How big is the market for AI-related services? 

*For those too young to recall the internet bubble of the late 1990s, this is the term Alan Greenspan used to describe market optimism that lacks the real foundation of fundamental valuation, but which rests instead on psychological factors. 

Sizing any professional service is inherently difficult: Unlike breakfast cereal, you can’t put it in a neat and tidy box; unlike wine, you can’t bottle it. Most models of the industry rely on high-level assumptions: Consulting represents a percentage of GDP; X% of a bank’s costs on average go on consulting; a firm’s own share of wallet in a single client can be extrapolated across an entire industry. Estimates of forecast growth that rely entirely on how much clients think they will spend are likely to be overstated by senior executives’ bias towards optimism. Meanwhile, technology vendors, keen to attract investment, have a strong self-interest in talking up particular markets.  

You can argue that the reliance on assumptions in estimating market sizes isn’t confined to consulting and professional services. But there are three specific challenges to sizing this industry: 

Capacity is not infinite: The notion that the services sector can enjoy endlessly increasing returns on investment was first proposed in the mid-1990s—not uncoincidentally alongside the internet bubble. It assumed that services didn’t need raw materials or factories and would therefore not be subject to the diminishing returns of, say, the manufacturing sector. But this turned out not to be the case. Ideas require a different type of factory, or they become outdated as quickly (perhaps even more quickly) than goods. They also still rely on people to deliver them. The application of AI to the delivery of consulting and professional services may change this but hasn’t done so yet. 

Definitions vary: One firms’ definition of an AI-related service may be quite different to another’s. Take, for example, demand for help understanding the legal and ethical risks of deploying AI at scale. One firm might see this as an aspect of AI strategy and deliver it with technology strategists. Another may see it as more closely aligned with legal services. 

Most of the AI-related services market is embedded in other services: One of the defining features of the consulting and professional services industry in the last decade is that most growth has come from multidisciplinary work, which combines multiple sub-services and capabilities in new ways, rather than from traditional, standalone services. The result of this is a two-tier market. Imagine an iceberg: The ice you can see above the waterline represents demand for services that are explicitly badged by clients as being AI-related. A far bigger market exists below the waterline that consists of projects that have an AI component embedded in them—a supply chain management project, for instance, that incorporates a new AI-enabled logistics model. This hidden market is uncertain but simultaneously holds the potential for above-average growth: Growth in the AI-related services market depends on clients moving their expenditure on these services from below to above the waterline. To use our supply chain management example, over time, as more of the work is AI-related, clients will shift from hiring consultants to do supply chain management work that involves AI to doing AI-related projects for their supply chain. This is important because only some of the growth in AI-related services will be net-new: In fact, the lion’s share will come from rebadging—i.e., cannibalising—other services.  

Irrational exuberance occurs when enough people (clients, consultants, technologists, investors) are sufficiently optimistic about a market to create a positive feedback loop, a tipping point that quickly becomes a bubble. These three factors (capacity, definitions, and the extent to which embedded services create the prospect of rapid growth) explain why the consulting and technology services industry is especially vulnerable to irrational exuberance—and therefore why it’s important for consulting and technology services firms to be both cautious and ready to expand. 

So, how big is the market for AI-related consulting services?  

Our model reflects the factors we’ve discussed. To take capacity into account, it’s a supply-side model, built bottom-up by estimating the number of consultants working for specific firms, in specific countries and industries. We use simple, consistent keywords right across the extensive research that underpins the model to remove any definitional biases and identify how many people have a specific skill. Finally, our forecasts are based on the expectations of client organisations, the extent to which it’s possible for the supply side to grow its raw material (i.e., hire/train people with the relevant expertise), and the likely speed with which new, AI-related services will cannibalise traditional services.  

The result? Excluding implementation work, we think the visible part of the AI-services market in 2024—the tip of the iceberg—was worth around $12bn.  

What can firms do next? 

Want to map out the AI opportunities? Source’s wealth of market sizing data can help you track areas of growing demand, from technology to tax, risk to regions, and pinpoint the most important prospects for your firm. To find out more, get in touch.