The KPMG senior management obsession with predictive technology appears to be finally satiated. Last week's announcement of an alliance with the Formula 1 team, McClaren, is intended to help the big 4 firm evolve from its current “retrospective and subjective” business model to become “forward looking and predictive.” Which is great news - if it wasn’t such nonsense!
The desire to know in advance of chaotic shocks that could rock the firm or its market place is understandable. Yet having more data, even of the turbo-charged variety, seems less a breakthrough in predictive capabilities and more an attempt by a large organisation to differentiate its commoditised offering in an increasingly fragmented market place. “Multi-dimensional” data - the influence of weather, pit stops and wheel changes - may be able to predict the performance of a complicated machine, like an F1 car, in a known environment (a limited term event operating to clearly defined rules). But KPMG - and the clients its advising - are complex entities operating in deeply uncertain ecosystems where the rules change at random and prediction is not impossible but can be dangerous if its believed.
Complex entities, like organisations or markets, are made up of thousands of people engaging in an almost infinite number of interactions that are impossible to calculate. Furthermore, these interactions are non-linear - meaning small causes have large effects (e.g. US sub-prime mortgages and the global economy). Though events often seem obvious in hindsight (e.g. sub-prime mortgages look a bad idea now) this doesn’t lead to foresight as a complex ecosystem is constantly emerging and new patterns make past pattens obsolete. This is why generals don’t win wars by re-fighting the last one.
The futility of relying on past data for future planning was perhaps best exposed when Alan Greenspan gave testimony to Congress in October 2008. He expressed “shocked disbelief” at the global crisis unfolding because for “40 years” there had been “considerable evidence things were going exceptionally well.” The 2012 collapse of Monitor Group, the elite strategy firm of Michael Porter - the ‘father of modern strategic analysis’ - also demonstrated how rigorous analysis may explain success in hindsight but still fail to predict it.
A complex world, by definition, is unpredictable and more data doesn’t alter this reality. KPMG therefore faces a real danger of unintended consequences in adopting ‘sexy’ technology: complacency (believing the models to be right), retardation (why think for ourselves now) and ultimately worse decisions (as the outputs will only be as good as the inputs chosen and are supposedly more experienced partners going to learn how to do that or push it down the chain?). The result of a turbo-charged stream of data may be to merely obscure the key signals KPMG and its clients need under even more noise.
If KPMG truly wishes to distinguish its offerings there are better ways than embarking on (another) long, seductive foray into technology they don’t understand. Within the organisation’s engine room sits (probably 87% idly) its most sophisticated capabilities: auditors who’ve completed thousands of audits and can recognise emerging patterns of system failures and potential improvement paths. “How do you report those insights to decision-makers in the organisation” one such auditor was asked in an interview. “We don’t” was the answer. “Why not?” they were asked. “Because my manager has told me to keep my damn mouth shut” was the reply.
So, before you buy the latest snake oil to solve your organisational problems consider what underused capabilities you already have that you can bring to bear on your strategic challenges. For any increase in the amount of data being collected must be coupled with the human capability and desire to seek insights and create action. Predictive technologies alone won’t turn a complex world, prone to eruptions of chaos, into an ordered one. But while it may not be possible to make an unpredictable world predictable - it is feasible to make your organisation better adapt to this reality through harnessing the natural sense-making talents of your people.
Leaders have a unique challenge in the 21st century. The ecosystems (the countries, markets and industries) their organisations operate in are increasingly volatile, uncertain, complex and ambiguous. And missing critical signals amongst the increased noise risks exacerbating existing fault lines in their organisation. What should leaders do?
Subject matter experts solve complicated issues at functional levels. But complex questions (e.g. top line growth, corporate culture, change or risk management) cut across specialised silos. Complexity therefore is always escalated up; making managing complexity the key strategic challenge for executives in the 21st century.
Yet complexity remains widely misunderstood; described as something ‘very complicated’ or confused with chaos theory. Complexity science itself ascribes distinct characteristics (non-linearity, emergence, unpredictability) that render traditional 'solutions' (e.g. best practice, ideal leadership styles) entirely context dependent.
To face a complex issue means to deal with a ‘brownfield’ context - never a ‘greenfield’. Complexity is located in the system (e.g. the organisation, market, population) and always has a history, yet is constantly evolving. And as we engage with it, it changes - often in unpredictable ways. This is why answers in a complex system often appear only in hindsight - though this doesn't lead to foresight (e.g. it seems obvious now that sub-prime mortgages were a bad idea, but likewise, today's Quantitative Easing is variously described as the only thing saving the global economy or creating an even bigger future crash. Only time will tell).
In a complex world context is king.
Idealised futures are an illusion - as are strategies based on certainties designed to get you there. The collapse of Michael Porter’s Monitor Group in 2012 showed that while rigorous strategic analysis can help explain how excess profits were created in the past, it's a poor predictor of how to generate them in the future. Even the best formal strategising can trap leaders into believing the future will be an extension of the past. But if the future fails to conform to expectations we are left naked and fragile, exposed to the elements.
Like King Canut ordering back the tide, we discover powerful natural forces defy command and control.
Yet, the natural world itself - of which man is a part - has adapted wonderfully to exploit complexity. Evolution works through a process of increasing variation (of options), basing selection on what works now, and replication (or starvation) of options based on hard evidence of suitability. Can leaders learn something from nature about adopting a rigorous external focus, increasing awareness of options through rapid trial and error, and creating mechanisms to amplify or dampen options in order to thrive?
Effective horizon scanning uncovers emerging signals that signal where and what to act on before its too costly or too late. Technology is a great enabler in this, if one caveat is kept in mind: technology without human interpretation is meaningless. Google may find anything you ask, but can’t tell you what to ask for. Uncoupled from humans technology merely increases the noise surrounding the signals. Data is dumb - to become meaningful information human knowledge must be applied.
Humans should be at the front and back end of technologically-aided decision-making - defining the issues to explore and discovering its real meaning. Technology therefore must be designed to fit the human - the way we are now, rather than the way we'd like us to be. It must augment our natural sense-making abilities, which have supported human evolution through millennia (a best practice case?).
Critical knowledge flows through organisations in human networks. Navigating these flows effectively can reveal the origins and dynamics of change. And as humans share such knowledge naturally, extrinsic rewards aren’t required to tap this. Humans naturally create and share knowledge in the form of narratives - ‘micro-stories’ - that are both universal (every culture has them) and democratic (no barriers exist to sharing). These are the 'water cooler’ stories that spread insights and enable other people to make sense of the world around them so they can act better in it. Harnessing these narratives is critical to making sense of and navigating complexity.
Critical knowledge can be leveraged at little extra cost.
Leaders must create the conditions for contextually-appropriate knowledge to emerge. Managing for serendipity (‘pleasant surprises’) means seeking fresh insights, rapidly field-testing coherent ideas and replicating success. But as genuine breakthroughs don’t come from established thinking patterns. Leaders must learn how to break through the hard-wired autonomic brain we rely on - which seeks first-fit, rather than best-fit, solutions - and instead become receptive to novel ideas. Strategic leadership is less about engineering future visions than it is about increasing awareness of the critical factors in our ecosystem, 'identifying the biggest challenges in them and devising coherent approaches to overcoming them'. Real strategy is about seeking the truth of the current position.
Navigating and exploiting complexity means leaders must take multiple perspectives to discover genuine insights. Going beyond objective numbers to understand the why. Rapidly testing coherent ideas as ‘safe-to-fail’ experiments. and feeding success, whilst starving failure of resources. No-one can ‘cut through’ or ‘simplify’ complexity - nor should we want to. Complexity contains rich opportunities in a changing world. Leaders employing naturalistic approaches can exploit complexity profitably.
SenseMaker® - an innovative technology first deployed by the Singapore government to detect weak signal terrorist threats - taps into mass organisational knowledge flows and helps join up disparate information from silos to form actionable knowledge. It presents whole network perspectives leaders can rapidly see and understand, helping unlock the organisation’s present evolutionary potential.
For more information about how to make sense of, navigate and exploit complexity for organisational success contact firstname.lastname@example.org
“Google can answer almost anything you ask it, but it can’t tell you what you ought to be asking”. Technology is immensely powerful, but severely limited. It’s not the amount of data that creates competitive advantage but how decision-makers make sense of it that counts. Genuine breakthroughs (innovation) come from a symbiosis of humans and machines.
A complex world compels organisations to rapidly discover new solutions to swerve or recover from increasingly inevitable ‘black swan’ visitations. But “historically” writes the internet expert Clay Shirky “we have overestimated the value of access to information, and we have always underestimated the value of access to each other.” The failure of the US security services, for example, to prevent the 9/11 attacks were not due to a deficiency of data collection but of humans beings sharing knowledge about what had been collected in cross-functional teams to make better sense of the data. If organisations want to become smart and resilient they must leverage natural human skills in parallel with technology. Less focus on big data; more focus on ‘big knowledge.’
Evidence from millennia of evolution testifies to the natural human ability to prosper. And we do this through social interaction, with each other, using technology as an enabler - not a decider. For example, the recent discovery of an unusual new planet with four suns was not made by the super hi-tech Kepler telescope alone, but in conjunction with two members of a web-based citizen science project accessing the astronomical amount of data NASA makes freely available to the public from it. NASA does this as it recognises the limitations of their own data-crunching computer programs which don’t know how to interpret variations that are signals that should be explored. Fortunately, this is something the human brain excels at. Engaging natural human sense-making abilities, by bringing multiple perspectives to bear on crucial challenges can create genuine breakthroughs and increase productivity at essentially zero cost.
Some organisations tap natural knowledge flows with huge commercial success - IBM is now repeating it’s hugely successful co-investment with Linux and Goldcorp discovered massive of previously hard to find gold reserves after making information freely available on the web. Real competitive advantage won’t come from investment in management fads - the learning curve is too costly. It will come from marrying multiple current capabilities together. This demands leaders reject the myth of the ‘one right answer’ and instead seek to stimulate the natural knowledge-sharing processes already occurring between humans; support them with technology they are able to use; and use their own human judgement to make sense of the insights to learn how to better act in the highly-complex world around them.
“Leaders who stay ‘above the details’ may do well in stable times [but]… riding a wave of change requires an intimate feel for its origins and dynamics.” Change is universal and accelerating, meaning leaders must cut through layers of delays and distortions to understand what’s really happening on the frontline and why. Failure to do so severely limits the potential to discover the critical insights needed to make effective decisions. Leaders must learn to disintermediate.
Disintermediation is a process of ‘cutting out the middle man’ - the layers of middle-managers ‘managing up’, only letting their boss hear what they think s/he wants to hear; the consultants who provide simplistic recipes to achieving ‘best practice’, invariably past practice devoid of context; or the knot of bureaucracy people complain about but feel helpless to change. A leader must learn to hear the signals despite the noise, recognise the wood amongst the trees, and understand the cause not just the correlation.
Big data is the current tool of choice for cutting through the jungle of obfuscation. But while Google can answer any question in nano-seconds it can’t tell you what to ask, nor can it provide you with any assurances that the answers you receive are not the work of madmen, fools or snake-oil salesmen. To make effective decisions beyond the routine re-ordering of inventory, or dynamic discounts decision-support technology must not seek to replace human intelligence, but augment it.
Technology must let decision-makers:
Effective decision-support systems should not make leaders dependant on third parties - quants to devise algorithms, programmers to code them, or experts to interpret them. Systems should augment the person accountable for the decision, heightening natural human pattern-recognition intelligence through visualisation tools; triggering the novelty-receptive brain through sharing rich, knowledge-based narratives (things this blog will elaborate on over the coming weeks).
This requires systems that work with the way people really are, not the way we wish they would be. For human beings have evolved over thousands of years as social creatures in networks (clans, tribes, families, and now organisations) and our success in getting this far suggests some secret formula for modern organisations to emulate - a best practice case if you like. And it’s this: technology works when in puts people in touch with people as beneficial variation then emerges.
Unless we’re willing to entirely hand over decision-making to machines we still need humans to make sense of the answers they suggest to us. This is just as well. For though humans, unlike computers, are very bad at calculating large sets of data; computers (unlike us) are very bad at providing sensible answers to complex questions. Making smart decisions in a complex world therefore requires employing tools to support the human, rather than employing humans to fit the tools we have built our organisations around.
Big data may have an important role to play in our highly-connected world. But big data excels in providing answers to a very narrow range of issues: inventory replenishment, automated underwriting, dynamic online offers and other ordered processes. Yet these are not the issues keeping executives awake at night. What disturbs their sleep is uncertainty: what will customers want tomorrow, what will rivals do next, and what will key staff members expect in future? Big data - of which we already have plenty (we currently produce more data every 10 minutes than we did in the 5,000 years prior to 2003) - is less a need than big knowledge: how to make sense of the data we already have to generate breakthroughs to the organisations most pressing problems.
The complexity expert, Dave Snowden, argues that ‘technology should therefore seek to support, not undermine human capabilities.’ We less want computers to make the big decisions for us and more need them to augment our natural decision-making abilities. Millennia of evolution suggests we humans are actually pretty good at surviving and thriving in a complex world. And our ability to manipulate tools to make this task easier is unrivalled in the natural world. What we mustn't do now is abdicate from our central role in decision-making in favour of a tool we've created. Humans must be firmly placed at the beginning and the end of any technology for it to have utility.
Therefore, before going down the path of more automation, more data (and more power to your CIO) consider for a moment why might big data not be the answer for you or your organisation:
Shape the Future
Don't just adapt to it