In today’s globalized world, competition is becoming more and more intense. Products are getting better and cheaper. Can this race be won? How do you protect yourself from being disrupted by new, innovative products?
During my time in Venture Capital and Consulting, I have seen many companies struggle to leverage innovation for a competitive advantage. Today, there is no doubt that Artificial Intelligence has the potential to become the most disruptive technology of the century. Companies do well to prepare for this. But, AI has a long history of being “the next big thing” so why should it be different this time?
Since the ImageNet challenge 2012, real-world applications have emerged, and AI grew out of being only a cool toy for geeks at university research labs. Today AI is slowly used by “old dinosaur” corporations who integrate it into their manufacturing or products. AI has moved from hype to production. If this hype and broad adaption continues, the golden age for AI has started. As a Co-Founder of WhatToLabel, an ETH Startup that helps companies to manage and prepare their AI data, we have some of the most innovative companies as our customers. Areas of AI applications include digital pathology, visual inspection, and autonomous driving. Thus, from my own experience, I’m able to tell you that AI has already started to become a constant part of our lives. Whether you talk to Alexa to order from Amazon or you use the driving assistance system of your car — you are interacting with a smart AI-based product. As we can’t deny the advance in technology anymore, it is of crucial importance to understand as a business leader, whether technology can pose a threat and how far its adoption has already spread.
To assess the widespread adaption of AI, it is best to have a look at the Top 100 AI Landscape analysis of CB Insights below. There are numerous AI startups active in a variety of fields today. Almost no industry is untouched, which shows how advanced the adaption process already is.
It is possible to segment AI applications by industry, which is essential if you want to analyze use cases in specific fields. However, at a lower level, it is imperative to determine whether those applications for AI are going to generate revenue or save costs for you. It is also crucial to understand what the strategic importance of those applications is for your business.
A thorough risk assessment, in line with your individual business situation should be conducted. For that, I recommend to stick with the proposed strategic approach of Raffi and Kampas:
Disruption isn’t inevitable, however. We’ve developed and tested a tool that can help companies detect potential disruptive innovations while management still has time to respond effectively. There’s nothing magical about it — no crystal ball that allows you to see deep into the future. But it does get a group of smart, opinionated managers to sit down together to think systematically about threats to the core business, and to surface ideas about how to avert or co-opt those threats.
You should ask yourself the following questions:
What we’ve learned in business school still applies to AI. We just need to transfer our knowledge to the current development. As Michael Porter already showed in 1985 a company can use technology for two different objectives and in two different ways.
The objectives to use AI in our specific case here can either be:
Whatever your objective as a company is, you must decide whether you want to be a follower or a leader in this race. Depending on your industry one or the other might be more important. The main difference lies in the strategic instrument one should choose to achieve the technological leader- or followership. As Robert and Berry proposed already in 1985 you can acquire innovation either through external innovation, internal innovation, or strategic partnerships.
As an example if you are a car manufacturer. You want to aim for technological leadership in differentiation (e.g., autonomous driving). It’s going to be hard to acquire that technology externally. Licensing is not a feasible option since that would leave you at the mercy of the licensing company. Thus, you will need to build it yourself.
I want to emphasize that both objectives: Using AI for cost savings as well as revenue generation are crucial to secure your companies long-term viability. Thus, a company should always pursue both.
Hereafter I will elaborate on these two objectives to highlight potential areas of application in your business and what strategic instruments should be used.
This one is probably the easiest way to integrate AI in your business. Why? Because unless you want to achieve technological leadership in cost advantage, you already have plenty of plug-and-play solutions to choose from. You rarely need to establish in-house tech competencies. The list below shows example areas as well as the companies who provide solutions for these:
However, there are use cases where it makes sense to go for technological leadership in cost advantage. For example, if you are a producer of a commodity good such as corn. In this case, we would assume that customers make their decision mainly based on price. This leaves you with almost no other way to differentiate. In this case, you want to be the cheapest producer to beat everyone else. It makes sense to automate your production with tools no one else has (e.g., computer vision.based ways to apply fertilizer & pesticides more efficiently).
Another reason can be if understanding customer needs and forecasting is so important for your business that you want to have the best solution in-house. An example for such an acquisition motivation is the recent $110M acquisition of Celect by Nike.
Whatever your motivation is, there is no doubt that AI can play a crucial role in minimizing costs and gaining a cost-based competitive advantage. But, a company should never solely focus on cost optimization. It is important to use all the potential of a new technologies also for increasing revenues.
This one is a bit more tricky. Here we are rarely able to purchase solutions to create new sources of revenue. Sometimes it can make sense to license a solution. However, as we have seen in the autonomous driving example, if those solutions are part of the core of one’s business, that can be a dangerous game to play. Technological followership is rarely the right decision in case of disruptive innovation. How should a manager deal with the current revolution to AI?
It can make sense to acquire leading startups in the field, but there is hardly a way around having certain competencies in house. This is often a tough decision because both acquiring startups and building up competencies are very expensive in the AI industry. Machine Learning Engineers are some of the highest-paid employees right now, while AI startups enjoy the highest valuation. Nevertheless, where there is a price, there is a reason. In this case, it’s quite simple the value you can generate through AI is also widely considered as one of the highest by many experts.
Here I have gathered examples where startup acquisitions might make sense:
Mining & Commodity industry: A proprietary way for metal & mineral prospecting, which is better than everyone else in the industry, would undoubtedly provide a player with a competitive advantage. Surely, is this field already filled with numerous startups such as, e.g., Earth AI, Goldspot, Kobold Metals.
Fashion & Apparel industry: A proprietary way to design more trendy products through AI. This a very novel field, but we know from examples that AI can successfully be applied to create new recipes for creative dishes.
Healthcare industry: A proprietary way to find new drug formulas to treat diseases. Here another example could be to provide physicians with software for AI-enabled digital diagnostics. Here we have players such as Roche, which are actively developing solutions themselves or startups such as paige.ai, Caption Health, and QuantX .
Retail industry: The retail industry has and still is being disrupted through e-commerce. However, the disruption did not stop there. There are initiatives such as Amazon Go, AI Retailer Systems, and Standard Cognition. Thus, the retail industry would do well not to make the same mistake twice by being disrupted again by the same e-commerce giant.
Automotive industry: To stay with my example of autonomous driving. In this field, there are two different approaches. We had acquisitions (e.g., Cruise by GM) and corporations that develop solutions mainly themselves (e.g., Renault). Both methods have advantages and disadvantages. The beauty is that you can combine different instruments. For example, GM acquired Cruise to get access to its technology. But, GM is still investing heavily in autonomous driving itself. Another example is the Google subsidiary and autonomous driving company Waymo which developed most of its technology in-house but still conducts acquisitions.
It is crucial that every time one looks at costs he/she also looks at revenue and vice versa. A change in costs can lead to an increase in revenue. Thus, it is essential to always have the bottom line in mind while evaluating strategic decisions such as integrating technology in one’s business.
In a nutshell, we can conclude that there is no industry going to be untouched by artificial intelligence. Managers do well to prepare their company for the next years. It is important not to rush such important decisions but rather think thoroughly about different fields of application in your own company. Once you have made a decision, it’s important to execute fast and bold since in an increasingly intense economic environment competition is fierce and slow player risk being consolidated by the fast movers.
Matthias Heller, Co-Founder WhatToLabel.com
Thanks to Mara Kaufmann and Igor Susmelj for reading drafts of this.