Ambidexterity is imperative as leaders grapple with the implications of AI
How far will artificial intelligence impact your business this year? What about over the next three years? Today, leaders face few challenges as urgent as adapting to the emergence of AI and adopting it in their organizations. From delivering a step change in process efficiency and productivity, to turbo-charging innovation, to ripping up established business models and making possible entirely new ones: the potential for AI seems near-endless.
But how can organizations make sense of an inherently uncertain and fast-changing picture? As leaders attempt to chart a path forward, one of the most insightful thinkers they might consult is Alex Osterwalder, founder and chief executive of Strategyzer.
Osterwalder is ranked at #8 in the Thinkers50 list of management thinkers – jointly with Yves Pigneur, with whom he developed the Business Model Canvas and Value Proposition Canvas tools – and famed for his work on innovation and strategy. We connected with Osterwalder to discuss the transformative challenges facing leaders today – and his own journey in leadership.
Human constructs
Osterwalder’s passion for business began to emerge in childhood, although his is not the familiar story of entrepreneurs who started by selling to their schoolfriends. “My dad was a senior leader in a large company, and I found that fascinating. I was interested in business – in what business is.”
That curiosity was to inform Osterwalder’s educational choices – although his path took some unexpected turns. “I failed at business school, then did political science. But, competitive as I am, I went back to business school and did a PhD,” he laughs.
The decision to study political science proved pivotal, reigniting his interest in how corporations function. “Political science was the best thing that could have happened to me,” he explains. “At the University of Lausanne, several professors had sociological views of politics. They drilled into us that institutions and society are the way they are because people built them. That helped me later to question everything, including the so-called ‘laws’ of business.
“Steve Jobs talked about this,” he continues. “Everything around you is man-made – and changeable. It’s up to you to create.” This is a critical truth for any leader seeking to transform their organization. “Political science taught me the world doesn’t have to be the way it is. Business models don’t have to be the way they are.”
Osterwalder’s father had also sparked an interest in computing when, in the 1980s, he brought home an early DOS-based PC. He left his son to experiment, igniting an interest in programming and technology that, years later, influenced Osterwalder’s decision to embark on his PhD in management information systems – supervised by Yves Pigneur. It was the start of a productive and enduring relationship – but it almost didn’t happen. “I did a pre-interview with McKinsey but they didn’t want me,” recounts Osterwalder. “I got lucky with that failure!”
Osterwalder’s studies led him to develop the Business Model Canvas, a tool for exploring the nine components of how any given business creates, delivers and captures value. It was presented in his 2010 book with Pigneur, Business Model Generation – with the Value Proposition Canvas outlined in 2014’s Value Proposition Design.
Those tools have become boardroom staples, helping executives build clarity about the central fundamentals of their business – a task that is more relevant than ever as organizations grapple with AI.
Technology traps
In Duke CE’s conversations with C-suite leaders, the difficulty of distinguishing signal from noise around AI is a recurring theme. Osterwalder has observed the same in his discussions with leading global firms. “The current uncertainty is forcing leadership to think about innovation,” he reflects. “Before, they could get away with not doing it; now, that’s harder, both because of uncertainty and because of the technological opportunities.” This points to a fundamental truth. “It’s not a technology problem. It’s a business challenge. How do you do innovation? How do you de-risk?”
While running Strategyzer accounts for much of his time, Osterwalder has recently been spending more time in the field with clients – and he sees many companies approaching the challenge from the wrong direction. “I work with some clients who are really into technology, and they forget that it’s actually about the value proposition for clients. It’s so simple – but when you’re a big corporation, you can lose sight of that.”
Refocusing on the fundamentals of the business model – on what a business does for its customers – is critical. “Going back to basics is so important right now, and so powerful,” he says. He explains how organizations can lose their way. “Every startup or innovation starts with an idea – a technology, patent, opportunity, asset. You search for a solution – product, value proposition, business model. Once you find it, you scale; it becomes a management challenge.”
The problem is one of over-tilting toward management of an established model. Here lies the fundamental shift required of leaders today: rebalancing the two modes of business, “explore” and “exploit.” Entrepreneurs and creators explore; large corporations are optimized to exploit. “I’m working with senior leadership teams at of some of the biggest companies in the world right now, and this is the big theme: balancing explore and exploit,” he confirms. “These large companies were startups once, but they forgot how to create. They became execution machines. They have to rediscover how to create.”
This is not a new idea, he notes. “Academics have been writing about the ambidextrous organization for ages, but they haven’t really popped up that often, with a few exceptions like Amazon.” But AI means the idea is impossible to ignore. “What gets me really excited is that this has leadership consequences,” he continues. On the exploit side of the ledger are the traditional management questions: is work done on time, on budget, and with robust project management? With explore, the questions are very different: What did we learn? Should we kill the project?
“Here’s the thing,” he adds. “I show senior leadership teams how these two worlds, explore and exploit, are different. What I’m constantly fascinated by is that these are very, very intelligent people running multi-billion dollar businesses – yet such a simple concept can blow their minds.”
How is this explained? Most C-suite leaders lack entrepreneurial instincts or experience, says Osterwalder: they spend their careers running businesses, before taking the top seats in the biggest companies. One exception is Bracken Darrell, the former CEO of Logitech (now of VF Corporation). “He’s an entrepreneurial CEO who understands that those two worlds are different. He can toggle between the two mindsets,” says Osterwalder. That is a clear competitive advantage. “I think we’re going to see this more and more often: the CEO who’s good at running a business, has a vision of the future and is entrepreneurial will out-compete the managerial CEO.” That’s not to say that core management capabilities are obsolete. “You can never compromise on running the business. You can’t innovate yourself out of a hole. But you need both: you need that ambidexterity.”
The venture approach
One of the central tasks involved in building an ambidextrous organization is changing how resources are allocated. Learn from venture capitalists, says Osterwalder: they create a funnel for their investments, making lots of small bets while knowing only a few will pay out. This is critical for AI adoption, where big bets on unproven use cases could be ruinous. “Some executives say, ‘I need to put $10 million, or $100 million, in one AI project.’ That’s the best way to burn money! What you want to do is take $10 million and spread it over 100 projects – then after three months, kill 50% of the projects. That’s what the venture world does really well.”
Big bets can work in a crisis, he points out. When digital technology destroyed the traditional photographic film business, Fujifilm pivoted to cosmetics, applying their IP to skincare. “They had to bet the farm, because they had no time,” notes Osterwalder. “But in general, that’s the best way to maximize your risk of failure. The best companies experiment all the time, systematically building up a portfolio approach.”
The fact that many firms still make big bets drives perhaps the biggest myth about innovation – that it’s too costly. “People say it’s expensive and risky. Well, if you make one bet, it’s like roulette – but if you work like a venture capitalist, you spread your bets and you let the winning ideas emerge. Leaders think, ‘I need to pick the best idea,’ but nobody can do that – not even venture capitalists. So you manage risk by spreading your bets.”
The strategy is well-proven yet it clashes with how corporations operate. “We’ve optimized for running the business and killed all the exploration,” points out Osterwalder. “I call it the ‘corporate antibodies.’ Anything that looks different is killed too early, so companies can’t explore.” It’s akin to an allergic reaction: the body corporate is destroying the very thing it needs most for long-term vitality.
We tend to lionize innovative leaders for their creativity, but other qualities matter more. “Creativity is overrated. People who believe in ideas will always fail, because innovation is not about the idea. The hard part is iterating the idea until it works,” says Osterwalder. That makes decision-making key, starting with understanding the type of decision at hand. Jeff Bezos famously talked of one-way and two-day doors: irreversible and reversible choices. “There are decisions where analysis matters – you’re making a big bet and you have to think hard. That’s the execution type of decision,” says Osterwalder. With two-way doors – those small bets – it’s different. “Don’t think, just go and do. Experiment, experiment, experiment – knowing that half of the things you try won’t work, and that’s OK.”
Amazon has long stood out for its innovative capability, which has helped it create entirely new business models such as Amazon Web Services. Compare Google, which has successfully iterated its advertising business model, without building entirely new business models to the same extent.
Lessons can be learned from Chinese conglomerate Ping An, says Osterwalder. It grasped the need for different approaches to explore and exploit, and appointed Jessica Tan to head the explore side – with the title of co-CEO. That was significant. “The biggest challenge for innovation is usually the lack of power,” says Osterwalder. “What gives you prestige in most companies? The number of people you manage and the budget you have.”
The conundrum is that innovation, by its nature, starts small – and small initiatives don’t win attention in a corporate environment. “We have to make these tiny explorations sexy again. That’s what serial innovators are really good at – they give prestige to exploration.” They’re also consistently focused on customers, he adds. “Technology is a means to an end: it’s about creating value for customers, with value propositions embedded in scalable business models.”
The great accelerator
We turn back to the specific challenges around AI. The innovation process will be accelerated. “You can prototype a value proposition, use agents to do customer interviews – with real people or synthetic users – and stress-test a concept within a day,” he says. That changes the equation for success. “My hypothesis is that exploration is going to trump exploitation in an AI-first world,” argues Osterwalder. “Innovation culture is going to become more of a competitive advantage than the actual products and technologies. The ability to constantly reinvent yourself is going to be the key competitive advantage. We’re already seeing this.”
That spans every type of innovation – efficiency, sustaining and transformative, to adopt Clayton Christensen’s categories. Leaders need to consider where their efforts are focused. “You can use AI to improve your business model – but if your business model is dying, you’re just going to die more efficiently,” jokes Osterwalder. So where do leaders start? “The first thing I suggest is increasing your ability to experiment, because with AI, like with any innovation, we don’t know what’s really going to work.”
All options should be on the table. “You can use investments, intrapreneurial teams, M&A, open innovation, doing it inside or outside – use every part of the innovation toolbox,” he advises. “But the mindset – the philosophy – is the first thing to get right.” The playbook for building an innovation culture can also be codified, he suggests. “There’s the formal stuff: What do I pay people for? Do I pay them to execute, or do I give them time and money, and reward them in the 360, for exploration? Then the less formal part is how leaders behave in innovation meetings. Do they insist on a business plan, or do they ask, ‘What’s the evidence?’ It’s not rocket science, but it’s hard to deploy this at scale.”
Business schools have a role to play, too. They have long focused on perfecting Management 101, we suggest: prioritizing exploit over explore. Osterwalder is emphatic. “They are not educating leaders to create the ambidextrous organization: it’s a side topic when it should be the core.”
As AI transforms our world, starting with our businesses, the challenge will be existential. “Ambidextrous companies are the only ones that are going to survive in the long term,” says Osterwalder. Leaders ignore his analysis at their peril.
Vishal Patel is president of global markets at Duke Corporate Education. Patrick Woodman is editor of Dialogue
