AI and Business Model Shifts
While the vast majority of generative AI implementations today are embedded within existing business models, AI’s transformational potential is rooted in its ability to fundamentally shift industries from one business model type to another. Specifically, as business models move in directions that increase market affordability and access, value is created for consumers, founders, and investors. What follows is an illustration of how that plays out.
There are more business model types than the ones introduced here, but I will include three that have been well documented and encompass a vast majority of businesses. Consider a law firm - a group of trained experts that charge by the hour when providing bespoke service to each client. This is a solution shop1 business model in which experts are paid to solve hard problems. This business model is typical across consulting firms, doctors offices, creative agencies, and many other professional services firms. Because of the unstructured nature of customer problems, these businesses usually make money on a fee-for-service basis.
Outside of solution shops there exists a vast array of businesses that take inputs, make improvements, and sell outputs. Restaurants, manufacturers, software firms, and many service businesses fall in this bucket. These firms improve or transform raw materials with known costs, and the customer pays for the labor and process steps that turn those inputs into something of greater value. These have been called by various names, but I think it is easiest to refer to them as ‘input-output’ businesses. With established processes for producing outputs at known costs, input-output businesses generally increase market access to those who couldn’t afford to play in the solution shop game.
Finally, there are network businesses that help customers transact with each other. Ecommerce and social media fall into this bucket, and what is being paid for is both access to a network and well-facilitated transactions. These businesses usually make money with a take rate on transactions, membership fees, or through advertising. Network businesses typically represent the broadest form of market access, where supply and demand from various market segments can come together and transact with minimal overhead.
Each of the above business model types is defined by different resources and processes. For example, solution shops are defined by their resources (the experts) which are used to solve problems. The experts spent a lot of time and money developing their expertise, and this valuable resource is monetized accordingly. Unlike solution shops, input-output businesses are defined by their processes which have been refined to provide predictable outputs of value. Network businesses are a function of both resource and process developments, including the processes associated with facilitating transactions and the resource of readily available supply and demand on both sides of the network.
When a technology emerges that alters the fundamental economics of the resources and processes embedded within an industry, a business model shift can occur. Importantly, these shifts expand and create new markets as previous consumers and producers now have access to resources and processes that were previously available only to a select few.
As a historical example, retail evolved primarily from solution shops (getting clothes made and designed by tailors) to input-output businesses (retailers who could mass-produce clothes from material inputs) and then to networks (buyers and sellers transacting directly or via intermediaries such as Amazon or Shopify). These shifts were precipitated by technological changes (textile mills, sewing machines, internet, shipping, etc) that redefined the economic structure of resources and processes embedded in the production and distribution of retail products. These business model shifts came hand in hand with massive market expansion, tremendous benefits for consumers, new wealth for entrepreneurs and investors, and many new adjacent markets and infrastructure.
Because of its ability to fundamentally alter resource costs, as well as upend the traditional process constraints that deny many market participants from accessing certain goods and services, Generative AI reflects a technological foundation that will facilitate business model shifts across a range of sectors. Solution shops will become input-output businesses, and input-output businesses will become network businesses. Participants in these shifts stand to be rewarded.
The evolution of solution shops by AI was first introduced by Scott Belsky who described the death of time-based business models. More broadly, as generative AI dramatically lowers the cost of developing or applying expertise - the resources that underpin the economics of a fee-for-service solution shop - then solution shops can become input-output businesses. This might look like ad agencies that are not driven by expensive creatives developing unique storyboards with high-end design tools but instead operate like typical SaaS businesses.
Once the resource frontier is conquered, the next step is processes. Recent research on Compound AI Systems highlights how this shift may occur, with breakthroughs in performance not coming from better data, inference or training but through system design. As these processes are democratized, a greater number of people will be able to replicate input-output businesses allowing network businesses to emerge.
The pace at which AI will facilitate business model shifts is unique to each industry as the resource and process economics differ across sectors. Additionally, there will be more juice to be squeezed out of existing business models that will become more efficient as AI gets infused, whether it be financial trading, material design, or drug development. But the abundance era that awaits us is not purely from efficiency gains but through market expansion and creation. These outcomes are a result of business model shifts, and I’m eager to speak with anyone who is building or observing this in real time.
Solution shops are part of a business model taxonomy originally introduced by Oystein Fjeldstad who referred to them as value shops. The term ‘solution shop’ was used by Clay Christensen to refer to the same thing, and is a more descriptive title.