Transcription
Swift trust can help your communication and your teams be more effective. My name is Matt Abrahams and I teach strategic communication at Stanford Graduate School of Business. Welcome to Think Fast Talks Smart, the podcast.
[music]
Today I look forward to speaking with Melissa Valentine. Melissa is an associate professor in the management science and engineering department at Stanford and a senior fellow at the Stanford Institute for Human-Centered Artificial Intelligence. Her research focuses on how emerging technologies, including artificial intelligence and algorithms, are fundamentally transforming work, organizational design, and team dynamics. Melissa, along with her co-author Michael Bernstein, wrote the fascinating book *Flash Teams: Leading the Future of AI-Enhanced On-Demand Work*. Well, welcome Melissa. I'm really excited to have you join us.
Thank you. This show is so fun, and I love the guests you have, so I'm delighted to be here.
Well, thank you, and thanks for being one of them. Should we get started?
Let's do it.
Excellent. I have to say, I really enjoyed your book, *Flash Teams*, very much. Uh, really, really interesting. Uh, can you define for all of us what a flash team is and explain what people, processes, and infrastructure enable flash teams to get aligned and execute quickly?
So, my PhD is in organization science, basically. So I got my PhD at a business school, and then my collaborator is a computer science professor, and he was doing a lot of work on crowdsourcing.
So, what's cool about crowdsourcing is, um, like the crowd is like millions of people online, right? And you can get things done quickly because you just like put a task out, and then there's millions of people who are available to come work on the task. But what he was seeing at the time that he and I started working together was crowdsourcing was sort of stuck because they were doing like really, they call them like microtasks, but just really like small tasks. And then I had this toolkit, organization science. Um, and my dissertation had been about how you can use basically like team scaffolds or like sort of lightweight team structures. Um, and then like you, you populate the team structure with experts, and the team structure and the role structure helps them know how to work together.
So, we sort of were able to combine the logic of both of these. So you have these like lightweight structures that allow people who are like relative strangers to come together and to work together on really complex stuff. The system that we built with like a great team of like a PhD student, Danielle Rattellni, she organized this great lab at the time. We first did our flash teams, so we built a platform that basically like took all the logic of organization science and team science and then the speed and scale of crowdsourcing and then like put them together. So the way the system works, we have this platform where like I'm somebody who like wants to get something done, so I go to the platform, I like design the team, and then the platform integrates with something like we used Upwork. So there's like 10 million freelancers who are on Upwork. And then so I'll say, "Here's the task that I need done. Here's the role I need filled." And then the platform can integrate with Upwork, and then it's an open call to like the 10 million freelancers, "Who's got the right expertise?" They join the team. So you can just convene teams really fast. And what we showed with this research is that, um, like you can get things done really quickly. You can pull teams together, they build really complex stuff really quickly, and it was, it was really inspiring. I mean, it's even faster now with GenAI.
So, you set a framework and expectations and then leverage a tool that pulls people in, and then they can, they then the flash team is really the, the focused, concerted effort that it's all about.
Can you talk to us a little bit about the structures that you can set up for teaming? It sounds like you've done some research in that. What, what are, give me an example of what a structure for efficient teaming looks like?
There's like three structures that are very useful for a flash team. The first is the role structure. So that's going to be our relative expertise. So, let's say that I am a UI designer and you're a front-end developer. Like, we [snorts] get together and we know I know what I'm responsible for. I know the skills I need. I know the tools, the platforms, and you know that too. And since we're both so expert at what we do, we can kind of negotiate it fast.
Right?
I can say, "How is this screen going to work for you?" Or you can say, "Like, where, where is this functionality going to come from?" And we just like negotiate it quickly.
Because we know each other's role.
Exactly.
Yes. And because we know people of that role, somebody else could plug in if you get sick or move on. Excellent. I really like that.
Yeah. So the role structure. Um, so the second one is the team structure, and the team structures are like what help a group know that they are accountable to each other. And so, for example, one team structure could be like everyone being responsible together for something. Like, maybe there's a team rating instead of individual ratings.
I see. So that's what you're going to be looking for, individual ratings, then everyone is going to be like, "Every person for themselves." If it's a team rating, like everyone pulls together, and it's like, "We're all in this together."
So, it's how we team and how we assess the teaming. Very powerful.
Yeah.
Students don't like group projects, right? Because like...
Yes.
...being all in it together can either be the best thing or the worst thing. So...
Absolutely. Yeah.
So, the team structure. And then the last one, actually, is hierarchy.
Um, and hierarchy as a structure is about, um, who has the right to make decisions, right?
Uh, in some ways, again, people don't necessarily think they like hierarchy, but hierarchy is so useful for coordination because you have, you know, five strangers show up together. Maybe everyone has a different idea of what should happen, but if you have the sense, it's like another role structure, if you have a sense of who's the final decision-maker.
Right.
A good enactment of hierarchy is they take in all the information, they understand like what needs to happen, and they just communicate it clearly, right? So, like, the coordination of the group happens through hierarchy.
That's fantastic. So, so three primary roles that, that you just identified are role, that is how we relate to each other and our expertise; the teaming, how we assess and work together and reward, get rewarded; and then hierarchy, or decision-making, which really determines how we can actually move forward.
That's right.
And I think all of us, regardless of if you're putting together a flash team or not, can think about those structures in the team that you currently have to help optimize it, or if you're forming a new team, to think about that. Thank you. That's great.
You know, one of the essential elements of any teaming or groups coming together is trust. Uh, what are some of the communication skills or other skills that people can deploy to establish trust quickly? I'm assuming part of what flash teams do, besides accomplish their work, is they're able to coordinate action, perhaps through building trust.
Yeah, trust in flash teams is essential and complex, right? Because it is relative strangers. You get together and you work together, really intensely, and then you disassemble. [snorts]
Um, so I think that there's this idea in the literature called like swift trust. Okay. And there's like an assumption that you just assume trust. Like you just meet somebody for the first time, you assume they're competent, you assume that they know how to do their role, you assume you're both there for the same reason, to like work hard, get it done. Um, so that, yeah, the assumption of swift trust is really useful. Um, it doesn't always work out. So I think it's, yeah, so but the, the sort of like offering of swift trust is part of it. Um, and then, um, yeah, paying attention to when it's not going right and like being willing and able to kind of repair quickly is also important.
I love this idea of swift trust. So, so you approach from a place of trust, assuming that I, you are trustworthy, and I will trust you, and you will do the same.
Exactly.
Uh, can, can expedite things because there's a lot of this testing that goes into trust and time. And if you start from that place, and obviously, uh, you know, it might fail, or people could take advantage of it. But the second part of what you said, I think, really is the helpful part, is to have that meta-awareness and watching for it, and then being willing to speak up and try to repair it.
Yeah. So [laughter] I think it's not only swift trust, being open to trust, but being open to repair when it's not there.
Yeah, exactly.
I really like that idea of of swift trust. Um, so much in my life, I think, would be better if, if we just approached everything with this notion.
Yeah. I mean, there is a real difference, like, cuz I think of like my, the person I wrote the book with, Michael, he and I have like worked together for 15 years now. And that is a different kind of trust than like swift trust. But swift trust is a tool that's useful, right? It's and it's a way to at least initiate, and then you can build deeper trust and relationship that can only happen, I think, after repeated exposure and time.
Totally. Yeah.
You talk about discursive diversity, and I had to practice saying that multiple times, by the way. Can you share what you mean by discursive diversity and how can teams recognize when they need to shift their communication style from a more open-ended brainstorming, ideating, to a very highly coordinated approach?
Yeah. So that is, um, from the dissertation of one of my really brilliant PhD students. So her name is Katarina Elix. Um, and she kind of invented that measure. Um, so what she was able to do, she had all of the sort of like Slack transcripts of different flash teams. Um, and she was able to, she did this, um, NLP processing.
NLP is natural language processing.
Yeah. Yeah. Before it was cool. She learned how to do it, and she, um, was able to create this measure where she was looking at, like, the similarity of the language that people were using. So you have all of the Slack interactions, and you can basically compare people talking on the team and how similar are the words that they're using, the sentence structure, right? Are they using the same adverbs, like adjectives, sense of urgency gets encoded in all of that. And what she was able to show is the, the sort of similarity of the language that people are using, or the the difference. So this is where, let's see, right, discursive diversity. Yeah. So, like, the diversity of discursion, I guess, like the diversity of words and languages.
Exactly. Yeah. Yeah. So that sort of would predict the team meeting a milestone in different ways. Um, so there were times when you really want the language to diverge. So this is a moment of, um, brainstorming, ideation, right? You want a lot of creativity. You want a lot of divergence. But then as you get closer to the deadline, you really, for the team to meet the deadline, then the language needs to converge. Like, people need to start sounding more like each other to hit the deadline.
So, one, it's a measure, really. So you can, you can look at, you know, if I'm a manager or somebody looking back at doing a post-mortem after a project is done, and I can look at the communication, be it Slack, Jira, whatever the tool is, um, and I can actually see the language might be a hint as to where we are.
Yeah.
And as you're trying to get closer to the decision, more similar language is a sign. What I'd be curious, and I don't know if there's any research on this, is can you actually use language to influence that? So, can I, can I as a manager encourage more diverse language through ideation, but when I want us to start moving towards a decision, encourage people to start using similar language? I wonder if you, if it's causal in that way.
Right. No, I think that's, I think that's a great research question.
I think that's really interesting. So, I encourage everybody listening to think about the language you're using for the different tasks you have, and start to notice what that looks like, and maybe you can run your own experiment and play with that.
Um, I want to keep on flash teams and then I want to make a have a broader question. But when flash teams hit a snag, for example, like a miscommunication, what is the most effective way to fix that so they don't fracture and deal with too much friction, uh, in those situations?
So, I'm so [clears throat] glad you asked this question. I have, um, an unpublished paper. Okay.
So, we ran three like really, um, complex flash teams one summer, and we had all of the, uh, DMs. So we had all the Slack front stage and then all of the Slack DMs, the backstage. Um, and the thing that was really interesting in analyzing all of, um, that data is a lot of repairs happen backstage.
So something would happen in the public Slack channel, and then a DM would go to the back channel, and they'd be like, "What is this bug?" And they would like kind of sort it out and then come front stage with the solution.
Oh, interesting. So it created a lot of like smooth operations. Like the flash teams, they all, you know, accomplished their goal. There didn't seem to be a ton of conflict. So this pattern of backstage repair was really interesting to discover. Um, I think, I mean, like any human behavior, it's got, you know, different sides to it. So I think it did support smooth operation. Um, but what I saw in the analysis that I have not yet published is,
[clears throat]
um, that the people who were more involved in backstage repair ended up with more influence over time. So these are the workers and the managers who were like, really helping, like make the decisions. So it's something about like access to where the real problems get solved was like creating a lot of influence for people. And I think that, I don't totally know what to do with that. As someone who is interested in flash teams going well, um, I would maybe say you need like a post-mortem on a flash team. What were the problems? Who helped solve them? What do we learn about this together?
It's really, really intriguing that, uh, the people involved in that repair or that dealing with those problems end up being more influential.
Yeah. Because it's like they had like the realer story of what was going on. And sometimes it was like in the backstage repair that decisions were made and then presented as if they were quite done, if that makes sense.
Yeah.
That's really intriguing. I, I'm, I'm putting on my manager hat and thinking, if I were in the midst of any team, be it a flash team or a regular team, how I could be aware that the folks who are involved in and helping fix, repair, dive deep into problems, uh, what that means for them in terms of their currency and how we interact on the public stage. And I just think it's interesting that you, we now have these two channels and ways of communicating that can be measured.
Totally. I know you think about storytelling and and change. Um, how can everyday employees in an organization use storytelling to drive broader systemic change, uh, and build coalitions to help achieve whatever it is they're trying to achieve?
Yeah. So, one of the studies that I did, um, when I first got to Stanford, was actually an academic cancer center, and I saw stories be really influential in this study. So it, it feels a little different than what [clears throat] we've been talking about with like AI and algorithms and so forth, right?
But it still, to me, is a study of organizational design, which is what I'm really interested in. Um, so in this, what happened in this study, so I was studying as this cancer center was trying to do a transformation. They were trying to kind of decrease the coordination burden on patients and their families. Um, and during the study, um, there was a group of patient advocates or activists who really wanted the cancer center to understand how much coordination they were having to do.
So they, the way that they made their case, um, this is why I'm thinking of it, is they, they had, um, an initiative they called "Patient Stories." And so they collected, they went to great effort to collect a lot of patient stories where patients just talked about, like, you know, like a family member had to like quit their job to like become the full-time coordinator for the patient. So they had, um, they came up with, you know, dozens of patient stories. And then, um, in this paper that I have, um, I analyzed like how they basically use the stories to make the case and to kind of spread the, the movement, spread the message across a cancer center. And at the end of the study, they had, um, kind of convinced the cancer center to take on some of the coordination work. So, it ends up being this like, kind of end-to-end story of like patient stories, patient narratives actually changing the organizational design, which is why I'm thinking about it. It was like a very powerful thing that these patient activists did.
Stories can really resonate and motivate people. Was there anything specific to the stories that you think were really important? Was it the fact that the stories accounted for all the specific coordination events, or was it the emotionality, or was it the fact that these people just collected a bunch of stories and, and the cancer center was like, "Wow, there are all these stories, we should probably do something"?
I think a lot of us interface with the healthcare system, and we could just like see ourselves. So the patients' stories were just like, "This is us. This is you and me." So there's a connecting piece there.
Yeah, really connecting piece. Um, and then it illustrated the problem really specifically. So, like, somebody's like, you know, sitting in a wheelchair in a hallway because they sort of been dropped or something like that. And so it's like, "Wait!" [laughter]
Like, I think the, the cancer center is so full of people who care, and then hearing like the specific instance in a really relatable story, they were just like, "Ah, that's a problem that I can help with."
I think you hit on it. It's, it's relatable and real, and we, and we can connect. Thank you. Yeah.
Can you take what you've learned about flash teams and give us any insight that we can just apply to, to a normal team? So, I've got a team that that's been around for a while, and we meet regularly. Is there any insight from your work on flash teams that could perhaps help my team be more efficient, feel more connected, ideate better?
Um, one thing that, um, Michael and I used to say when we were writing the book, we were trying to think of like, what's the mindset here that like any manager would find empowering? Um, and the phrase we came up with is, um, "experts everywhere all the time." [clears throat]
So you've [snorts] got your like trusted team. Great. You know, keep working together, taking good care of each other, working hard. Um, but like, really recognize that your organization, the world, the internet, your friends. Like, the world is just full of experts who can help, and people like helping and bringing their expertise to something. So it's just like having a mindset of like recognizing how much collaboration is available, how much expertise is available, and, um, yeah, so it maybe it like invites you to kind of like think of the, the boundaries of the team a little more porous.
That's that exact, that's funny you said that because that was exactly the word I was thinking, porous. Is that it can be very easy to insulate a team and say, "This is the team." But, but being a little more open and pulling in expertise. Uh, when I was managing a team, uh, we had a, I, I ran a learning and development group, and we had this very tricky thing we had to train people on. And it turned out that somebody in the company who was an admin, who was using our tool, and they actually knew how to do things that we were trying to figure out how to train people on better. And then that person was an expert. And we, and by just dumb luck, we, we figured that out and incorporated them into the team. And, and I like that idea. I think there are a lot of tools that can be very helpful to managers, um, in terms of how the team, like, works together, coordinates, and stuff like that. So, you know, like having an eye towards the other thing we said in the book is like AI-driven team design. So, just making use of the tools to be very thoughtful in how you're structuring your team.
So, AI can help us structure the team and maybe even identify who some of the experts are that we might know. I love that.
Well, Melissa, this has been great. Before we end, you know, I like to ask three questions. One I make up just for you, and two I've been asking everybody. Are you up for that?
Yeah, let's do it.
Beyond the academic research you do, you're also a creator. Uh, and you study the algorithms that serve up information on social media. What advice do you have for other creators?
Okay. So, I love Instagram and TikTok. Okay. I like, I have, I think it's people have like complicated relationships with social media, but I, um, so the attention stuff is real. My attention span's shot, but I have learned so much from content creators. I am such a fan. My advice is, keep going. [laughter]
Like, if you, if you're an expert, put it out in the world. Like, people will find it. And it, I mean, like different, like life transitions I've gone through, different, like health stuff, different fitness stuff, different hobbies. Like, I have learned so much from the internet. So thank you, content creators. [laughter]
You know, as a creator myself, and it took me a while to identify as a creator, but, you know, I think many of us are motivated to help people.
Exactly.
And it's that motivation that actually helps us get through. I mean, it's hard to put out a lot of content, and the algorithm can sometimes help you and sometimes not. So, so it's that passion and that desire to learn and to help, I think, that really motivate. And I can hear that in your voice and and see it in the things that you post and create.
Question number two, who is a communicator that you admire and why?
So, I'm at a moment where I'm thinking a lot about communication where we like really, like, take each other in in like a really deep way. So I have a friend who is, um, a therapist and a Buddhist, and she does, like, she has like a quality of like listening and expression that it, that allows for like true communication. So I would say I'm going to, I'm going to nominate her.
Excellent. And it, it's really that presence and listening and and being there.
It's like a fascinating thing about like being very aware of yourself that allows you to be very aware of others. And just, yeah, like it is a lot of presence, as you're saying, like a lot of kind of presence and connection.
Yeah, I like that self-awareness and other awareness. Thank you.
And final question, what are the first three ingredients that go into a successful communication recipe?
Something that I've come to appreciate is how useful, like, a deep awareness of your own experience is because then you're even more able to take in another's and all of its complexity. So, yeah, self-awareness, other awareness, and then working from there. I think what's possible between you is like new. It's like something that neither of you would have done, right, [clears throat] without that kind of connection.
So, I sort of bristle at the word synergy, but in this case, that's exactly [laughter] what you're talking about. Is it, it's the power of the two is greater than either individually. But that comes, as you said, from a, a self-awareness and then an other awareness.
Well, thank you. Thank you for the, the three ingredients and thank you for the conversation. It, your work to me is really intriguing because it focuses both on technology but the human aspect of people coming together, and the fact that there are structures that we can rely on to help expedite and make things more efficient. Melissa, this has been fantastic. Thank you. And thank you for the work you do.
Totally. Really fun for me too. Thank you.
Thank you for joining us for another episode of Think Fast Talks Smart, the podcast. To learn more about groups, listen in to our episode 241 with Colin Fischer. And to learn more about social media algorithms, please check out episode 268 with Angelie Christine. This episode was produced by Katherine Reed, Ryan Campos, and me, Matt Abrahams. Our music is from Floyd Wonder, with special thanks to Podium Podcast Company. Please find us on YouTube and wherever you get your podcasts. Be sure to subscribe and rate us. Also, follow us on LinkedIn, TikTok, and Instagram, and check out fastermarter.io for deep dive videos, English language learning content, and our newsletter. Please join us for our Think Fast Talks Smart Learning community at fastermarter.io/learning. You'll find video lessons, learning quests, discussion boards, an AI coach, and book club opportunities. Again, that's fastermarter.io/learning to become part of our Think Fast, Talks Smart learning.