Jun 25, 2025 | AI, Customers, Lean, Operations
The idea of the OODA loop is to get inside the decision cycle of your opposition. Once inside, you control the outcome in the absence of some externality.
Toyota used this idea to destroy Detroit.
The Andon cord placed the power of tactical decision making about quality right at the point where it was needed, with the workers on the production line.
By this means, quality problems were identified and fixed before they moved a further step towards the customer.
It also did something else.
By identifying and fixing problems at the source, the cycle of problem fixing was accelerated greatly. Not every problem can be fixed immediately at the line, but there are processes for escalation, from the front lines to the lowest level that is empowered to address the problem. That escalation involved suppliers when the problem was caused by a supplied part that was substandard.
By contrast, Detroit was driven from the top down, being run by spreadsheets (handwritten until the 90’s) by executives who may never have seen the inside of the factory.
A problem as it escalates up a chain of command has many opportunities to be buried, forgotten, miscommunicated, all of which will happen, driven by all sorts of human frailties and power games. The end result, the little problem in the factory compounds and becomes a big problem with customers, which costs a lot to address, and ruins reputations.
Toyota got well inside the time it took Detroit to respond to problems. While Detroit was escalating or hiding quality problems, Toyota was fixing them and moving on the next improvement.
They were inside the OODA loop of Detroit, and it destroyed the American car industry.
AI is now giving users an easy tool to get inside the decision cycle of their competition, while seeing the productivity benefits drop to their bottom line.
How are you going to deal with that?
Jun 2, 2025 | AI, Communication, Governance
Analysis implies the intelligent interrogation of data, the use of differing ‘frames’ through which to see the data, and to enable those non obvious connections to be made.
First you need ‘clean’ data, without which, nothing that follows will be worth much.
Thoughtful, critical analysis of data leads to insight, from which comes that elusive lightbulb moment.
Reporting is the opposite, it simply requires the cleaning, summarising and posting of the data without the critical thought from which real insight evolves.
No lightbulb.
AI is good at that, while not being good at generating insight.
Being data rich but insight poor is now a very common problem.
AI will not solve it for you, people are needed. Not just any people, seat warmers, but the right people with the curiosity and ‘why not’ attitude of youth, combined with the wisdom of experience and domain knowledge.
Unfortunately, these people do not grow on trees, ready for the picking. You have to grow and nurture them yourself, while recognising that many will move on at some point. The old adage of a rising tide lifts all boats is nowhere more relevant.
May 30, 2025 | AI, Marketing
The market research industry turns over big dollars by providing reassurance to marketers that when they are wrong, they have an acceptable excuse.
The recent election campaign and associated polls demonstrate comprehensively how broken market research can be. It should have been simple. Is Labor going to win, will the conservatives take the lollies, or will it be a split result? It was almost a binary choice, but no poll I saw was even close.
Given that failure, how in heaven’s name can we reasonably expect such a broken system to deliver reliable answers to challenging questions about the future behaviour of customers and potential customers in a competitive and volatile environment? Add into the mix the inability of most marketers to understand the competing forces in their market sufficiently well to ask good questions and therefore write quality research briefs. That delivers a perfect recipe for pissing money against the wall in pursuit of reassurance.
Over the years as a much younger marketer, I spent a lot of money on market research. It took a while, but I did come to realise the data was only a tiny proportion of the game. The real challenge was building the wisdom, insight, and market gutfeel to be able to ask those really good questions. Then, when a surprising response emerged, have the curiosity to further interrogate it from a positive and genuinely inquisitive perspective to get an answer to the eternal question: Why is it so?
In the last year or so, and accelerating at an astonishing rate, is the ability delivered by AI to gather and process market and behavioural information that can be used to ask those challenging; why is it so’ questions.
The process of market research has been totally up-ended. No longer do you need to spend tens or even hundreds of thousands of dollars over months to get shallow and often dated results. Now you can replicate the task quickly and cheaply with superior results using AI, and what is emerging as ‘synthetic research’.
Traditional research is good at telling you what has happened. It is good at counting. However, if you need to know what will happen in the future, and you use yesterdays tools, good luck!
May 26, 2025 | AI, Governance, Strategy
Innovation using physics is forging ahead at an accelerating rate.
Remember the speed at which a covid vaccine was brought to the market after the first identification of the virus. Instead of the usual 10 to 15 years we suddenly had that process compressed into 18 months.
And yet there remained those who refused to accept the vaccination for a range of personal and behavioural reasons which many would say are irrational.
Somewhere the line between the technical innovation involved in the hyper-rapid final stage development of the vaccine and the humanities driving behaviour crashed into each other.
As the rate of technical innovation across every domain accelerates it is likely we will continue to stumble across this barrier to adoption, and a fragmentation of adoption across a range of behavioural parameters.
Simply another social tension driven by the speed at which the modern world is evolving. It is way beyond the speed at which our DNA allows behaviour and attitudes to evolve.
The situation in front of us right now is the degree and manner in which AI is accepted and adopted by organisations and by individuals.
We managed this dilemma in the motor industry as it became obvious that it was profoundly important to incorporate safety into the vehicles as a means to save lives. As a result, it became mandatory to design crumple zones into cars, and install seat belts. Regulatory intervention and oversight 60 years after it became obvious that a car could kill its occupants.
Where will the equivalent crumple zone emerge in the arena of AI, and will it be in time?
Apr 28, 2025 | AI, Change, Governance
American Roy Amara first coined what has become known as Amara’s law.
‘We tend to overestimate the effects of technology in the short term, and underestimate the effects in the long run’.
‘It was put more simply by (I think) Reid Hoffman who said: ‘the future is like a windscreen coming at a moth at 100mph.’
The initial excitement, hype, enthusiasm for the idea is followed by a period of underperformance, and disillusionment, before the real impact of the technology kicks in and changes the way we do things. Gartner’s well thumbed ‘Hype cycle’ is a better known version of Amara’s law.
Time and again over the last 30 years we have seen this effect on vivid display.
The internet, smartphones, AI, electric vehicles, Hydrogen as an energy source, (just entering the disillusionment stage) and many others.
It can also be applied to wider contexts, we just need to look for it.
Advertising.
No new TV ad campaign was ever released into the world without exalted expectations about the sales that would result coming from the ad agencies and those often clueless advertisers paying the freight. Then, unexpectantly, the ad is shown to be a dog, and is quietly euthanised.
Climate change.
Remember the hype and enthusiasm for ‘doing something’ that accompanied Al Gore’s influential doco ‘An Inconvenient Truth’ back in 2006. Nothing happened, the hype and enthusiasm was drowned by hubris and short term individual, corporate and political self-interest. While it seems unlikely at the moment, I remain confident that realisation will hit soon that we must take remedial action now in order to mitigate the long term becoming worse. Meanwhile. continuing to do nothing more than provide lip service ensures the moth will hit the windscreen in my grandchildren’s lifetime.
Business.
There are cycles of ‘fashionable’ management frameworks that seem to come, become the next great management breakthrough, undergoes the hype, then is shown to be np more than an emperor dressed in some transparent new clobber. Sometimes they re-emerge rebranded to go through the process again. Michael Hammers 1993 book ‘Re-engineering the corporation’ was such a fashion. I recall sitting around a board table listening to a very slick but hollow (even obvious to me at that time) presentation by a high priced consultant making promises of easily won great profit improvements from an aggressive ‘re-engineering’ of my then employer. That business hit the windscreen several years later, having cherry-picked the easy bits of the process, while ignoring those that actually made the long term difference because they were too hard. A few years later, Al ‘Chainsaw’ Dunlap had another run at it which made him a fortune, but left chaos in his wake. There are many more examples, the fall of GE from the largest corporation in the world to being virtually broke being one.
Politics.
Governments are relentlessly hyping the impact of their latest policy, more intensely than usual around election time. They whip up enthusiasm, at least amongst their acolytes, then falling into the trough of hubris. Usually, there is a renewal under a different name at a later time, often the next election. Remember the ‘Gonski reforms’ to education hyped by the then government, and supported in principle at least by the then opposition? Swept under the carpet of hubris and self-interest, again. Similarly, the 2010 Henry tax review was received by a grateful government who then shelved it. We may now have reached a point where the dust will be partially removed by necessity.
Americans are in the midst of waking up, again, to the reality of a second Trump administration. My contacts over there indicate dismay bordering on horror, and most of the working class Trump voters are about to learn the cost of the hype to them. The US moth seems likely to be splattered over the windscreen by the 2026 mid-term elections.
Artificial Intelligence.
Occasionally, the outcomes go way beyond what was originally envisaged. AI has been evolving for decades, but it exploded into the wider public awareness when ChatGPT was launched in November 2022. We are still experiencing the upswing in the hype cycle; I am certainly playing my small part. However, at some point I suspect soon, the tsunami of tools emerging, the sheer complexity of choice being forced on us will overwhelm all but the few, and we will collectively throw our hands in the air when the robot that does our washing does not appear. This collective action, if that is the way it occurs, will just let the first movers race away with the lollies.
The hype cycle remains around us, daily impacting on our lives. Its greatest risk is that we let it drive our decision making by making short term choices that are strategically flawed.
Apr 24, 2025 | AI, Marketing, Strategy
‘Lean thinking’ is a mindset and toolbox to drive optimisation. Little more, beyond the use of common sense and humanity.
Prominent amongst the tools, and the one I probably use the most is ‘5 why’.
AI has given us an entirely new use case that leverages the insights that a 5 why process when done thoughtfully can deliver.
Prompt development.
There are now hundreds of prompt templates and mnemonics emerging from the woodwork, many claiming to be ‘the one’.
All I have seen use a variation of the Lean ‘5 why’ tool.
Most AI users look at the first output of a prompt into any of the LLM tools, and it is sub-par. Generic recitations of what the trained information base reflects as best practice. The beauty of these data driven assistants is that you can push back as much as you like without them taking it personally.
You can point out areas of failure, misinformation, gobbledy-gook, or imagined fairy tales. You can ask for specifics, deeper analysis, sources, or give it examples. The output then improves with each iteration.
You can also ask it what you might have forgotten to ask, or has been missed for some reason, and ask for suggestions. This interrogation of the tool can reveal things you would not have thought of under normal circumstances.
Go through that process 5 times, and in all likelihood, you will not only have something entirely different to the first response, but it will also be infinitely better, and tailored to the need. You will have cleared away the unnecessary, banal, insignificant, and generic, leaving a response that equates to a first principle response to your evolved prompting.
Continuous improvement by AI driven lean thinking.
What a boon!
