Thinking what nobody has thought: how to get corporate innovation right

George Zarkadakis
12 min readFeb 12, 2021


In my career I have witnessed many companies and industries that did not innovate enough and are now extinct. Let me mention just two of them. Silicon Graphics was an innovation legend in the 1990s with a bold vision about the future, such as VR systems that would revolutionize education, healthcare and work. I am proud to have worked for “SGI”, and sad that the original company is not around anymore, not because we were not smart or talented enough but because we had placed all our bets on workstations and supercomputers and did not see the interconnected PC revolution coming. Second victim: the print media, where I worked in the 2000s. The corporation cemetery is strewn with the bones of big magazine and newspaper publishers who saw the web as just another channel and did not realize its disruptive power for their business models, until it was too late. What worries me is that I keep seeing those same errors of judgement repeated by many of the corporations that I work with today, despite the tons of books and papers on innovation, and the armies of innovation experts and gurus. It seems to me that innovation is easy for a small and dedicated startup, or a research team in a University lab, but remains hard for large corporations. In this article, I want to explore why, and what to do about it, so that successful companies that create value for the benefit of their customers, their shareholders, their communities and their employees, can survive — and thrive — in an age of exponential technological disruption.

Getting it wrong: barriers to innovation

In his seminal book “The Innovator’s Dilemma” Clayton Christensen identified how one of the principal causes of creative destruction in a free market system, namely technology, can blindsight corporate leaders: by challenging established notions of acceptable business risk. This is particularly applicable to highly successful businesses and industries at the top of their game. The examples I gave before, of Silicon Graphics and the print media industry, are typical. The higher the feelgood factor from reading your cashflow statement the higher the probability of your catastrophic extinction by a newcomer who will eat your lunch wielding some brand-new shiny tech, one that you dismissed as irrelevant — or too risky to invest in — in your last board meeting. Nevertheless, since the publication of Christensen’s book, and the rise of the big tech behemoths and the startup billionaires, innovation has become part of every CEO’s briefing vocabulary. In fact, it is rare to find a company nowadays that does not present itself as one that boldly pushes the boundaries of technology further and beyond. It would seem that we are innovators now. Really? Let me summarize below what I think are the key symptoms of not getting corporate innovation right, and why the walk does not always follow the talk:

Lack of clear vision at the top: Top leaders have a vague idea about innovation, or think of innovation as something that happens naturally, like weeds in a field, or that it is marginal to their core business. Typically, those leaders’ key metrics of success are around sales and market share, rather than number of experiments conducted and successful ideas generated. They seem to ignore that we live in an economy of intangible assets where real value has shifted to knowledge, ideas, software, patents, algorithms and data. If in doubt, just look at Tesla’s valuation and compare it to the other car manufacturers.

Inadequate investment and commitment: as a consequence of the previous factor there is inadequate support and investment in corporate innovation. I have seen C-suite executives actively avoiding getting involved with innovation as they consider it too risky for their careers. Innovation is a long game. You need to take several losses over a period of time before you score the winning goal. But when your pay package and bonus are based solely on successfully hitting annual revenue and efficiency targets this not a game for you.

Ideas rarely get productized: So we now have an innovation platform on our intranet that collects proposals, and a fresh-faced innovation team of ex-hackers and startupers to lead the charge, have captured a bunch of ideas in the company’s hackathon, and our workforce are really excited about getting their ideas out there. But as those ideas are taken through a prototyping phase, they quickly wither and die without ever been set up for success. And if they ever manage to see the light of an MVP, they rarely become full products. Why? Because no one at the top really wants innovation (if it is going to cost too much); they just want to talk about it to market analysts.

People are not recognized: as for those employees who got really excited, woe to their bonuses and chances of promotion, as they will realize come annual review. Without reward-related innovation metrics, the effort of participating in jams and design sessions is discounted as time spent on a discretionary basis, like having-fun-with-colleagues on company time sort of thing. And so, the excitement among the workforce subsides, and fewer ideas land on next year’s company hackathon.

It is very doubtful that companies that recognize themselves in the list above will survive in the post-pandemic era. The degree and pace of digitalization that has taken place because of lockdowns and remote work, means that technological disruption in the 2020s will be unprecedented. All of us, at least those who live in the developed world, have changed how we learn, work, consume, save, invest, transact, entertain, travel (or not), relate, care or love. The risk of not understanding the magnitude and direction of change, and of not responding rapidly and innovatively to new consumer demands and societal trends, is nothing less than obsolescence. That is why getting innovation right is more important now than ever before.

Levels of innovation

My favorite definition for innovation is the carrying out of new combinations of existing things. This makes it different from invention that aims to find something that did not exist before. But once something is invented, it becomes a “thing” that can be combined with other things in an innovation process. Invention is the mathematical solution to the Byzantine Generals Problem. Innovation is using the solution to disrupt the global banking system. The relation between innovation and invention is key to envisioning effective strategies for corporate innovation, and I will return to discussing some more this very important link. But let me start by distinguishing three types of innovation. This distinction is based on the “3-horizon model” of innovation by McKinsey, which introduced the idea of the “ambidextrous” organization, one able of executing on an established business model while simultaneously creating new capabilities. I have somewhat adapted the McKinsey Model to the exciting new possibilities of the 21st century. So here are my “three levels of innovation”, which I like to visualize as successive waves spreading across a canvas of possibilities (see figure 1).

Level 1: Incremental. This is the basic level innovation that a company must do in order to remain relevant in its market. It may involve improvements in its products and services, or technological upgrades that deliver more efficiency and productivity. For example, incremental innovation is to improve decision-making in financial investments by combining standard financial data with alternative data (e.g. satellite data, or text mining company reports). Often, incremental innovation needs to align with the digital transformation agenda. Depending on the organization, incremental innovation can be a centralized shared service or dispersed across the business. My favorite organization structure is a hub and spokes model whereby a Centre of Excellence (CoE) for Innovation is the hub, while the spokes are business segments and citizen innovators on the organization’s edge. Following this structure delivers significant efficiencies, particularly as organizations aim to monetize their data and transform into data-driven businesses. A CoE for Innovation, for instance an innovation lab for data and AI, can provide the necessary vision and technological leadership, setting up the organization for success by enabling maintainability, serviceability and scaling of successful products and ideas. For example, having a cloud-based microservices infrastructure together with an easy to access and mine data infrastructure can accelerate innovation cycles. It is vital that innovation cycles from ideation to MVP testing should be as short as possible, a couple months or a few weeks at most.

Level 2. Transformative. This level of innovation aims to enable significant changes in the existing business model and how the company serves its customers. These goals often entail specialist capabilities, technologies, and skills and knowledge that may not exist within the company. Transformative innovation targets strategic, but still immature, technologies that could make a huge difference for the company. For example, if you are a logistics company a key transformative technology would be image recognition. For an energy company, batteries are transformative. For a financial services business it would be privacy-preserving technologies such as homomorphic encryption, advanced AI techniques such as reinforcement learning, algorithmic explainability, complex data exchanges, or blockchain solutions for creating new classes of digital assets. I tend to think of transformative innovation as core R&D with a high degree of “invention” involved. But, unless you are Google and you can sink several hundred million dollars in an R&D outfit such as DeepMind, the only way to execute on transformative innovation is through a partnership ecosystem. Such ecosystem would include links to Universities, specialist research labs, industry consortia, and startups that innovate on the outer limits of technological possibilities. Often, transformative innovation is encouraged by government subsidies. It is usually the space where private-public partnerships are best utilized. The timescale for transformative innovation can be long. It is important, however, that the company has an effective and efficient incremental innovation capability (“Level 1”) in place that can absorb the results of transformative innovation and integrate them quickly into new products and services.

Level 3: Disruptive. This level of innovation refers to combinations of technologies that disrupt traditional business models and replace them with new ones. The example I gave at the beginning with the digitalization of the print media industry is a typical one. The automotive industry is currently undergoing a similar disruption: it is becoming a fintech business. I can easily imagine a day (very soon) where I will be taking my car anywhere, and I mean take it in my smartphone and carry it in my pocket: my “car” will be personalized software code and personal data secured on a blockchain (including my driving habits, routes, driver risk score, but also where I go shopping, sports, etc.) that I will carry along and connect to any “car-machine” hardware. Guess where most of the value will be: in the hardware, or in the software and the data? Disruptive innovation used to be thought of as something that took many years to develop. The McKinsey Model sets the timeframe of “horizon 3” innovation between 5–12 years. That is not true anymore. Remember that innovation is the new combination of existing “things”. The example I gave on how the automotive industry could be disrupted does not need new discoveries to take place. It can happen tomorrow, if a team of entrepreneurs backed by capital puts their minds to it. So how can incumbent companies defend against such devastating threats? There are four main strategies for getting disruptive innovation right.

· Self-disrupt: IBM creating IBM Watson as a new AI company.

· M&A: Salesforce acquiring Slack

· Copy them: what car manufacturers scramble to do to compete with Tesla.

· DIY: plug into an ecosystem of disruptive innovation that includes startups and Venture Capital, and be bold enough to take quick decisions, take risks, and place bets.

Like with Transformative Innovation, having a strong Incremental Innovation basis is important to guide effective strategy, execution, and adoption of Disruptive Innovation. That is why in figure 1 all three levels of innovation are connecting at the bottom: they are interlinked. Strong Level 1 innovation capabilities must be in place in order for innovation across all three levels to be strategic and not tactical. I have seen too many companies that are participating in startup accelerators and incubators, seeking the hedge against disruptive innovation, but lacking an effective Level 1 incremental innovation capability. Those companies are not able to harness disruptive innovations, even if they partly own them, and are at high risk of becoming extinct.

Getting it right: an operating model for corporate innovation

Getting corporate innovation right requires changing the company mindset, which is a huge undertaking. Let me therefore give some practical tips for ambitious business leaders that are sincerely aspiring to transform their companies into innovation powerhouses.

C-suite sponsorship: this is where it all starts, and often ends. A top executive must be mandated to become the sponsor of corporate innovation. They must be passionate about new ideas and talented in coaching and inspiring people. They should be responsible for shaping and executing innovation strategy across all three levels.

Organizational agility: I mentioned a hub and spokes model for a CoE for Level 1 Innovation that is key for harnessing all three levels of innovation. Every employee should be given the opportunity to become a citizen innovator, by democratizing tools such as Robotic Process Automation, data analysis, and AI, and putting them in the hands of the many. Agile organizations are best for executing innovation at scale. It is therefore recommended that digital transformation aligns with agile transformation, where teams are incentivized to innovate and experiment. In an agile transformation scenario the hub and spokes model for Level 1 Innovation evolves into an organizational enabler: the spokes are now the agile teams ideating, developing, and testing new ideas. A key lever for the necessary cultural transformation for becoming agile is to rethink performance metrics and to reward innovation, and that means rewarding failure as well as success.

Skills & competencies: the innovative organization is a skills-based organization, where continuous reskilling and upskilling is part of the company’s DNA. Moreover, it is an organization that is permeable to internal as well as external talent, capabilities and skills on a demand basis. Skills are the new currency of operational excellence, including excellence in innovation. A way to do this is to develop a “knowledge architecture” that represents how skills are distributed across the company as well as how they dynamically match against the requirements for innovation. A marketplace for skills that connects external and internal talent can accelerate ideas through the innovation process and ensure success for those ideas that are shown to be viable.

Technology stacks: Speed is more important for innovation than completeness, maintainability, or scale. Reducing the time from ideation to market testing is key to success. Speed enables to increase the number of experiments, and therefore the chances of success. To increase speed organizations must have the right technology stacks in place. A cloud-native organization leveraging microservices and composable systems to provide sandboxes for quick experimentation is, in effect, a platform for fast innovation. All C-suite stakeholders must align to get technology purpose-ready for innovation, for instance by participating in a “Technology Council” that shapes strategy around data, software development frameworks, and integrates emerging technologies from innovation levels 2 and 3.

Ecosystems: one of the greatest barriers that I have encountered in my numerous discussions with executives across all industries when it comes to collaborative innovation is the fear of “opening up”; for example, exposing corporate data to external partners, or joining consortia to share know-how, or partnering with startups to serve a company’s customers. The fear is mostly abound losing competitive advantage by sharing the company’s crown jewels or failing customer expectations. But in an era of exponential technological disruption a wait-and-see strategy is a lose-the-train strategy. Companies need to find ways to meaningfully co-innovate with their existing partners (e.g. suppliers, customers) as well as with new ecosystems of value, in an ethical and compliant way. Legal and privacy experts are instrumental to getting this right while working together with business leaders and innovation evangelists for the common cause.

Getting corporate innovation right needs an operating model that aligns five critical success factors: strong C-suite Sponsorship, organizational agility, skills& competencies, the right technology stacks, and networking with ecosystems (see figure 2). It is hard, but not impossible. Many popular quotes capture aspects of what means to combine things in new ways, but my favorite is by Dr Albert Gyorgyii, the Hungarian biochemist who discovered vitamin C. He said: “innovation is seeing what everybody has seen and thinking what nobody has thought”. All the big and small ideas that will change our world are in front of our eyes. We just need to see the beautiful patterns of their combination, and act.



George Zarkadakis

PhD in AI, author of “Cyber Republic: reinventing democracy in the age of intelligent machines” (MIT Press, 2020), CEO at Voxiberate @zarkadakis