Figma
A glimpse into GenAI within the teams of tomorrow
Insights
Exploring GenAI’s impact on team dynamics when in real-time brainstorms
The Challenge
Figma, the leading collaborative design platform, wanted to understand how collaboration and team dynamics change when generative AI becomes a member of the team. We conducted an experiment to take a look.
Generative AI has begun to work its way into people’s everyday workflows, and we’re starting to see how it can augment individual tasks and productivity. But in practice, much of our work is collaborative. When AI “enters the chat,” it has the potential to impact not only productivity and outcomes, but also the synergy between teammates.
The Outcome
We found that generative AI has a big presence in the collaborative room, bringing in energy that shakes up team dynamics considerably, especially for teams with less generative AI maturity. Specifically in our research, we outlined four potential impacts and made recommendations to help teams incorporate generative AI into their workflows for best results.
We authored an insights report of the findings Figma published publicly, as well as joined their team in a livestream to the Figma community to share highlights from the research and participate in a panel discussion with a Q&A.
From the research
“The speed with which [GenAI] can ideate or create something tangible with just a few keywords is really powerful. I feel like if someone is able to harness that power, they can go really far. But if they don’t know how to harness it, the horse is just gonna kick them off and feel overwhelming.”
Observing teams to learn how AI impacts collaboration.
We recruited real teams and observed each as they completed two tasked-based challenges. The first challenge each team completed as they normally would. In the second, teams were instructed to involve generative AI in any way the team saw fit.
With 1-on-1 breakouts in between challenges and follow up in-depth interviews with select participants to reflect on their experience, we were able to get a combination of quantitative observational data and qualitative data to bring into analysis. Additional interviews with subject matter experts helped us situationalize the data in broader trends, contexts, and potential futures.
Our process
Recruit real teams
We conducted a series of challenges with existing teams that worked on products, including design teams at mid- to large-sized companies and design students at the University of Washington. Each team member already had some familiarity with using AI, as well as using FigJam, Figma’s white boarding tool, where we conducted the challenges.
Observe task-based challenges
We tasked teams with solving two ideation prompts. First, we asked them to complete the challenge according to their typical ideation process. Then, we asked them to use generative AI at some point, however they saw fit. Prompts encouraged the teams to ideate around generating ideas such as “envision a new conference experience in the era of remote working” or “imagine a way to help people better manage their digital clutter.”
Follow up in-depth interviews
We matched the number of facilitators in each challenge with the number of participants. This enabled us to pull each participant into a breakout room 1-on-1 with a facilitator to get quick real-time qualitative reflections on the challenge. After the team session, we followed up with select participants for in-depth interviews to hear more about the experience of working with AI on the team.
Talk with experts for broader context
Alongside the observational research with teams, we also spoke with five subject matter experts in industry and academia. Experts were able to share perspectives on AI related to their work and research in foresight, machine learning, designing with AI, product design, and digital humanities. These interviews helped the working team contextualize findings, insights, and recommendations and ensure good storytelling in the final report.
GenAI had a big presence in collaboration.
Groups spoke an average of 40% more, and made 22% more artifacts in their FigJam boards in the challenge using AI.
While genAI had the potential to jumpstart brainstorms, its outsized energy risked drowning out quieter voices on the team. People found the energy exciting and inspiring, yet chaotic and overbearing at times. Below are a few of the risks and opportunities from the research.
Risk
Moving faster and falling prey to shortcuts.
Teams shared that the technology quickly ushered them into generating solutions, often at the expense spending time considering and defining the true problem that needed to be solved beforehand.
Risk
Going in different directions and falling out of sync.
Teams were able to generate more artifacts overall and carry ideas further as individuals. A higher quantity of ideas made coming together as a team and aligning on one cohesive solution a challenge, putting team cohesion in jeopardy.
Opportunity
Onboarding AI to the team like a new team member.
One team had developed norms around when individuals were to use generative AI, and those use-cases rolled up to their teams’ strengths and values. Setting expectations for what and when to use AI helped the team divert many of the pitfalls and chaos other teams encountered in the challenges. Doing so enable to them to use generative AI to prop them up when they encountered a gap or obstacle no one on the team could fill.
Opportunity
Looking at AI’s contributions like a teammate’s: unfinished.
Most participants perceptions of AI fell into two buckets: bossy and definitive, or unfinished and suggestive. While the former found themselves frustrated with lower quality ideas, the latter were inspired by what they saw as initial kernel or spark—something to get them thinking, like a teammate’s first idea might.
Get all of the insights in the full research report
Grab your copy the takeaways and insights from our research report in collaboration with Figma. The report was originally published in Figma’s “Reports and Insights”. The Artefact team presented key insights from the report on Figma’s “Inside the Minds” livestream.
Next project
PRISM