% accurate assuming the world works in purely tangible ways’. Computers and data analysts love tangible things. Mainly because they’re easy to measure.

But 100% tangibly accurate has nothing at all to do with 100% meaningfully accurate. People are emotional creatures. People make decisions for two reasons: ‘the good reason and the real reason’. Tangible analysis is all about capturing the good reasons and nothing at all to do with capturing the real reasons.

If we’re to capture what drives peoples’ behaviour the analytics need to delve deeply into the world of intangibles because this is where we find all the ‘real reason’ stuff. Things like:

–   Does the data come from a person who is psychometrically relevant to my target audience (e.g. if you’re trying to test a mass-market toothbrush design and all the product reviews you’re analysing are coming from Feudal-thinking, ENTPs, they’re not going to tell you anything useful at all about the future mass-market appeal of your design)

–   Does the data come from a person with a relevant opinion about the subject? See my earlier TripAdvisor case study – is it sensible to listen to the comments of a person that stays in a hotel once a year? Is it sensible to listen to the comments of a person that tends not to be listened to by other people? Sometimes, maybe it is (if we’re designing products for dimwits), but the important point is that I would be well advised to understand the difference between the two and listen to only the relevant people.

–   Does the data come from a person who is speaking reliably about the subject – are they telling the truth in other words or are they playing one of the 4Gs game:

–   Can the analysis identify that the person’s behaviour is going to be consistent and congruent with what they’ve said. People often say one thing and then do something completely different. Back to the good-reason/real-reason dilemma, ‘congruent’ data means data that has successfully captured the between-the-lines ‘real-reason’ content.

Only when an analysis capability is able to achieve these four intangible things – Representative-Relevant-Reliable-Congruent – should we be talking about ‘100% accurate’. 100% accurate, to my mind, means we’ve accurately captured what people mean rather than what they’ve merely said.

‘Sheep in Fog’ is the metaphor I’ve been carrying around for a while now when I look at the innovation activities taking place – or not taking place – within a majority of organisations around the world right now: very little bravery and even less direction clarity.

It’s also, as it happens, one of my favourite poems by Sylvia Plath. Well, ‘favourite’ is probably too strong a word. ‘Admire’ is probably better. The poem is overwhelmingly bleak and I’m only glass-half-empty-level bleak. I’ve always read it as a list of metaphors describing how she felt about the world at a particularly difficult time in her life.

Only lately have I come to connect my innovation metaphor to the poem. The more I think about the connection, however, the more I think Plath’s words tell us about the ‘Hero’s Journey’ from an innovator’s perspective. Four things stand out for me.

First, she uses personification – the stars ‘regard me sadly, the train has ‘breath’, and the fields ‘threaten’ her. All of this creates a sense that nature pities her, or finds her presence problematic. She does not belong in it. Plath as the prospective innovation Hero, and nature as the ‘efficiency engine’, everyday world she unwittingly finds herself in.

Secondly, she uses enjambment, a poetic device in which a single sentence is broken across two verses. Her discussion of the horse in the second stanza extends into the third, while the discussion of the morning is split between the third and the fourth. The technique is used to great effect in the poem to link all of the disparate metaphors together to create a profound sense of estrangement and uneasiness. While I’m sure Plath had no overt conception of s-curves and the idea of the innovator as the navigator between one s-curve and the next, it feels that, somehow, instinctively, she did. The enjambment makes an awful lot of sense, in other words, as a representation of the discontinuous jump between the current world and the next:

Thirdly, the title of the poem references how Plath (the ‘innovator’) feels – a lost sheep wandering in a murky and meaningless world. She feels like she continually disappoints those around her, all while she sees the world blackening. There is a clear paradox here in that she calls this terrifying place a ‘heaven’.  This dark, fatherless heaven was used by Plath as a telling metaphor for her personal life, but perhaps it makes for an even more powerful metaphor to represent the ‘Ordeal’ of the Hero’s Journey? This ‘dark heaven’ is the Contradiction.

The Hero’s Journey stage connection follows, too, fourthly, when we step back and connect the poem to Plath’s life. Plath didn’t prevail over her Ordeal. In the Hero’s Journey, after the Ordeal, something has to die. Plath killed herself a month after the final changes she made to the poem. She never made it out of the fog. In this regard, the poem is certainly bleak and hopeless. But then again, perhaps, all the more sophisticated because it captures, in a very few lines, a profound ambivalence towards death. It is in this regard too, I think, a telling metaphor for the life of the innovator, and the (98%) likelihood that the innovation-sheep don’t make it out of the fog either.

Call it a warning. Or a roadmap. Or, maybe, just a call to arms.

In that regard, finally, I find much to think about in the final changes Plath made to the original draft of the poem in those final weeks of her life: