Ready or Not, Here I Come! 5 things AI is teaching us about innovation and market behavior
Updated: Jun 13
“We shape our tools, and then the tools shape us.” – Winston Churchill
The first time I used a search engine was magical. In that 1990s moment it was game-changing to find “any” website, even if you didn’t know the URL. Few modern innovations have had quite the same magic feeling (at least for me) as my first online search.
This year a lot of people are having a similar magic moment with ChatGPT. It’s rare for technology to stop society in its tracks with such power and elegance. As always, there are things to be both excited and cautious about. It shouts “ready or not, here I come.”
An upside of ChatGPT is that it’s helped to update mainstream society’s intuition about what technology can do.
ChatGPT may very well be the first contemporary application of AI that anyone (technical or not) can understand. It’s a tangible, accessible example of the kinds of things that are now possible with enough creativity and computing power. Only 18 months ago it would have been painfully difficult to explain ChatGPT to a non-technical person, likely leaving them confused and skeptical. There was no basis upon which most people could form an intuition about it, or anything like it.
In a similar vein, my team uses AI to study privately-held companies (such as startups) and the markets they occupy. Looking at private markets though the lens of data science has led to new lessons, but it’s often painfully difficult for people to understand where we’re going because they have little intuition about the AI we use to get there. It can feel a bit like trying to explain Uber to someone in the 1800’s. “That’s impossible,” they might object, “you can’t possibly hail a ride by touching a piece of glass.”
Thanks to ChatGPT and applications like it, society is updating its intuition about what technology can do. AI is also reshaping how we understand market behavior.
For example, here are five things AI is teaching us about innovation and market behavior:
1. Markets are far larger and more complex than most people imagine. If I asked you to guess how many telehealth companies there are in the world (like Teladoc), what would you estimate? 25? 50? 100? 1,000?
The last time we checked, there were more than 11,000. That’s an enormous number. As we use advanced computing to mine and analyze new types of market data we’re continually struck by the scale and complexity of markets. We rarely encounter markets with company populations in the tens, or even the hundreds. Today we’re in a world of thousands, or even tens of thousands.
Traditional methods of market scouting, opportunity evaluation, strategy and competitive analysis – like SWAT - are increasingly inadequate. As a result, companies are vulnerable to being blindsided by realities they simply can’t account for.
This begs a need for new analytics and computational tools to help managers better understand and navigate market realities at-scale. Said differently, AI taught us that markets are huge, and that we need even more AI to understand them. Clever AI.
2. The more crowded a market grows, the more that luck (versus skill) determines who wins or loses. Skill matters. Obviously, it’s important for businesspeople – especially entrepreneurs – to be good at their jobs. This hasn’t changed. What has changed, however, is the link between effort and reward.
As markets become bigger, more competitive and more complex, the odds of market leadership and even survival itself decrease for each competitor. At the same time, the role of luck increases when determining who will win or lose. Think of it this way, the more competitive a sports league becomes, the more that random events like a star’s injury, bad weather or one bad call by a referee can make or break a championship.
As a result, relying on inductive causation about which business will win, and why, is increasingly less realistic. This lesson is for executives, venture capitalists and other managers who place bets on businesses. At a portfolio level, you have to abandon the temptation to believe you know who will win, or why. When the road zigs and zags, having strong beliefs, putting on blinders and staying the course is a great way to end up in a ditch.
Today the bedrock of corporate strategy needs to be an assumption of ignorance, not knowledge. It’s the opposite of how most executives have been trained to think. Only then can they stand a chance of constructing a portfolio with the right level of flexibility, hedging and fluidity that are necessary to maximize the odds of success, regardless of how the dust settles.
3. Market leaders determine the dominant technology, not the other way around. This is perhaps one of our more heretical findings.
For as long as I’m aware, the bulk of business and academic thought leadership has assumed that dominant technologies produce market leaders. In other words, once in a while a superior a new technology shows show up and rises to dominance, putting the companies involved with it in a position to win.
This is why companies spend inordinate time and money trying to figure out which technology will win. Will it be RFID or NFC? CPUs or GPUs? Lithium ion, fuel cell or supercapacitors? Solar or Wind?
Dominant technologies can, indeed, help related companies succeed; a rising tide can lift all boats. After all, it certainly doesn’t pay to bet your company on a technology that goes nowhere. That said, we’re finding this notion (i.e. a superior technology rises to dominance, causing related companies to win) to be the exception more than the rule.
Let me explain… in the 1980s when Bill Gates made his historic deal with IBM that put DOS into most home computers, I don’t think anyone would have considered DOS the best operating system. Apparently, Gates ran out and bought DOS for around $50,000 because he needed something quickly for the IBM deal. It was originally called QDOS which stood for Quick and Dirty Operating System.
As IBM dominated home computing, DOS became the de facto technological standard in the industry. Said differently, Microsoft’s market leadership (the fact that it won commercially) led to DOS becoming the dominant technology, not the other way around.
Regardless of how winners become market leaders in the first place, by vice or virtue, their position of commercial leadership establishes the dominant technology – theirs. Others in the industry must then operate within the downstream constraints, technical or otherwise.
Meanwhile, anyone in the late 1970s who would have tried to guess which operating system would win based on technical merit or industry trends would have missed the mark, badly. This is partly why the best technologies often fail and mediocre or even inferior technologies can surprise everyone by becoming dominant. If you want to know which technology will win, you need to figure out which company will win (not the other way around).
4. Markets are growing more interconnected, which creates more extreme, zero-sum outcomes.
We all know markets are increasingly interconnected. Digital technologies are becoming ubiquitous; there used to be “tech” and “non-tech” industries. Today that distinction seems almost quaint. Markets also used to be more divided by subject matter (where boundaries continue to blur), supply chains (which continue to mix) and geography (which continues to flatten).
Most businesspeople also can’t escape the feeling that markets are getting more extreme. For example, extreme outliers such as “Unicorn” startups with billion-dollar valuations have become almost ordinary. The first Unicorn list I remember had around 40 companies.[i] That seemed like a lot.
Today, even if you eliminate all the Unicorns that have been acquired, taken public or are more than 10 years old, you still end up with over 200 of them. Not only are there more extreme market outliers but they’re rising and falling at an accelerating rate (shortening lifespans).
To put a cherry on top, markets are also becoming increasingly zero-sum (or ‘winner take all’), where the top 1 - 3 market leaders take the lion’s share of customers and profits while everyone else gets little to nothing. For example, there isn’t much value in being the world’s 10th search engine (with Google alone having an estimated 93% of search market share[ii]).
While all these factors can make markets seem wild and out of control, they aren’t a coincidence. It’s interconnectivity itself, as a property, that leads to more extreme and zero-sum outcomes. This is especially the case to the extent that markets have network effects. I’m reluctant to over-simplify this topic but also don’t want to launch into a diatribe about modeling complex adaptive systems, so think of it as the good ol’ “butterfly effect”.
The more that things in a system are connected (versus not connected), the more interactions and feedback across the system can transform tiny changes in inputs into enormous differences in outcomes.[iii] Or, in a world where all your friends are on Facebook (…that world being 10 years ago…), the more incentive you have to join it and less incentive you have to join a different social network where you don’t know anyone. So you join, which incentivizes anyone you know to also join, and so it continues exponentially. Meanwhile MySpace, Orkuk, Bebo, Google+, Friendster and Vine go out of business. It was the interconnectivity – in this case between you and your Facebook friends – that caused an extreme outlier with zero-sum impact on a generation of companies.
This too has affronting implications for managers and investors, especially in large companies. Understanding the more extreme, more volatile, more zero-sum and less predictable nature of markets shouldn’t just change their approach to growth; it should change everything about their approach to growth. For example, at a minimum, large companies need to be structured to rapidly identify and capitalize on ever-faster, ever-larger, ever-shorter growth opportunities at ever-increasing volumes.
That’s a good segue…
5. In the past, competitive advantage stemmed from a company’s ability to create internal value. Today, competitive advantage is increasingly about a company’s ability to identify and capture external value.
Competitive advantage is the idea that a company can create more value within its walls than others can create outside its walls. In other words, the fundamental premise of a business is that it can create internal value in a way that produces a sustainable position in the market. The bigger a business grows, the more its advantages multiply.
While this remains true for many businesses, we’ve found that it breaks down when companies reach certain stages in their lifecycles. In other words, the classical theory of competitive advantage has upper limits. When companies hit these limits their growth stalls, often followed by a relatively abrupt fall rather than a soft decline (think… Atari, Blockbuster Video, Borders Books, Circuit City, General Motors in 2009, Kodak, Nortel, Polaroid, Silicon Graphics, RadioShack, Sears, Toys R Us).
The acceleration in market sizes and complexity is undermining the very idea of classical competitive advantage itself.
No company, no matter how large, has more resources, talent or innovation than the aggregate of all its competitors combined. This is true for WalMart, Microsoft of even Samsung. Even the mighty Amazon, with its trillion-dollar market cap and $73 billion in R&D last year, is smaller than the estimated $16 trillion dollar e-commerce market as a whole[iv].
When the growth of a large company stalls and its yearns for a cadence of new billion-dollar businesses becomes deafening, it’s a matter of necessity to move from an inward-looking view to an outward-looking one. If more growth and innovation is happening outside, they need to fish where the fish are.
This means shifting the company’s own view of its competitive advantage from one of creating internal value, to one that identifies and captures external value in a differentiated way. It’s a 180 degree turn.
Identifying and capturing external value is a different skillset than operating an existing business. The problem is, most operating executives don’t have the skill set, at least not at an elite level. Even if they do possess the skill, they can run smack into limitations caused by their own businesses when they attempt to put those skills into practice. It isn’t about making a few little tweaks so much as it’s about a massive overhaul of the business, its goals, its identity, its talent and its organizational structures.
AI is giving unprecedented visibility into new markets, resulting in discoveries with existential implications for managers, executives and investors today.
Data-centric technologies continue to evolve and update society’s intuitions about what is, and isn’t, possible; be they my team’s market analytics, ChatGPT, search, or whatever has given you a feeling of magic. These technologies can vastly improve our lives, although they also have a way of exposing inadequacies and gaps between the world, as it has become, and the world as we assumed it was. As always, there are things to be both excited and cautious about. It shouts “ready or not, here I come.”
[i] Unicorn List: https://www.cbinsights.com/research-unicorn-companies
[ii] Estimated Google Search Market Share: https://gs.statcounter.com/search-engine-market-share
[iii] For a good exploration of these dynamics, see: Morris Holbrook (2003). "Adventures In Complexity: An Essay on Dynamic Open Complex Adaptive Systems, Butterfly Effects, Self-Organizing Order, Coevolution, the Ecological Perspective, Fitness Landscapes, Market Spaces, Emergent Beauty at the Edge of Chaos, and All That Jazz." Academy of Marketing Science Review, Vol 6. 1-15.