As artificial intelligence immerses itself in the daily workflow of almost any organization, many business leaders are confronted with restructuring departments, units, teams, job functions, and tasks. They face the conundrum of a simultaneous need to redesign their organization to move fast and rapidly innovate while building stable organizational foundations for sustainable business growth. They know their businesses must rapidly build upon AI advancements to improve products and services because moving too slowly can result in missed opportunities for competitive advantage and market relevance; those that lag in adopting AI may find themselves outpaced by more agile competitors, losing market share and facing obsolescence. At the same time, they understand that forming stable organizational structures is necessary for assessing probable risks in deploying these developments. Hastily deploying AI solutions without thorough risk assessments may result in unintended biases, privacy breaches, or inaccuracies in AI-generated outputs, potentially damaging customer trust, exposing the organization to legal and regulatory liabilities, and tarnishing its reputation.
The Speed-Stability Continuum
This tension can be considered a continuum between speed and stability, two conflicting principles that manifest differently in how people collaborate in organizations, departments, and teams. For example, small teams foster rapid innovation because they operate autonomously, allowing for bold experimentation and rapid responses to emerging challenges. Within these smaller units, a culture of continuous learning thrives as team members readily share knowledge and adapt to new information. This nimble approach empowers them to swiftly address evolving customer needs and seize opportunities in the ever-shifting market landscape. Larger teams, on the other hand, provide a robust foundation for organizational stability and sustainability, much like a diverse ecosystem fosters resilience. Their strength lies in specialized expertise, allowing for a division of labor where individuals can focus on specific areas and contribute their unique knowledge. This enhances efficiency and problem-solving and creates a safety net for handling increased workloads and navigating unexpected challenges.
How can leaders think about redesigning whole business units, departments, or teams, and what principles apply when determining who can execute the business strategy, which is the optimal span of control, and when to prioritize speed or stability in operations? Understanding how teams are structured today and learning how two companies have successfully utilized speed and stability to address a specific organizational need can help make such a decision.
Team Size and the Nature of Work
Recent research has shed light on the nuanced relationship between team size and innovation. A comprehensive study analyzing a vast dataset of over 65 million papers, patents, and software products uncovered a compelling trend: as team size increased, the disruptive potential of innovation markedly decreased.1 The first thing leaders should know is if and how their current spans of control – otherwise expressed as team size – may either promote or hinder their efforts. Detailed knowledge of team size and its effect on performance, managerial load, collaborative efforts, and innovation potential is an elusive concept for many organizations. While many estimates about the ideal span of control exist, only companies with mature people analytics capabilities thoroughly understand their own team structure and its consequences.
Our recent study of 144 enterprise-sized organizations in the USA, Canada, and Europe, including over four million employee records in the Visier Community Data, revealed a team size range of six to ten members. (see Figure 1) 2
Strikingly, despite the disruptions wrought by the pandemic, the average team size has remained consistent throughout the observed period. Either organizations have not adapted to the virtual work environment, or these changes are yet to come. The fact is that these numbers have remained stable for years. We did find, though, that the nature of tasks in various industries is influencing these team sizes, so we detected significantly larger teams in the Healthcare, Retail, Wholesale, Transportation, Warehousing, and Manufacturing sectors. In contrast, workers in Finance and High-Tech industries are typically part of smaller teams. Looking at the job domain, we found that a significant variability is influenced by the nature of the work these teams engage in. Different domains, such as healthcare, support, and sales, typically operate with larger teams of eleven or more members, whereas domains like finance and IT tend to have smaller teams on average (see Figure 2). Research indicates that when complex tasks are to be performed in healthcare settings, for example, there are benefits to increased team sizes that can foster greater collaboration.3
While these observations across our large database of real-time employee records are not yet helping us with the innovation conundrum, they suggest that team sizes are tailored to business needs, the scope of responsibilities, and the types of problems addressed within each domain. For instance, the size of facilities teams corresponds to the workload of maintaining office buildings, while technology teams are sized according to task. When tasks are routine or predictable, wider spans of control can be maintained, allowing managers to oversee larger teams effectively. Moreover, larger team sizes are essential for efficient management in areas like healthcare, where shift work is standard. Larger teams tend to perform complex tasks requiring diverse expertise and perspectives more effectively.4
Building on the principle that team size design should be driven by the end goal – the tasks to be executed – we next investigate how the speed-stability continuum can help solve the dichotomy of rapidly innovating while also leveraging stable structures in assessing future risks and challenges. We propose to consider these dynamics within what we call a Speed-Stability Continuum (see Figure 3), which has the potential to solve the dichotomy of rapidly innovating while also leveraging stable structures in assessing future risks and challenges.
A case study from a recent Organizational Network Analysis (ONA) conveys how the speed-stability continuum has been applied by team design based on the nature of the respective task.
Speed Matters in the Age of AI
The Advanced AI Development Center is an innovation hub with 137 employees focused on AI and machine learning. It comprises 13 specialized teams of 7-14 members each, emphasizing small, agile teams for rapid innovation. The center’s overarching mission is to provide a comprehensive suite of services, tools, and resources that enable businesses to harness the full potential of AI for their specific needs; with a keen focus on rapid development and scalability, the center endeavors to unlock value from data for enterprises, startups, and nonprofits across every line of business, all while ensuring end-to-end security, privacy, and AI governance.
These teams operate autonomously, fostering creativity and quick iteration. The center’s ability to prioritize agility and innovation allows teams to respond swiftly to market trends and customer needs. Research has shown that large teams tend to develop and further existing ideas and designs, while small teams disrupt new ways of thinking with new ideas, inventions, and opportunities.5 At the heart of the center’s success lies its strategic emphasis on small, agile teams as catalysts for innovation. Within the organization’s structure, 13 specialized teams (see Figure 4), ranging in size from 7 to 14 members, collaborate seamlessly to pioneer groundbreaking AI solutions. Smaller teams, typically 7 to 9 individuals, serve as the center’s innovation engines. These teams operate autonomously, free from bureaucratic constraints or rigid hierarchies, fostering a culture of creativity, experimentation, and rapid iteration.
At the same time, the organization also manages to create a balance between speed and stability. The below network diagram from our organizational network analysis (see Figure 5) depicts the company’s approach to managing the speed-stability continuum. While small teams are pivotal in driving innovation (small teams on the network’s periphery), the center recognizes the importance of integration and scaling for sustainable growth. Larger teams (shown in blue and gray), situated at the organization’s core, are tasked with integrating and advancing disruptive solutions on a broader scale. These teams collaborate closely with smaller innovation teams to ensure that breakthrough technologies are seamlessly incorporated into the broader organizations’ (not shown here) operating structures. For example, a larger team (dark blue teams) specializing in AI governance and security collaborates with smaller innovation teams to develop robust frameworks for ensuring the ethical use of AI technologies. By integrating these frameworks into the center’s offerings, the larger team helps to instill trust and confidence in AI solutions among clients and stakeholders.
Practical implications: optimizing collaboration for speed and stability
The size of teams plays a uniquely important role in the ability of executive leaders and organizations to innovate against the backdrop of a rapidly evolving AI technology landscape. Our data-science-backed deep dive into team sizes across job domains revealed the status quo of spans and layers in a large data set of real-time employee records. They showed that task design and the nature of work inform the optimal team size in many organizations. Furthermore, the provided case study demonstrates how balancing operational stability and speed is both an art and a science as executive leaders and managers assemble teams for innovation. This increasingly important management priority of determining cross-organizational team composition and size is compelling leaders to explore the use of actions and tools to bring a more disciplined and bespoke approach to assembling a small or large team to achieve and maintain competitive advantage for product innovation.
Actions for Leadership
In answering these questions, executive leaders should then consider the following when determining the right balance of speed vs stability:
- Consider that Small Teams are the Engine of Innovation
Organizations should prioritize forming small, cohesive teams to foster innovation and agility in the face of rapid technological advancements and market shifts.
- Weigh Speed against Stability
Organizations should leverage small teams for initial innovation and larger teams for stability and scalability as AI initiatives mature and expand.
- Remember the Disruptive Power of Small Teams
Organizations should empower small teams to drive innovation by fostering a culture of experimentation and risk-taking and providing them autonomy and freedom from bureaucratic constraints.
- Remove Dependencies and Distractions
Organizations should maximize the potential of small teams by providing autonomy, resources, and a clear mandate while fostering a culture of psychological safety and open communication.
- Realize the Operating Power of Larger Teams
Organizations should utilize larger teams to provide the necessary infrastructure, resources, and expertise for scaling and sustaining AI initiatives over the long term.
In navigating the complex landscape of AI adoption, organizations should look to their small, specialized teams as catalysts for innovation and leverage the operating power of larger teams as initiatives mature. However, as shown above, these teams mustn’t operate in isolation. Engaging IT and Legal departments early and often ensures that AI initiatives comply with technical standards and regulatory requirements, effectively mitigating potential risks. This collaborative approach strikes a balance between innovation and prudence, allowing organizations to move forward with confidence.
ENDNOTES
1 Wu, L., Wang, D. and Evans, J.A. Large teams develop, and small teams disrupt science and technology. Nature 566, 378–382 (2019). https://doi.org/10.1038/s41586-019-0941-9
2 Visier Insights Report: Designing Impactful Teams: Data-backed insights about effective team size. Visier Solutions Inc; 2024, https://www.visier.com/lp/ideal-team-size-organizational-design/
3 Mao, A., Mason, W., Suri, S., and Watts, D.J., “An Experimental Study of Team Size and Performance on a Complex Task,” PLoS ONE 11(4): e0153048. Doi: 10.1371/journal.pone.0153048 (2016)
4 Andrew Day and Dev Mookherjee, Does Team Size Matter? (Metalogue, March 19, 2024)
5 Dashun Wang and James A. Evans, “Research: When Small Teams Are Better Than Big Ones,” Harvard Business Review, February 19, 2021.
