By Bernard Casse, Ph.D.
Reposted with permission by rios.ai

The world is experiencing a “black swan” event — an extremely rare event with very severe impacts. COVID-19 has caused unparalleled disruptions to our lives and to many businesses. It is testing every organization’s nimbleness to react to a rapidly changing landscape. Darwin’s evolutionary theory is now unfolding before our eyes — the survival of the fittest — either companies restructure and adapt or die. Even though the pandemic has created challenges, it is also creating an opportunity for companies to re-invent themselves. This re-invention is embodied in many forms — innovative products and services that this new world will embrace, improved processes, novel business models, or new value propositions. But all this can only stem from one thing: experiments. Building forward-looking and high-growth businesses is all about taking calculated risks and large-scale experimenting. To tell companies to take risks in an already ‘risky world’ can appear like a very contrarian idea (or even an unsound idea), but it stands on timeless principles. Experimentation “done the right way” is the only way to reimagine companies in this new reality. All wildly successful companies (e.g., Amazon, Facebook, Google, etc.) ran experiments and tested many “crazy” ideas, even in unprecedented times. To quote Jeff Bezos, Chief Executive of Amazon: “Our success at Amazon is a function of how many experiments we do per year, per month, per week, per day… Big winners pay for so many experiments.”

Getting to a big payoff by capitalizing on the “fat-tailed” probability theory

When we run a lot of experiments, we expect tremendous success as the ultimate outcome. But we should also bear in mind that there are going to be failed experiments, and we must somehow get comfortable with those. Most people aren’t. Even at Amazon, some experiments were epic failures (e.g., the Amazon Fire Phone), but this giant experimental engine yielded successes even beyond Bezos’ imagination (e.g., Amazon Web Services / AWS). AWS currently generates revenues north of $35 billion. I have been fortunate to work at Xerox PARC and learned firsthand about risk taking, experimenting, and open innovation. PARC has a well-known track record of experimenting with blue sky ideas — it enabled the company to pioneer computer-centric technologies (i.e., Ethernet, GUI, Alto PC, WYSIWYG, etc.) and dominate the field of computer science for almost half a century. This notion of getting to a big payoff by running thousands of trials is rooted in the “fat-tailed” probability theory (i.e., a distribution with a “tail” that is “heavier” than an exponential function; the bulk of the gains come from a rare event that’s difficult to predict at the onset). This “fat-tailed” probability concept is ubiquitous — it is applied to areas ranging from financial returns (investments), insurance payouts, to baseball matches. Nassim Nicholas Taleb, professor at NYU and author of “Antifragile: Things That Gain from Disorder”, summarizes this idea in the graph shown below. According to Prof. Taleb, 1 in 1000 trials can lead to 50% of the total contributions and critically we do not know the winner ahead of time.

Develop an open innovation strategy to get an influx of ideas

For companies that want to be more resilient, agile, and stay ahead of the curve, they constantly need to innovate — and innovation starts with experimenting with a series of small pilots. The idea is to evaluate these pilots rigorously, rapidly terminate those that do pan out, and scale up those that work. To get to as many pilots as possible (i.e., our fat-tailed probability), or to get an influx of as many ideas as possible, companies need to embrace open innovation (i.e., drawing on resources and/or working with people outside the organization to fill in gaps). Many forward-looking companies have realized that not everything has to originate from within. Open innovation has a host of benefits — it can reduce costs, reduce risks, accelerate time to market, increase differentiation in the market, and create new revenue streams for companies. The best sources for open innovation are academia, R&D organizations, and importantly startups, like RIOS. The appeal with startups, compared to purely research-oriented organizations, is their commercial bent. It is important to note that to remain competitive, companies need to constantly seek “bleeding-edge” innovations — not the off-the-shelf technologies that all their competitors have wide access to.

Minimize financial risks when running experiments

To reduce financial risks and maximize return on investment, each pilot should be as cost-effective as possible. It is vital that any failed pilot has negligible impact on the company. “Cost-effective” does not necessarily mean zero cost, although many established companies will seek unpaid pilots to maximize their gains. We cannot really blame them if we think solely about the economics. But they fail to bear in mind that paid pilots create a greater incentive for startups to go the extra mile, and spur investors to invest more capital. Importantly, it also ensures that the companies themselves devote internal resources for these pilots — when there is a financial commitment, companies take pilots very seriously and are more vested in seeing these projects succeed, which in turn increases the likelihood of success. However, there is a way to structure unpaid pilots in a mutually beneficial fashion — the key is to tie pilots to a real promise of a substantial commercial opportunity down the road. For startups, there needs to be a reward at the end for the risks that they are taking.

By conducting experiments at large-scale and avoiding the common pitfalls (which I’ll discuss in my next medium post) will enhance the odds of companies in getting to big winners, re-imagining themselves and be ahead of the curve even in unique times.

About the Author

Bernard Casse, Ph.D., is CEO and Founder of RIOS (www.rios.ai). Dr. Casse is a seasoned leader and entrepreneur, leading a world-class team developing dexterous, AI-powered robots. RIOS is a technology company helping global customers automate their factories, warehouses, and supply chain operations by deploying a new class of AI-powered and dexterous robots. RIOS blends design, engineering and robotics.