Julian Birkinshaw is Professor of Strategy and Entrepreneurship at London Business School. His particular field of interest is the challenges for large companies who want to work in more innovative and agile ways. I heard Julian speak at the Front End of Innovation conference in June 2017, and then read his latest book, Fast/Forward, and he was kind enough to spend some time on the phone with me.
Hi Julian, to get us started, could you tell me a bit more about the field you work in?
There’s a perennial need for large companies to try to get a balance between economies of scale and market size and power on the one hand, and responsiveness and adaptability and innovation on the other. Most big companies are very good at being big and they’re typically fairly bad at being responsive and flexible and adaptive.
There is always an appetite for advice on how to do these sorts of things better, so that’s the broad scope of my research over the last 19 years.
What I really liked about your book was the idea of emotional conviction being vitally important. I often feel that strategy is a post-rationalisation of a gut feel or a hunch, and was wondering what led you to be interested in this particular aspect of corporate entrepreneurship?
There’s two answers to that question. One is that over the last five years or so analytics, big data, artificial intelligence and machine learning have found their way to the top of the conversation in big companies because, as always, technology enables you to do things that you couldn’t do before. The further that conversation progressed, the more worried I became, because I felt that if you allow the computers to take over you could find your way into a set of unhelpful organisational pathologies.
One of these is the risk of analysis paralysis, the feeling that we’ve got to keep on gathering data, because there’s so much available, before we get to the right answer. Another is the risk that we make sterile decisions which completely lack any sense of context or meaning. If we’re using computers to tell us the right answer and our competitors do exactly the same thing, there’s the risk that we end up with a completely undifferentiated product, or we end up investing in exactly the same things as everyone else because that’s what the smart data tells us. So all of that led me to suggest that there’s still a role for humans in the process. As you say, human insight and intuition and creativity have always been a big part of strategy making and we mustn’t lose that.
Jonas [Ridderstråle — Julian’s co-author] and I came up with the twin notions of decisive action and emotional conviction as being the essential human qualities in any strategy making process. They’ve always been part of the story, and sometimes the stronger the analytical capability, the more important emotional conviction is in order to rise up above the crowd of companies around us.
The twin notions of decisive action and emotional conviction are the essential human qualities in any strategy making process.
The second reason was that Jonas and I had always had the view that the traditional hierarchical perspective on structure was rather limited. We know that structure and position and hierarchy matters but we also know that for organisations to be effective, you’ve got to think in terms of the way that people actually act and behave on a day-to-day basis and you have to think of the feelings, the emotion, the intuition of individuals. The four dimensions of formal structure, knowledge, action and emotion — each one of them matters if you’re trying to make sense of a big, complex organisation. There’s nothing new here, all four of these elements have very, very long research lineages, but we wanted to give some traction to the idea that companies put a lot of time into formal structures, and many of them nowadays think about knowledge management, but they don’t give as much attention to the action imperative or to the emotional imperative.
So that’s the other way of getting to decisive action plus emotional conviction, through the theory rather than through current rhetoric.
It seems like there’s a contradiction in that companies want their people to be emotionally engaged with the purpose of the company, but strategies should be rational and involve spreadsheets and numbers. But you explicitly say that a lot of strategy is about making those emotional leaps.
That’s right. I think any wise, established executive gets that, but the structures and the processes put in place all lean towards rationality and quantification as the only acceptable means by which you can convey an argument in the boardroom. If you’re in a board meeting and somebody says I see the data but it just doesn’t feel right, there’s a lot of pressure on them to justify their emotional dissonance around the rational data. We’ve come some of the way on this: there’s considerable behavioural economics literature building on Kahneman and Tversky’s original insights into the biases and flaws and limitations in our cognitive processing capabilities.
Their ideas have become so successful because they provide a way for us to rationalise non-rational action. They’ve taken us some way down the road to opening up the black box of non-rational insight, and now we need to go even further to give a voice or a licence to people to bring their intuition to bear.
I was working in the public sector when the Nudge book became incredibly popular, and the government set up their Nudge Unit. It was interesting because you could imagine that what really appealed to the government at the time about Nudge was it gave them a way of bringing rational predictability to influencing seemingly irrational citizen behaviour.
Absolutely — that’s what people need and the more regulated the industry the more everything has to be evidence based. Yet for private companies we need to find a way of allowing people to make these non-rational leaps based on intuition and emotional belief. Because, remember, emotional conviction and intuition is not people acting on a whim. Almost always it’s based on some deep insight borne of experience. Intuition is a point of view that cannot be fully rationalised, but is based on some sort of subconscious understanding of something.
Intuition is a point of view that cannot be fully rationalised, but is based on some sort of subconscious understanding of something.
Yes, you’re processing something in order to get there, aren’t you? You’re just not aware of that processing happening.
Thousands of books have been written about expert judgement and true experts often know more than they can tell. They’ve built a body of tacit knowledge and they will often come to a point of view and don’t even know themselves exactly why it is what it is. That’s been proven in many walks of life and we need to make sure that we don’t lose that in our decision-making process in big companies.
In many privately held or private equity companies or founder entrepreneur led companies, like Amazon and Facebook, you see executives making huge gut-based decisions. The worry is that the traditional plc environment or partnership type government structure doesn’t give people the same licence to act on their intuition. It’s perhaps not surprising that most of the exciting companies we talk about today have very, very narrow power bases: a small number of people who are making the big judgement.
It doesn’t mean they’re always right but they have more of a licence and, of course, this is classic survivorship bias, but the ones which have succeeded have been right more than they’re wrong.
You talk a bit about risks and failure in the book and, course, as soon as you choose a particular path right and wrong fall away, though they? Because you don’t know what would have happened if you done it differently. How do you think organisations can move away from this idea that there is a right strategy or a right answer?
It’s now axiomatic that in order to innovate and change we have to be prepared to tolerate a reasonable level of failure along the way, and it has also now become the norm to argue for experimentation as a means of trying out ideas in a low risk way. I won’t say anything against that as a principle, but I’ve got a couple of observations on it.
Observation number one is that an experimental approach to building a new strategy is based on the notion that we need to mitigate the downside risks, so we put a small amount of money in and then when we get to first base, then we can put more money in and we sequence investment in that way. That’s a good idea and I think most companies now are wise to that. The part of the story which sometimes gets missed is that failure by definition involves benefit as well as costs. There’s always some value in failure. In the book we came up with the notion of thinking in terms of return on failure, just like we think in terms of return on investment. There’s a whole host of things we get out of a failure, one of which is obviously learning directly from it. The second is the signal we are sending to the organisation that trying things out, even when they don’t work, is a good thing.
So my view on experimentation is that, yes, we need to experiment but we’ve also got to build in a structured process to ensure that when something doesn’t work, rather than sweeping it under the carpet or pretending it didn’t happen, we explicitly review quickly and thoughtfully what we learn from it so that we can then move on and try the next thing. I think they call it the triple layer: fast, frequent and forward looking.
The notion that we would experiment in organisational changes or learning development initiatives, for example, is still anathema.
The other point is that there’s something about the psyche in large companies in particular that makes them struggle to do experiments. The notion that we would experiment in organisational changes or learning development initiatives, for example, is still anathema. If a company is thinking of a couple of different models for a big executive training programme, then what they should really do is try a pilot of one option, a pilot of the other and then do a compare and contrast. In my experience, they never do that. They decide on one that they prefer then they run that one on a simple pilot basis, which says ‘we’re doing it but we want to have a little get out clause just in case it doesn’t quite work out’. That means they will never do the explicit compare and contrast of the two options.
I’ve even caught myself making this mistake. I’m head of degree programmes at London Business School, and I think about whether we should try running one elective or another. What we should really try doing is one of each and then review after each elective course has run. Yet too often we shy away from the pain or the indecisiveness that seems to come with not making a choice. There are still massive behavioural or structural impediments in organisations that get in the way of doing experimentation.
I suppose in some way doing that kind of experiment would involve the organisation admitting that they don’t know what the best option is. And that’s quite a difficult thing, I think, for people to admit.
Yes. The only way that that becomes legitimate is from the top. If your boss is looking for clarity and decisiveness, then it’s difficult for you to say, “I don’t know, that’s why we’re doing an experiment.” Decisive, in my language, doesn’t necessarily mean going all out, it means trying something out, doing something. The best way to resolve uncertainty is by taking action — a small experiment might be the best way of moving forward.
Another aspect of the book that I really appreciated was the fact that you don’t talk about this being the end of bureaucracy. There’s a lot of books out there that say quite dogmatically that all organisations must espouse a particular way of working. Whereas you seem to be suggesting that different approaches or even combinations of approaches are right for different types of organisation. Do you think there’s a danger that companies end up somewhere in the middle, not doing any of them terribly well?
Some people do say they want to be a mixture of all three models in the book (adhocracy, bureaucracy, meritocracy). But they’re missing the point because, by definition, each model involves some combination of formal structure and knowledge and action. The question you’ve got to ask is what is the right model for this particular unit at this time? You have to make a choice about which one of those three models will be the dominant one.
Now, where it gets interesting, of course, is that you could have a manufacturing facility being run as a well functioning bureaucracy, because they are simply producing high volume products with low cost. But within that single manufacturing facility you might also have an operational effectiveness team experimenting with new and better ways of working. The person managing that factory has to be able to work across the two different models. Otherwise what happens is that the unit, which is meant to be moving much more quickly, ends up getting sucked in to the decision making processes and the routines and the risk aversion of the mainstream bureaucratic unit.
What we’re trying to do is give people a language and a set of practices that help them name the different activities they’re operating so they can make sure that their style of working, rules and practices are fit for purpose. For example, in my current role, one day I’m sitting in amongst my academic colleagues, which is a pure meritocracy. No one cares about titles, it’s all about discussion and debate and getting to the right answer. But I also have an administrative role where I’m operating in a much more traditional bureaucratic way with lines of authority and clear KPIs and so forth. So my job is to switch my style back and forth between these two ways of working.
That makes sense. I was thinking that there’s a danger that we forget that not all jobs are knowledge jobs, not all jobs involve those kinds of skills. I worry about the inequality gap widening between the people that are doing jobs where they’re managed in a bureaucratic way and the people that are more free to think and create. Do you think it’s possible to involve everybody in these kinds of ways of working?
Where I think I differ from some of my fellow management thinkers is that I still see a very strong place for traditional, well functioning bureaucracy. I don’t think that getting rid of bureaucratic methods across the board is actually possible or appropriate. But we do run the risk that by putting people in a bureaucracy box, their job becomes very narrow and uninspiring and we end up with all the problems around disengagement that blight big companies around the world.
I still see a very strong place for traditional, well functioning bureaucracy.
We’ve got a bit of a challenge: a lot of big companies have no choice but to get a whole lot of very tedious work done. Yet, at the same time, they’re operating in competitive markets and they’ve got to try to find a way of getting the best out of their people and keeping their top talent.
A common solution to this problem is the notion of a shadow, two speed or dual organisation. John Kotter wrote a book on this recently called Accelerate and I’ve given it my own particular spin. Essentially, imagine the traditional hierarchical model where people have formal roles and responsibilities and then imagine, behind that, a more organic or flexible shadow model. People in traditional hierarchically managed jobs spend 80% of the time working in their line role and they have to do things by the book, through their job description and according to their line manager’s wishes. But they get one day a week, 20% of their time, not just to do what they like but to join a project in the more organic shadow organisation.
So for one day a week I’m on a special project working at much more rapid speed, and it might be working on an area which is an area of personal development for me. Those temporary projects have deliverables, but they are managed in a more ad hoc way. It’s a way of getting some things done more quickly whilst giving individuals, who are doing mundane jobs, an opportunity for self development.
I’ve seen versions of this in some Indian software companies, like Infosys and TCS, I’ve seen a version of it in Bayer, the German pharmaceutical company, and I’ve seen quite a lot of companies expand with informal projects in a smaller way. The challenge in making that work, of course, is that it’s very hard to have enough of this sort of organic project-based work for it to work on a systematic basis. But it’s one way of resolving the problem because if we just take high potential people and throw them off into a separate skunkworks team to do all the cool stuff, it creates a problem with the people who aren’t selected. They feel their job is to earn enough money to pay for other people to do the interesting stuff, and not surprisingly people get a bit fed up with that.
I’ve seen a lot of people writing over the last few months about the importance of having side projects, so I guess this would be side projects within an organisation. That brings me on to the other thing I wanted to ask you about. I’m a freelance consultant so I go into different companies and work two days a week here and three days a week there. You’ve talked a bit about how opportunity focused organisations make the standard salaried employment model less attractive. Do you think the future is going to be much more about people having various different roles that and perhaps that 20% time coming from their own personal side projects?
We know that these virtual companies are on the rise. We know that in the professional services sector there’s a lot of these network companies emerging in law and consultancy and on the other end of the salary scale you’ve got the gig economy. The broader trend is towards more flexible employment contracts in which individuals have, for example, an 80% contract rather than 100% contract. That’s what my personal arrangement is with London Business School, and it’s always been like that, I’m allowed to spend one day a week working for third parties. I think a reasonable number of professional organisations, particularly for more senior roles, have created such opportunities and I think they’re going to be more and more frequent, absolutely. But I do not believe that we’re going to move away from traditional integrated companies.
There was an article by Tom Malone called the E-Lance Economy which suggested that digitisation and computerisation were making traditional integrated companies irrelevant. I see absolutely no evidence that traditional integrated companies are going away, there are still absolutely enormous benefits in having the coordination, sharing of knowledge and shared identity of being part of a company. But the model whereby people get a little bit of freedom on the margin to dabble in their own self personal development, a bit of flexibility to try new stuff, that actually works quite well.
You’re quite clear in your book that you don’t think that the information age is done yet, that we’re currently in a maturing of the information age. Do you think we’re heading towards the Information Age, as Rita King calls it?
I go back and forth on this. In my more traditional mindset I think we’re still in the thick of this information age, we’re absolutely still reaping the productivity benefits that come from automation and computerisation. But it’s impossible that it will go on forever, so there will be something beyond it, for sure, but it may be a long way down the track. We need to move from the information age, where information and knowledge is king, to the agile age where it’s all about getting the right blend of knowledge, information and human agency. Ray Kurzweil claims that computers are going to do everything for us, so the singularity is when human and machine consciousness somehow blur together. I don’t go with that, I think he’s deliberately being provocative in his own way.
I don’t get too hung up on pure futurism. I’m very comfortable that for the next decade we will be harnessing the powers of technology like driverless cars whilst also firmly reconnecting with the human capabilities which computers will never be able to match. Getting that balance is what smart companies need to do. Most company executives can’t even think beyond the next couple of months, let alone the next couple of years.
What do you think is the most important thing that organisations could learn from reading your book?
Two things. Point one is decisive action plus emotional conviction, that is the heart of any successful organisation in a fast paced business world. Point two is about the adhocracy, but more specifically it is that an action orientation should be at the heart of what we do as individuals. So if we can’t quite get our heads around decisive action and emotion conviction as a corporate imperative, we all, as individuals, can figure out a different way of acting. That means taking more small steps, more frequently, and learning from those things and for those of us who are in positions of management and responsibility, enabling that to happen around us. The most successful companies are the ones who have built a group of mid to senior high quality managers who have enabled those around them to be corporate entrepreneurs.
That means taking more small steps, more frequently, and for those of us who are in positions of management and responsibility, enabling that to happen around us.
The idea that we could somehow make things happen through formal structures or through one or two visionary people at the top has never made much sense. The people on the frontlines are the ones who understand what’s going on, the ones who have the capacity to make things happen and we need to create a culture in which their entrepreneurial endeavours are encouraged.