Despite generative AI’s remarkable advances in recent years, adoption of the technology remains largely confined to the same large corporations that have historically led the way in deploying emerging technologies. But GenAI is evolving and so, too, is the company profile best suited to extract value from it. Increasingly, it is mid-sized companies that possess the right balance of resources and agility to accelerate adoption, drive meaningful outcomes, and reap the benefits of GenAI as the technology matures.
Large enterprises currently dominate GenAI adoption. A 2024 BCG survey found that 46% of large companies reported mid-to-high levels of GenAI maturity, compared to just 28% of mid-sized firms. With greater access to talent, data, and capital, larger players have had an edge in taking on traditional, resource-intensive AI initiatives.
But that advantage appears to be fading. Recent research from the London Business School (LBS), the Institute of Directors (IoD), and Evolution Ltd suggests that mid-sized companies—particularly ambitious, growth-focused firms such as those backed by private equity—are well placed to overcome adoption barriers. Mid-sized firms are more agile than large companies, a key attribute as GenAI has become more accessible, making these companies prime candidates to leverage the technology and unlock their potential.
On the whole, while such firms are still behind, they may be poised to rebound. Research by Oxford Economics found that only a quarter of mid-sized companies surveyed had adopted AI in 2023 but 51% were planning to adopt AI in 2024; the adopters were expecting it to improve their outlook, specifically in new products and services (43%) and marketing and sales (48%).
Until recently, it was (very) large companies that benefited most from GenAI, as the advantages of scale outweighed the challenges of organizational complexity that accompany size. Yet as technology evolves, large firms find themselves slow to adjust. Extensive layers of management, entrenched processes, and siloed operations can slow down the adoption of fast-evolving technologies like GenAI.
In large corporations, GenAI implementations can suffer from “death by a thousand pilots,” in which individual teams or functions develop proof-of-concept products and tools yet do not manage to scale them due to the enterprise complexity and lack of clear governance. As a result, large companies frequently struggle to fully realize the potential of new tools despite extensive investment in digital transformation efforts.
Mid-sized firms, by contrast, can benefit from leaner structures that allow for quicker decision-making and implementation, given the right leadership and governance. Their agility, when combined with the right strategy, enables them to adapt more quickly to new developments in the technology and more easily operationalize GenAI. (Mid-sized firms here refers to companies with revenues between $50 million to $1 billion, and although the precise definition will vary from country to country, this broadly refers to companies that are still small enough to have relatively simple operations and remain agile.)
While the benefits of size and scale provide once-decisive advantages in access to specialized talent and capital-intensive infrastructure, the evolution of GenAI as a technology—particularly the development of GenAI as a service, the emergence of streamlined platforms, and growth of customizable models—is creating a more level playing field between mid-sized and large firms.
GenAI providers, for instance, are significantly reducing the need for up-front investment and extensive IT capabilities, by offering models and infrastructure as a service. Streamlined platform solutions like Google Vertex AI and Snowflake also simplify the AI ecosystem, providing integrated tools for data management, model customization and deployment, all of which lower technical barriers and accelerate time-to-value.
The advance of customizable models through technologies like retrieval-augmented generation (RAG), meanwhile, allows mid-sized firms to leverage their proprietary data effectively without an army of in-house data scientists. Much of the coding needed to build traditional AI has been replaced with natural language prompt engineering to create GenAI-powered tools tailored to the company’s content, expertise, and workflows.
In addition, updates to existing software platforms including ERPs and CRMs are incorporating AI features, giving easy access to AI functionality on the existing tech stack. Mid-sized companies are well placed to adopt these rapidly, given they generally have less complex and less customized instances of software, so integrating new releases is simpler and faster than for larger companies.
Beyond adoption, mid-sized companies are well positioned to create value from GenAI, as it may help them tackle the operational constraints that often hold them back. Mid-sized firms often struggle to attract specialized talent, such as data scientists, and do not have the scale to make it economically viable to hire a full-time position. GenAI tools can expand the capabilities of existing staff, as demonstrated by a recent BCG experiment where management consultants were each asked to complete three basic data-science tasks outside their core consulting capabilities: data cleaning, predictive analytics, and statistical understanding.
Using GenAI to perform the tasks immediately expanded the consultants’ aptitude beyond their current abilities. These augmented participants showed a 13- to 49-percentage-point improvement over those working without GenAI and came within 12 to 17 percentage points of the benchmark for data scientists. Function- or role-specific tools are now entering the market and enabling companies to further expand the capabilities of existing employees. Sisense, for example, enables companies to build semantic data models without coding that users can then query through natural language queries, enabling managers to incorporate data-driven insights into their decision making without the need for data analysts or data scientists.
Another constraint often found at smaller companies is a lack of sufficient proprietary data to create differentiation. The recent study by LBS, IoD and Evolution Ltd. found just 56% of smaller firms with annual revenues of £10 million to £50 million stated they believe that proprietary knowledge is somewhat or extremely important to their business, compared with 72% of mid-sized companies with revenues over £50 million. Large companies, on the other hand, are already using traditional AI to extract value from proprietary data, having invested in cleaning and curating datasets.
Mid-sized firms, however, often have a wealth of unstructured data—from which they’ve struggled to extract value. A mid-sized company, for example, may have handbooks for customer service agents outlining product details and troubleshooting tips, along with transcripts of real customer support calls. With GenAI, such a firm could now unlock those insights without needing to hire a team of data scientists, using company data to make new connections, and creating and disseminating highly tailored organizational knowledge in real time. The result is improved customer service at a reduced cost—something that these companies would previously not have had the resources, capabilities or infrastructure to do.
Mid-sized companies backed by private equity firms have additional operational strengths—strategic alignment, financial and human capital, and focused implementation—that make them prime candidates for GenAI adoption. PE firms’ clear objectives and timelines for their portfolio companies, focusing on value creation within specific investment horizons (usually five years), enable decisive action to prioritize and implement GenAI applications. Companies backed by PE can also access the necessary financial and human capital for GenAI projects, giving these companies the capacity to invest heavily in leadership and advisory teams in anticipation of growth. As a result, they are often more willing to take calculated risks based on potential for high returns.
Mid-sized companies may now have some structural advantages for GenAI adoption compared to larger players, but that doesn’t guarantee success. Here are five strategic steps they can take right now to increase their chances of successful GenAI adoption on the road to value creation.
-
Build a scalable and flexible GenAI stack: Invest in scalable AI-as-a-service platforms that can grow with the company without significant additional investment.
-
Move to ‘reshape’ and ‘invent’: Move beyond deploying GenAI for incremental improvements to current processes, and rethink your business model and how you can reengineer entire functions. A recent BCG survey found that the companies at the forefront of AI adoption derive nearly two-thirds (62%) of the value they get deploying AI and GenAI in core business functions, with the remaining third (38%) coming from more peripheral support functions. The takeaway is clear: Go for deep applications that reengineer core functions and prioritize those that leverage unique, proprietary data to create a moat.
-
Look at what complements GenAI, not just the technology: As a recent Evolution Ltd white paper suggests, a key reason for disappointment with GenAI is an overemphasis on the technology itself with too little attention paid to what lies upstream—data engineering and proprietary data—and downstream—integrating GenAI into strategic decision-making and creating learning and experimentation loops.
-
Establish clear governance and leadership: Success with GenAI requires a strong commitment from a company’s leadership to implement governance structures that facilitate efficient decision-making and prioritize investment for the mid-term, not just immediate returns.
-
Enhance workforce capabilities: Use GenAI to augment employee skills, enabling them to perform tasks beyond their current capabilities.
Mid-scale companies, once considered too small, may be “just right” to make the most out of today’s GenAI. To do so, however, they need a clear strategy and a tight focus on where GenAI can make a difference—not just reducing costs, but generating revenue and value. Those that are able to stay laser-focused on effective implementation will find the AI revolution is not just for the industry incumbents or nimble startups—it can be an inclusive wave that mid-sized companies are ideally suited to ride.
Read other Fortune columns by François Candelon.
François Candelon is a partner at private equity firm Seven2 and the former global director of the BCG Henderson Institute.
Michael G. Jacobides is the Sir Donald Gordon Professor of Entrepreneurship and Innovation at London Business School, academic advisor at the BCG Henderson Institute, and the lead advisor of Evolution Ltd.
Meenal Pore is a principal at the Boston Consulting Group and an ambassador at the BCG Henderson Institute.
Leonid Zhukov is the director of the BCG Global A.I. Institute and vice president of AI & Data Science at BCG.X.
Some of the companies mentioned in this column are past or present clients of the authors’ employers.
This story was originally featured on Fortune.com
