Building the Team
Summary
- Design the team on purpose. High-performing innovation teams rarely happen by accident; they come from selection, investment in team building, and clear guidance on roles and tasks (Tidd & Bessant, 2018, p. 194).
- Match structure to how new the work is. The newer the innovation, the more autonomous the team should be (van den Ende, 2021, p. 155), while incremental work suits functional or lightweight teams (van den Ende, 2021, p. 156).
- Keep it small and deliberately diverse. About ten people is the ideal (Benraouane & Harrington, 2021, p. 141), and talented but similar teams performed below mixed, average groups in one noted experiment (Tidd & Bessant, 2018, p. 195). People default to homogeneity, so the person forming the team must build diversity on purpose (Benraouane & Harrington, 2021, p. 141).
- Name the roles, and watch for champions attached to pet ideas. Secure a sponsor with the power to remove obstacles (Tidd & Bessant, 2018, p. 179), and document who does what once the initiative takes off (Benraouane & Harrington, 2021, p. 188). A technical champion may not be prepared to let go of a pet idea (Tidd & Bessant, 2018, p. 179).
- Recruit against explicit criteria. Interesting initiatives attract everyone and uninteresting ones attract no volunteers, so define selection criteria up front (Benraouane & Harrington, 2021, p. 187).
- Free the leader, and choose for expertise. At a minimum, the team leader should be single-threaded, not splitting the week with other projects (MIT Sloan Management Review, 2023, p. 30). No universal list of leader traits exists (Bessant & Tidd, 2015, p. 267), but expertise plus cognitive-processing skills matter (Bessant & Tidd, 2015, p. 268).
- Invest in psychological safety, and measure it. Project Aristotle found it the strongest predictor of team success (Benraouane & Harrington, 2021, p. 142). A ten-item team check, scored 1 to 5 by each member, gives a baseline (Goller & Bessant, 2017, p. 142).
- Set a cadence with visible checkpoints. Innovation boards typically review funded teams every 30, 60, or 90 days (MIT Sloan Management Review, 2023, p. 31), and without a sustaining program the initial surge of ideas tends to ebb (Bessant, 2009, p. 60).
Choosing a structure
For a new product or service, the basic choice is among functional teams, cross-functional project teams, and some form of matrix between the two (Tidd & Bessant, 2018, p. 562). In a matrix, no dedicated team is formed; members stay in their functional groups but are designated to a project group (Tidd & Bessant, 2018, p. 562).
van den Ende sharpens the choice with four structures that differ in their lines of authority (van den Ende, 2021, p. 154):
- Functional: tasks are coordinated by the functional managers of each unit, which is closer to a regular work structure than a true team (van den Ende, 2021, p. 155).
- Lightweight: members from different units are coordinated by a project manager whose authority is limited to planning and reminding, without power to make major changes to the project’s definition (van den Ende, 2021, p. 155).
- Heavyweight: the project manager has far more authority, including a say in members’ yearly evaluations and more responsibility for the outcome (van den Ende, 2021, p. 155).
- Autonomous: the project leader has full say over the unit’s members and is the only person evaluating them (van den Ende, 2021, p. 155).
The fit depends on how new the innovation is.
The newer the team is, the more autonomous the team should be, because higher uncertainty demands flexibility and strong integration of functions (van den Ende, 2021, p. 155). For incremental work whose tasks can be defined up front, functional or lightweight teams are more appropriate (van den Ende, 2021, p. 156). In practice, many firms apply the same structure to every project regardless of newness (van den Ende, 2021, p. 156).
Cross-functional teams carry a trade-off. Functional diversity extends the range of knowledge available and the number of options considered, but it can hurt cohesiveness, project cost, and decision-making efficiency (Tidd & Bessant, 2018, p. 797).
Size and diversity
Most researchers agree the ideal innovation team is about ten people, since larger teams increase conformity pressure and weaken team identity (Benraouane & Harrington, 2021, p. 141). MIT Sloan Management Review is tighter, citing Amazon’s two-pizza rule, a maximum of eight people (MIT Sloan Management Review, 2023, p. 30). In my own experience, a team size of five to seven people balances effectiveness and speed.
On diversity, the guidance is to mix disciplines, professional backgrounds, and cultural and geographic backgrounds, with cognitive diversity supporting divergent thinking (Benraouane & Harrington, 2021, p. 141). Diversity also helps teams avoid two named pitfalls (Benraouane & Harrington, 2021, p. 141). Concisely:
- Groupthink is when a cohesive group converges on consensus to preserve harmony, and dissenting views get suppressed.
- The Abilene paradox is when a group pursues a course of action that few or none of its members want, because each assumes the others do and nobody says otherwise.
The most effective teams show diversity in background, ability, and behavioral style; in one noted experiment, talented but similar “Apollo” teams performed below mixed, average groups (Tidd & Bessant, 2018, p. 195). There is a limit, since too much heterogeneity can create tensions (Bessant & Tidd, 2015, p. 270). Diversity also does not appear on its own: people default to homogeneity, so the person forming the team has to engineer it deliberately (Benraouane & Harrington, 2021, p. 141).
Making diversity known to the team
Without psychological safety, teams tend to focus on knowledge everyone already shares and fail to use the diversity of knowledge in the room, a pattern called the common knowledge effect (Goller & Bessant, 2017, p. 139). Two practices help surface differences. Leaders can hold face-to-face meetings early in a project to align professional values and make differences explicit; Edmondson’s example, as reported by van den Ende, contrasts builders, who prize getting it right the first time, with software engineers, who are used to experimenting (van den Ende, 2021, p. 173).
Make differences known to build mutual understanding of each other.
A “relationship map” uses get-to-know-you questions, such as strengths, best and worst days at work, and what creates pressure, to build mutual understanding (Goller & Bessant, 2017, p. 144). My suggestion is to end the kickoff with a one-page team map recording each member’s background, strengths, and working style, and to keep it visible.
Team roles: an aggregation across authors
Different authors describe different kinds of roles, so what follows is my synthesis across several of them.
Organizational roles (Tidd & Bessant):
- Technical champion: the source of critical technical knowledge, who also supplies inspiration and motivation when problems seem insoluble (Tidd & Bessant, 2018, p. 179).
- Organizational sponsor: a person with power and influence who can remove obstacles; sponsors need not know the technology in detail but must believe in its potential (Tidd & Bessant, 2018, p. 179).
- Business innovator: someone who brings the broader market or user perspective (Tidd & Bessant, 2018, p. 179).
Agile role (van den Ende):
- Product owner: oversees an iterative process, is responsible for the outcome, and is not a member of the team (van den Ende, 2021, p. 155).
External-relations functions (van den Ende, summarizing Ancona and colleagues):
- Ambassador: connects the project to the company’s power structure (van den Ende, 2021, p. 156).
- Scout: monitors markets, technologies, and competitors (van den Ende, 2021, p. 157).
- Task coordinator: aligns tasks and schedules with other teams (van den Ende, 2021, p. 157).
Behavioral team roles (Belbin, as summarized by Tidd & Bessant): Belbin proposes nine roles describing how people prefer to contribute, and argues most people are comfortable in only two or three (Tidd & Bessant, 2018, p. 195). Each pairs a strength with a weakness; for example, the Plant is a creative problem solver who can lack detail (Tidd & Bessant, 2018, p. 195).
So one person may be a technical champion (organizational role), a Plant (behavioral role), and a scout (external function) at once. Once an initiative takes off, documenting who does what reduces confusion and motivates people (Benraouane & Harrington, 2021, p. 188).
A caution belongs with the champion role. A technical champion may not always be prepared to let go of a pet idea (Tidd & Bessant, 2018, p. 179). This is close to the “not invented here” phenomenon, a bias against ideas that did not originate inside one’s own group. However, the pet-idea problem is its mirror image: overvaluing what is invented internally. My suggested remedy is to keep outside voices connected to the team and make weighing evidence, not defending an idea, part of routine.
Recruitment: required versus volunteer
The recruitment question is a tension between motivation and coverage. Volunteers bring energy; assigned members can fill functional gaps.
If an initiative is interesting, everyone wants on the team, and if it is not, no one volunteers, so selection criteria are needed (Benraouane & Harrington, 2021, p. 187). Suggested criteria include technical skills, knowledge of the organization, and ability to perform in the new team (Benraouane & Harrington, 2021, p. 187). The same source adds criteria that identify intrinsic and extrinsic motivation (Benraouane & Harrington, 2021, p. 188). One large innovative organization, 3M, lists recruiting volunteers among the elements of its innovation culture, alongside involving customers and outsiders in development teams (Tidd & Bessant, 2018, p. 224).
A smaller example shows how volunteering and structure work together. A manufacturer of measurement equipment began high-involvement innovation at shop-floor level with three volunteer teams. Each received training that included storyboarding, and each presented twice, about three months in and at project close after 10 to 12 months, to the other teams, the manufacturing director who introduced the approach, and several managers (Bessant, 2003, p. 118). I include it because it links two choices in this framework: volunteers bring motivation, and scheduled, visible presentations bring accountability, which is where cadence comes in (see below).
One membership rule deserves emphasis. The best innovation teams have members dedicated full time, and at a minimum the team leader should be single-threaded, meaning not splitting the week with other projects (MIT Sloan Management Review, 2023, p. 30).
The team lead, and how that differs from the facilitator
Selecting the team lead is the most consequential choice in this phase, and it helps to separate two roles that are easy to blur. The team lead owns the team’s direction and creative climate. The innovation facilitator, who may also act as project manager, runs the process around the teams. In a small program they can be one person, but they are different jobs.
Start with the lead. An initiative needs a clearly defined leader who can secure tangible and intangible resources and who has the profile the initiative requires (Benraouane & Harrington, 2021, p. 187). The research warns against checklists: there is no brief, universal list of enduring traits that all good leaders must have (Bessant & Tidd, 2015, p. 267). What recurs is a combination of expertise and cognitive-processing skills, and a key part of creative leadership is providing feedback and evaluation rather than simply generating ideas (Bessant & Tidd, 2015, p. 268). The quality of leader-member exchange also influences subordinates’ creativity (Bessant & Tidd, 2015, p. 268).
Style matters too. Mainemelis and colleagues distinguish four leadership styles for creativity (van den Ende, 2021, p. 172):
- Directing: the leader is the creative one and expects the team to act on their ideas.
- Facilitating: the leader helps others become creative without trying to be especially creative personally.
- Integrative: the leader challenges both themself and the team to be creative.
- Non-leadership: the leader expects neither themself nor the team to be very creative, and so is not really leading creativity.
The right style depends on the leader’s personality and the team’s expertise in creative tasks; with low expertise, a leader may have to choose the directing style to show the way (van den Ende, 2021, p. 173).
Now the facilitator or project manager, where the guidance is about fit. Project managers range from “heavyweight,” deeply involved with the organizational power to make things come together, to “lightweight,” whose involvement is more distant, and the advice is to match the type to the situation rather than using a sledgehammer for a simple task (Tidd & Bessant, 2018, p. 179). My suggestion is that a facilitator who owns cadence and handoffs can usually be lightweight, provided the lead and sponsor hold the decision rights.
As innovation capabilities scale across more teams, the facilitator becomes the program and portfolio manager. However, scaled further, a couple programs may be necessary: a dedicated program and portfolio manager, and a separate train-the-trainer program that teaches project managers to be facilitators of the innovation framework.
Creating the environment for innovation
Google’s Project Aristotle is the most cited evidence here. It studied 180 teams and more than 250 attributes, and found psychological safety the strongest predictor of team success (Benraouane & Harrington, 2021, p. 142). Two further norms marked high-performing teams: equal conversational turn-taking and high social sensitivity (Psych Safety, n.d.). Colocation, extroversion, and seniority were not significantly connected with effectiveness, and neither was team size, though that page cautions that the range of sizes studied may not have been wide enough to show the limits of large teams (Psych Safety, n.d.). That caveat is a reason to treat the ten-person maximum guidance above as a default, not a proven threshold.
Edmondson defines team psychological safety as a shared belief that the team is safe for interpersonal risk taking (Goller & Bessant, 2017, p. 134). It enables learning behaviors such as asking for help, admitting an error, and seeking feedback (Goller & Bessant, 2017, p. 136).
Leaders can build it. Edmondson’s three tips are to frame the work as a learning problem, model curiosity by asking many questions, and acknowledge your own fallibility (Goller & Bessant, 2017, p. 143). Leaders can also stress that risk-taking and failures are accepted and use socializing activities at the start of a project (van den Ende, 2021, p. 173). Two simple team rules are equal air time for everyone and short feedback sessions after every meeting (Goller & Bessant, 2017, p. 142).
To measure the climate, Goller and Bessant offer a ten-item team check. Each member rates ten statements from 1 to 5, where 5 means the statement describes the team exactly and 3 means neither agreeing nor disagreeing, and writes at least one example for each score (Goller & Bessant, 2017, p. 142). Run at kickoff and again later, it gives the lead and facilitator a concrete baseline.
- It is safe to take a risk on this team
- It is easy to ask other team members for help or to get them to accept it from other team members
- Members of this team are able to bring up problems and tough issues without raising objections
- Mistakes made by team members are not held against them
- No one on this team would deliberately act in a way that undermines the efforts of another team member
- Team members accept differences and value distinct talents of team members
- Team members are eager to share information about what does and does not work
- There are good relationships between the team members
- On this team, it is easy to speak up
- There’s a high level of trust in this group
Cadence: quarterly, annual, or ad hoc
At the team level, agile methods use a rapid cadence of short sprints (MIT Sloan Management Review, 2023, p. 30). At the program level, boards that fund long-range innovation teams typically review progress every 30, 60, or 90 days before releasing the next tranche of resources (MIT Sloan Management Review, 2023, p. 31). By contrast, single-sponsor approval is described as inherently ad hoc and hard to end once a sponsor’s reputation is attached (MIT Sloan Management Review, 2023, p. 30).
There is also a warning about momentum. Many organizations start an innovation effort, enjoy an initial surge of ideas, and then watch it ebb without a strategic program to sustain it (Bessant, 2009, p. 60).
My recommendation is a layered rhythm: form cohorts on a fixed annual or semiannual schedule so recruitment and onboarding are predictable, hold quarterly checkpoints aligned with 90-day funding-style reviews, and reserve ad hoc teams for urgent, well-scoped problems. Whatever the cycle, schedule visible checkpoints, as the measurement-equipment teams did, so the effort does not fade.
Summary
Team formation is a design problem with several levers. Match structure to how new the work is. Keep teams small and deliberately diverse, then make that diversity visible so people can use it. Distinguish organizational roles, behavioral roles, and external-relations functions, and watch for champions who cling to their own ideas. Recruit with clear criteria, using volunteers where motivation matters and assigned members where coverage matters. Choose a team lead for expertise and creative style, and treat the facilitator as a separate role. Invest in psychological safety and measure it. Finally, set a cadence of short cycles and visible checkpoints so momentum survives the first burst of enthusiasm. The output is a named structure, a documented team, a lead and a facilitator, a safety baseline, and a calendar, ready for innovation discovery.
Bibliography
Benraouane, S. A., & Harrington, H. J. (2021). Using the ISO 56002 innovation management system: A practical guide for implementation and building a culture of innovation. Routledge.
Bessant, J. (2003). High-involvement innovation: Building and sustaining competitive advantage through continuous change. Wiley.
Bessant, J. (2009). Innovation. DK Publishing.
Bessant, J., & Tidd, J. (2015). Innovation and entrepreneurship (3rd ed.). Wiley.
Goller, I., & Bessant, J. (2017). Creativity for innovation management: Tools and techniques for creative thinking in practice. Routledge.
MIT Sloan Management Review. (2023). Innovation systems: Advancing practices to create new value [Special report]. MIT Sloan Management Review.
Psych Safety. (n.d.). Google’s Project Aristotle. https://psychsafety.com/googles-project-aristotle/
Tidd, J., & Bessant, J. (2018). Managing innovation: Integrating technological, market and organizational change (6th ed.). Wiley.
van den Ende, J. (2021). Innovation management. Red Globe Press.