Disruption at OCC
Director John Boyle recently gave a talk on technology-led disruption and why a company must act as its own disruptor, continuously challenging the way it does things to remain innovative, efficient and an interesting place to work.
Disruption is not a new concept in business. Disruption’s latest incarnation is the sharing economy, which is disrupting traditional business models, including those of the information technology age itself. Some argue that the zero-hour contract models of Uber and Airbnb are taking us back to the days of The Grapes of Wrath, where the foreman selects the workers he needs each day, and sends the others away with no income. We are starting to see social unrest around such business models, including strikes, street protests and class-action lawsuits.
The sharing economy’s behaviour is largely unstructured and unpredictable. It primarily follows what is termed emergence systems theory, where larger entities, patterns and ‘intelligence’ arise through interactions among smaller or simpler entities that do not themselves exhibit such properties. It is a model that can be used to describe how our brains and the internet work and how sharing economy models prosper. The importance of emergence theory is that it lies at the heart of the opportunity, and some might say threat, of artificial intelligence, which is, in turn, itself a disruptor. The reason it can be seen as a threat is that emergence theory eschews the idea of a master plan or controller guiding the emerging ‘intelligence’.
Emergence theory is important because it shows the potential for complex behaviour to emerge where feedback mechanisms exist. The most famous models are ant colonies. If you put an ant colony in a large and controlled space with a food source, the ants create the same complex social structure each time, with an ant hill, a rubbish tip and a cemetery – and these three points will automatically lie at the maximum possible distance from each other.
Emergence theory is a model for today’s technology-led, disruptive forces. However, because of emergence theory’s lack of human-led design, it may be a dangerous theory to have at the heart of our technical evolution.
I was first alerted to the importance of disruption by an excellent series of talks by Clayton Christensen. Over 3 hours, he essentially describes how the US steel industry died because US firms allowed Chinese companies to disrupt it. The Chinese did this step by step, first in low-grade steel, then medium-grade. The error the Americans made was to keep saying, “Oh, it doesn’t matter if that small and low profit area is disrupted.” Before they knew it, three-quarters of their industry had been taken over.
Disruption at Oxford Computer Consultants
While emergence theory is useful to describe how patterns evolve in technology, at OCC we cannot trust to emergence theory because it does not allow us to control and plan our disruptive path and because emergence theory works on scales and timeframes that are not appropriate for our size and industries. We need to control our experiments and put our own rapid evolutionary models in place.
At OCC we must be our own disruptors but we must disrupt ourselves intelligently. We need to be able to explain and justify the decisions we take, be those new architectures, pricing models or project management techniques. This means we need rapid feedback loops to understand, validate and improve on the decisions we make, and our own rapid evolutionary process capable of rejecting poor mutations and evolving those that bring measurable advantages.
If we fail, our products will be surpassed by new solutions. Indeed, the reason we replaced the long-term incumbents in social care software is because we disrupted the way social care finance is integrated and implemented with case management and we were able to disrupt the pricing models in this sector. I believe the reason our Innovation Delivery team has grown its scientific and engineering development, despite us being essentially an IT company, is because we have never lost sight of the need to understand the science as well as the IT technologies; that is a novel behaviour.
Not all our disruptions have worked. The lesson I’ve learnt is that we need to test and validate our disruptions as early as possible, through technical design reviews, feedback from customers, evaluation of costs and listening closely to what our own staff are telling us. A good disruptor has a good feedback mechanism.
Disruption – a word of caution
I worry that my discussion might provide intellectual cover for us losing sight of the evolutionary behaviour that is the reality of most progress. If you interpret this talk just as saying you should ‘disrupt’, this can be used to justify inventing an octagonal wheel, because how is using a round wheel disruptive? We must disrupt intelligently!
There is an important reason why I believe disruption is better than just improvement. Part of this is about seeking to identify competitors that are vulnerable to being attacked with the tools at our disposal. In the case of our new products, we identified that the incumbents were weak at implementing integration between case management and finance. By implementing, we mean delivering into a live and working environment, not writing the code. We also identified they were weak at using web technologies to change the relationship between supplier and provider and they were weak at realising that new UI experiences were emerging.
So we must recognize that disruption is often not technical but rather a novel use of existing technologies.
The social impact of disruption
However, we must think about our social obligations in the way we deliver software. This is not just about the dangers of the decision-making potential of technology itself but also the social problems. Faced with the ever-increasing presence of digital technologies, it is easy to feel disconnected from our surrounding environment.
In his Christmas interview with Prince Harry, Barack Obama talked about how “we have to find ways in which we can recreate a common space on the internet. One of the dangers of the internet is that people can have entirely different realities. They can be cocooned in information that reinforces their current biases.”
Let me take this back to a real example in OCC work. Local authorities and health trusts are concerned that, with people increasingly using websites to obtain information and advice, those websites guide them safely. We hope people understand the purpose of a site they access but often they don’t – for example, thinking that a local authority social care site can help them with medical emergencies. This raises challenges for the way in which we design software. Design needs to ensure people understand the difference between the virtual and physical world. People need to be helped to understand how each tool they use can, and cannot, help them.
Allow me to go back to a paper I wrote over 27 years ago but which is as relevant today, certainly to much of the work OCC does. It relates to the simple calculator. I argued (and still believe) that a calculator should have a ‘Test’ rather than ‘On’ button. Imagine a simple calculator with just the basic buttons: + - / *. The test should be a random question such as 7 * 4. Only if the user can get the right result using the calculator, should the calculator allow the user to proceed. Why does this matter? After all, calculators always get calculations right. But what if I type 1200 * 7 when I meant to type 120 * 7? If I have no appreciation of numbers I might not notice the error and I may make a disastrous decision.
I would like to apply the same principle to our own products. For example, our finance system could ask the maximum value of payments that should be made to a provider before allowing use of the software. Perhaps our thermal rating software, developed for National Grid, could ask the melting temperature of copper, and only allow users to proceed if they get this roughly right.
In 1983, Stanislav Petrov, an officer of the Soviet Air Defence Forces, prevented a retaliatory nuclear attack, based on erroneous data, on the United States and its NATO allies that would have probably resulted in immediate escalation of the cold-war stalemate to a full-scale nuclear war. For those of you who are interested, there’s a documentary film called The Man Who Saved the World. It’s a topic that interests me because of my father’s role in the Cuban missile crisis. He was the navigator on a Vulcan bomber, which waited, engines running, with a set of codes and the equivalent of 8 Hiroshima bombs on board, for the command from Kennedy.
We all enjoy the benefits of technology but I’m keen to ensure we never forget our social obligations and that human connectedness should be enhanced by what we do. There are dangers to society if the virtual and physical worlds become too separate and humans no longer understand what they are doing.
Disruption for employees
Some of you may be thinking how does this talk of OCC behaving in a disruptive manner affect me? I believe everyone needs to be their own mini-disruptors, learning about new technologies, frameworks, design concepts and standards. There are new people coming along the whole time, be they off-shore, graduates trained in new ways of working or people who want to innovate. These are perhaps uncomfortable truths and we try to help our employees with this by offering training every year, tech talks, lunchtime events and the opportunity, for those who work on products, to become deep and valuable domain experts.
Change should be enjoyable. It should stretch you, give you the opportunity to improve the way you work and make you feel part of a company that will never rest on its laurels.