2021 NYU Stern Innovation Conference: Enabling Distributed Innovation

2021 NYU Stern Innovation Conference: Enabling Distributed Innovation

Show Video

[Music] i'm now going to introduce the moderator of our next panel uh she's a professor here at nyu it's natalia lavina she's the toyota motor corporation term professor professor of information systems technology operations and statistics she's a fascinating incredible intellectual scholar and friend uh so glad that she's here with us her main interest is in understanding how people span organizational professional cultural and other boundaries while producing and using technological innovation so her main research interest could not be more central to the topic of today's conference so we're extremely lucky that she's here and i'm going to turn it over to her now to introduce her other panelists thank you natalia thank you melissa what a great start i truly enjoyed hilah's latest presentation and i haven't heard it before you guys getting absolutely the latest and greatest research so that's great um so i it's my pleasure to introduce our my two co-op panelists our uh the first co-panelist i hope chris if you can turn on your camera is uh christucci uh he's a longtime friend and a fascinating researcher uh chris is uh currently uh in the imperial college uh london he wasn't faculty here at stern many of us still miss him uh he has his phd from mit's loan school uh chris uh is a leading research on crowd sourcing especially from strategy perspective so kind of continuing with some of the themes he'll have brought in um chris is leading the center for digital transformation now um some of you who have taken mba classes and innovations might have been exposed to chris's earlier book on digital model transformation and uh his more recent book on crowdsourcing i was a part of it's a great book as well uh chris is a fun person to talk to always and today he will be covering uh the topic on the strategy level uh of governing if i uh distributed innovation uh uh take it away chris and millis i'll introduce melissa valentine right before her talk so it stays fresh in our minds and uh my talk will be third and we all will hopefully leave some time for questions and thanks melissa for keeping us on schedule so we'll have time for q a chris go ahead okay thank you very much thank you natalia uh great to be here thank you for the invitation let me try sharing my screen see if i can pull this off um i am at home and i just got my internet connected three minutes before i started so i was tethering to my phone for three weeks so let's hope i hope you can see the screen um i'm going to talk a little bit about distributed innovation and crowd governance today and so i will uh explain what i mean in a in a minute or two um uh actually it's a little bit about me here but the one thing i wanted to mention uh thank you for the for the intro um and i didn't tell you about this but i'm working on a new campus concept right now for imperial college so we have a new campus area in white city in london and we're going to have cross faculty cross school within imperial uh collaboration on ai data and digital and so uh we're working on that i'm actually one of the co-directors of it and working on all the education programs for that so it's kind of a very exciting development that i've been doing uh working on for the last year and a half um okay let me jump right into this um to this crowdsourcing um governance type type question so just to delay the the groundwork here i wanted to talk a little bit about open innovation just to say you know um open innovation's got many different aspects to it many facets to it and this is you know uh from henry chesbros book you can see the different you know the the the boundary of the firm here there's lots of things that happen outside the firm and then there's many things that happen inside the firm and you know the whole point of the book was to um make people aware that they could do things outside as well as inside in addition to uh just to reference a little bit of hilar's talk there and i think what i say will be very very um consistent uh i hope is a compliment to what hilar was just discussing so you know you might have corporate new venture groups a spin-off a spin-out um you know taking you license things in etc uh this is this is sort of time passing or this might be the progression of the projects university research corporate venture capital and then over here you've got this big one on distributed innovation and i'm going to zoom in on that a little bit and think about how we could have different um uh how we could have uh different aspects of distributed innovation this is firm sponsored stuff you know obviously and i'm calling this a temporal sequencing because basically i'm thinking of this like an upstream on the left and downstream on the right so you know the you could have and it's mainly outside the firm that's why you see it below this little dashed line here but of course you could always have an internal idea contest a lot of companies have them you know very interesting you could also have a hackathon that spans internal and external maybe something very upstream to come up with some my new ideas sort of a very very compressed crowdsourcing exercise and then you know the one that a lot of people studied including myself studied the most um is you could have a kind of crowd sourcing for problem solving um in addition to that you know as you move further and further toward the market you may actually have something like crowdsourcing for marketing or crowdsourcing for for demand you know this is the kind of thing like a threadless type type argument where people propose designs and vote on them and then they kind of know which designs are the most popular amongst the the crowd and then they basically turn around and sell those to those same designs to the crowd you know so you're you're basically trying to figure out the demand as opposed to coming up with any kind of basic r d or any kind of applied r d um so you know as a slight explanation for this i mean this is a little paper i wrote with alan afua um almost 10 years ago now and you know this is a i'm going to call it a sponsor right now we call it we called it a focal agent in that paper because you know you're on some kind of knowledge landscape let's just say as an analogy and you want to get to the highest peak in the knowledge landscape and so you want to have you know access to these other people b c and d for example and they are going to be well placed to do local searches around where they are and so this one d here could get you a very very good solution exactly where you want to be but you can't get there yourself because you know you're too constrained by your current knowledge by the people you have the training they have the skills they have so it's kind of like a making a lo a local search you know for these guys into a distance search so in other words making a um making a local search for a in some sense okay and i'm going to come back to this picture later because i'm i'm going to question when we don't have a sponsor what do we do so um basically here i'm adding a new a new parallel oval and i'd like to differentiate between sponsor organized and crowd organized innovation and so basically what all we've been talking about so far have been really oriented toward a focal firm you know or sponsoring an idea and trying to get as much input as as possible into their own um into their own process that they then churn out and and take to market and what i'm proposing here uh something i've been working on um uh quite a bit lately uh and to some to various degrees of success but i i'm actually very i'll be very happy to get your input and your questions here uh and see how we could we could think this through a little better and improve it but basically saying that you know it you may not have a sponsor you may not need a sponsor maybe there are some circumstances under which a self-organizing crowd could actually be very very interesting for innovation even if a firm would be to benefit from it maybe you don't need to control the crowd in such a way so that's a and i don't think it's it really doesn't have anything to do with a lot of the other things that we talked about when we talk about distributed innovation it's really uh kind of a different dimension i would say an orthogonal dimension that's how i pictured it like this um so you know what is what is crowd organized means and this is the paper i've been working on with for for a while now uh with my former phd student joanna pereira and my colleague gigi biscuisi um who's at imperial college with me um and we're trying to basically understand what does it mean to have a self-organized crowd you know so so far we've come up with the following list and they're not completely independent of each other as you're going to see in a minute but you know one thing is that norms could be emerging you know so it's not so obvious how interactions occur and how behaviors occur and people who have studied you know offline crowds i mean real world crowds always say that you know if you feel like you're anonymous in a crowd or even if you're not anonymous per se but if you even feel a little bit anonymous because you're in a crowd you might have more so-called deviant behavior which which brings you know which i really enjoyed that discussion that we just had you know saying what is an error here you know and this doesn't have to be a machine generated error it just be more variance you know so so what i think of is more deviant behavior could be more creative behavior more creative ideas um uh on the other hand it might be something that's anti-social or illegal or something like that but i would say probably the variance of the ideas you know would tend to grow if you have a self a self-organized crowd um the scaling could be higher typically if you don't have a gatekeeper now that's not always the case you know clearly uh you could have an open call and try and get as many people in the in the crowd in the in the world to participate um but on the other hand if you know if you have a linkedin group that's by invitation only you know that's going to do brainstorming for you i mean that's not that's not got a very big growth rate that's not a very high scale and so what i'm arguing here is that it might be more scale it might have bigger scale and be more scalable and on the other and you may have more democratic decision making so in other words um the sponsor isn't going to call all the shots on what people are going to work on but rather or even the problems that are posed so the crowd itself could um could turn around and decide what problems are interesting and what kinds of solutions would be would be interesting as well now when we talk about what is an error or what is deviant behavior i think there's a this this is a great example and i'm sure that most of you have heard of boaty mcboatface but if you have it you can search for it online it's a very you know it's it's it's a very funny story but it also and the headline in the new york times is what you get when you let the internet decide so you know the idea was hey let's throw up a challenge to the whole you know to anybody and let's see what they come up with for the name of this prestigious sophisticated new research vessel in the uk so that was the basic challenge and people came up with names and at some point someone said how about bodhi mcvote face you know and then people um latched on to that and started voting for it and and originally they had decided that they were going to literally take the winning like the one with the most votes was going to be that's how they're going to name the boat which is actually a mistake maybe we can talk a little bit more about that it kind of relates to the question by isabella uh from a few minutes ago so in the end they had to solve it so they actually this name did win but they declined to name the whole boat body mcboatface they decided to take one of the sub vessels on their research vessel and call that body mcboatface and they called it something very sober i can't remember what it was now in any case what i'm trying to do here along the bottom is trying to articulate roughly um a kind of a you know very closed i would say or a very controlled or firm sponsored process on the left and moving all the way to the right we have more of a crowd driven process let's just say you know so you know center threadless you know we talked about these all the times um you know sapiens idea contests uh then you've got you know your your current stage of wikipedia which actually is and this is the this little logo here is the actual original logo of wikipedia and um i would say that wikipedia was actually fairly crowd driven at first it was it was a kind of crazy wild west you know of anybody putting anything they want in there and then over time they kind of put controls and made editors and voting and you know and and made much more of a structured process out of it and then you know you've got your things like one billion minds which was a which was a very interesting site for global social challenges and people could propose projects and people would work on the projects in the crowd and then maybe over here you've got something like bitcoin or some kind of cryptocurrency where you the where the constituencies themselves you know decide how they want to migrate the entire system uh if at all so i think there's there could be a very interesting range of options here that that that incorporate less and less firm sponsored control and and and that brings me to this to this landscape now we haven't had very much um traction with this so far um but you know what we were thinking was you know here you have these people on a landscape and you've got your this focal agent the sponsor is deciding you know what the problem is and so they are are they somehow the knowledge landscape is laid out for them because of what they decide they want to work on and these peaks are representing different ways that you can approach the problem let's just say you have people scattered all over the landscape but what if these different people scattered all over the landscape could themselves decide what problems to work on you know and would that make a would that make a more uh rugged uh landscape um i imagine it could would there be more interdependencies probably and so i've been trying to think of different analogies of how a rugged landscape searching on a rugged landscape could change if you could have the crowd itself determine the problems and the solutions so now i just wanted to kind of conclude here with some different implications for outcomes um so uh you know you can see crowds represent different types of degree of organization and so i'm going to go from you know crowd organized over here to sponsor organized and then i'm going to think of an output and outcome of this which would be the knowledge generativity so as i was saying before you know if you have a crowd organized you're probably going to have more dev so-called deviant behavior you might have more so-called errors but actually that might give you some crazy inspiration or some crazy idea that you could then run with at some point so you know what we're arguing here is that your knowledge generativity is probably going to be higher in a crowd organized situation than in a sponsor organized and you'll be have less so i'm not saying it's zero just probably lower amount of generativity from you know very firm controlled um outcomes and this could trace out a kind of a um what we're calling what we what we have called the generativity frontier here you know where you've got you know like along the frontier you're efficient and within here you're inefficient because you could either go this way or this way uh and and improve on one dimension or the other so for this amount of of generativity you know you could you could probably um have a much close much more controlled um environment for your crowd and up here is you we would just say it's impossible because you know you can't go beyond the edge of that frontier so implications here so what are the risks and costs of mysteries so misunderstanding type of collective in other words you know depending on what you want to do and what you want to get out of it and maybe even a sponsor could be interested in seeing what the crowd works on itself that could be interesting so even if they're not controlling the crowd so you know you might over value or underestimate the amount of investment the kinds of output that you think you're going to get from a crowd um uh the crowd characteristics may lead to more or less knowledge generativity and you know the higher the level of of centralization the more control the sponsor exerts over the crowd probably the lower the knowledge generativity of the crowd so um that's all i wanted to say about this i i am happy to hear any feedback or ideas that you have for um improving this it's kind of uh things that in progress for the last couple years we've been thinking about this and it was a very nice opportunity for me to um collect my thoughts uh today so thank you very much uh look forward to the rest of the panel and to the q a thank you very much chris uh it's great talk again brand new stuff so very exciting uh i wanted to ask that our audience can put in questions into the q a but we're going to move on to the next talk and we'll have question time for joint q a at the end with that said it's my great pleasure to introduce our next panelist professor melissa valens valentine from stanford uh i also have a pleasure of knowing professor valentine for many years since she was a phd student she is now a tenured associate professor at stanford university management sciences and engineering department in center for work technology and organizations here research is fascinating and spent spends multiple domains which is rare i got to know her work when she was doing stuff in that organization health care teams uh looking at how to but organize emergency room departments she uh then extended this work to working with oncology clinics but um later working with others at stanford she has pioneered this idea of flash teams which we're going to hear about more today and now uh professor valentine is working on research on the use of artificial intelligence technologies in organizations with diverse experts so i can't wait to hear her talk and please welcome professor valentine thank you so much natalia it's great to be here um i actually want to start with a selfie and a story uh this selfie is the day that i met dana lewis as the caption reads dana lewis is the sort of innovator of the artificial pancreas and the open innovation community that sort of innovated and produced this so dana has diabetes and she wanted her insulin pump to work more like a pancreas but there were not any companies who were developing that product despite a lot of user demand and a lot of user need for it so she and an open user innovation community produced it they developed it and they produced it so now many people have artificial pancreases you can see one you can see hers it's actually attached to her body like pictured here um at the time that i met her so when this picture was taken um artificial pancreases were still not being mass produced by companies but still innovated and produced by the open user innovation community i think that her story and sort of her vision is what draws a lot of you here to this great conference organized by the great faculty and staff at nyu she saw a need that was not being filled by current solutions by current companies by current offerings and she looked outside of the firm she looked at distributed resources distributed expertise and figured out a way to sort of bring those together and combine those in a really new way to produce this thing that has been very beneficial to a lot of other people so i think that's what's really exciting about what we're talking about here today i think that's what kind of the shared vision of this conference is which is great to be a part of so my my sort of piece of this and what i want to invite us to kind of think about while we're pursuing this vision together i want to talk about the crowdsourcing platforms in particular that make this possible and i actually want us to focus specifically on looking at the structure of work on the crowdsourcing platforms so the crowdsourcing platforms to put in the language of the other panelists these are outside of the firm these are drawing together people who are outside of the firm i want to look at the structure of work on those coordination platforms and think about how the structure of that work enables or constrains different kinds of innovation when i talk about work structures uh often we're thinking about bureaucratic organizations if you've been in a big organization i'm guessing that you have been a part of possibly pitched a reorg reorganization reorg if you've ever said you know we should have we should have more boundaries spanning or we should have more cross-functional teams or maybe we should try a matrix structure if you've ever said any of those things then you have thought of work structures you have like a you have an intuitive sense of what structures support what outcomes so we're taking that kind of thinking that way of thinking about work structures and we're going to apply it to crowdsourcing platforms to sort of introduce this um idea and kind of like tune our tune our analysis towards this i'm going to kind of walk you through a research conversation that was happening at the time that my collaborators and i worked on the idea of flash teams so the conversation that was happening in the crowdsourcing community at this time so i'm thinking of like sort of academics who work on crowdsourcing platforms and kind of innovate what's possible using crowdsourcing so this was the research conversation that was happening at this time i think that um this community like so many of us and so many of the other panelists and so many of you all see how computation and sort of like networked labor is just fundamentally changing the way that we do work so i won't belabor this point but i'll just say the specific aspect of this that we were focused on was open call sourcing meaning when you have a piece of work when you have a challenge when you have something that you would like done you can you can sort of source it you can solicit participation online from anyone online who's sort of in the network so open call sourcing changes the nature of work pretty fundamentally at the time that we started our research crowdsourcing was largely focused on microtasks and algorithms that integrated the microtests so i have sort of stylized images here of what those microtask workflows might look like these might be familiar to you by now so this would be um you know if you're ever labeling an image or if you sign into recaptcha for example when you're signing into your email that's that's a microtask so most of the crowdsourcing platforms at this time were structured around microtasks and what was really cool and innovative at this time was that the microtests could be algorithmically combined so when you have a large group of people all contribute a small task you can combine all of those tasks in really kind of innovative and cool ways so this was mostly the crowdsourcing conversation at the time so again taking the analysis or just taking the sort of like work structure this is a lot more akin to an assembly line where you have pre-specified tasks predefined laid out and then automatically combined in in pre-defined ways the achievements of this way of structuring work have been foundational and important this is how we got a lot of the machine learning images that support ai this is how a lot of the um sort of like writings and books have been transcribed that are now available through google you can also see um that these can support really cool innovative real-time services like supporting uh question answering for somebody can just post this and then you have a lot of people available online to sort of work on this but the crowdsourcing community at this time was recognizing a challenge it's it's cool and it's interesting and it's innovative because it's happening online with all of these really cool platforms but it actually is a classic work design problem the problem was they were wanting to accomplish much more complex goals using crowdsourcing things that are more interdependent that require new ideas combining things in new and novel ways so at the time that we did our research this is where the conversation was sort of um staying it was remaining where all of the open and the more complex innovation was kind of out of reach uh what is pictured here is actually a crowdsourcing paper where the crowd was working on designing a new chair and you can see that the designs sort of like go back and forth between um and it just didn't converge into something that actually ended up working out so what we saw and the conversation that we sort of had in our team is that this was a classic work structure problem um if you have a really open-ended and complex goal like producing or innovating a new product it's really hard to articulate modularize and pre-define all of the tasks that will be involved and all of the behaviors that will be required for that kind of complex work so our our sort of problem statement is that crowdsourcing at that time was confined to goals that were so predictable that they could be predefined and our question our vision was can we use the same kind of open call platform approach to crowdsource something that was much more complex and much more interdependent i'm going to add a comment here to sort of connect this to some of the other panelists i think the way i would think about this is we're looking for production that is outside of the firm but we're drawn we're going to draw on some of the coordination structures of the firm we're just going to apply them sort of outside in the more kind of open community setting we thought that this vision um was the way to sort of create or sort of advance what was possible with open innovation with more sort of networked models of drawing people together and supporting their coordinated work so i'll now sort of describe the platform that we designed and deployed in a proof of concept study as an organizational scholar what's really interesting to me about all of this is some of the structures that we innovated are pretty classic i love that um in the in the paper that we wrote about this uh one of the novelties was actually hierarchy which if you think about it is one of the oldest management structures you could possibly imagine but hierarchy allows for different kinds of coordination than like an assembly or a microtask workflow would allow for so our vision was to create a platform that supported an alternative crowdsourcing architecture something where the crowds were structured more like organizations but still drew on the speed and scale of crowdsourcing platforms so i'll give at this point really like a nod to organizational scholars and organizational theory a lot of what we drew on to design and test our platform had we drew on from a lot of research on temporary organizations so people have done really great studies of movie crews or incidents command or some sort of like disaster response team these are similar to what we were envisioning with flash organizations these are um these are basically like crowds that come together temporarily and do something very complex and very emergent and very innovative so the question that scholars took to these was to ask what makes this possible what makes it possible for a group of strangers to show up and coordinate something so complex and something so interdependent so the answer that was consistently found across all of these studies is that people drew on their roles their expert roles so when somebody shows up to a movie crew they know who the director is they know who the gopher is they know who the gaffer is so they all show up knowing who they are and what they're supposed to do and those roles help them know how they're supposed to coordinate with each other the second really important piece besides the roles was the hierarchical structure when you have a hierarchy it's clear who is supposed to issue the next decision some sort of emergence something comes up and someone has to respond and issue sort of an adaptive instruction at that point that's what hierarchy enables so we wanted to design and deploy and test a crowdsourcing platform that embedded the idea of role structures and hierarchical structures into the platform so this really great fun amazing group of phd students master students me and my collaborator who's a faculty in computer science we had um a couple of really fun years designing this platform and deploying this platform that allows for basically designing flash organizations that have these role structures built in that have the hierarchy built in but still support automated hiring that we expect from that sort of on-demand hiring and open open call that we expect from crowdsourcing um the sort of technology version of role structures i guess i'll say um is that the platform interfaces with the labor market at the level of or sort of around the role so if you're wanting to hire a ui designer then you would put that into the platform and then the platform interfaces with the labor market to find a ui designer so that's now introducing role structures into the crowdsourcing approach this is how we allowed for hierarchy to drive adaptation in the crowdsourced organization um so basically this is the platform visualized here um and we configured it so that it it could be edited by anyone basically um and from a technical point of view this is this is simple this is very straightforward this is like what git github does um so basically you make a copy of it make an edit to it and then merge it back in so here they've added another team member and another task and then they merge it back in and now the whole team can see that this change has happened they can see the organization has adapted there are a lot of different ways you could configure this for our proof of concept we ended up configuring it so that these suggestions had to be approved by a manager so it ends up being a very hierarchical way of sort of adapting the organization over time but there's fun with there's you know we had fun thinking about ways that you could do this like in a democratic way or um you know you could there are different different decision rules that basically could allow the group to think about whether or not to adapt in particular ways the final piece that i'll note about the platform is that it did continue to support this open call hiring because the platform interfaced with at the time we did this research with upwork which has you know millions of workers who are registered so it interfaces with upwork and it issues an open call to people who have the particular role that you're hiring from for our study we configured it to be first come first serve and there are lots of things to talk about with implications of that decision so to sort of conclude i'll just say we did a proof-of-concept uh study where we recruited people outside of our research team who didn't have experience with crowdsourcing who didn't have technical production experience and we gave them the chance to use this platform to crowdsource something complex to lead a complex project i've listed here what the projects were there's lots of documentation online if you're interested i'm happy to provide more detail but the proof of concept was largely successful our users were able to basically crowdsource entire organizations which we defined as 30 people so they convened 30 people in uh six weeks and they were able to accomplish their really complex projects and they hired their workers in uh 14 minutes on average using the labor market um here's a description of the complex deliverables and now i'm just going to sort of illustrate what was produced what i'm trying to show here is just that they were complex these would be very hard to crowdsource with a different infrastructure obviously um these are the workflows that emerged this is basically like a stylized image of what was designed on the platform these are tasks each of these little squares are tasks these are the emergent organizational structures that were produced and then finally here are sort of the important uh statistics that sort of proved our proof of concept this is the average hiring time for each organization and then we also um just counted the number of times that somebody requested a change to the organizational structure or to the plan uh what i'm illustrating here is just indeed these projects could not be predefined there needed to be a structure that could be very adaptable as the work progressed so in conclusion i'll just say the vision of this conference and what this research hopefully contributes is computational systems are really fundamentally changing the way that we work the point of view that i wanted to offer is to really think about the way that we structure the work and how that helps us accomplish certain kinds of goals um or possibly constrains other ways of accomplishing that work thank you very much melissa fascinating i love this work on flash teams and anybody who has tried to build anything on upwork that involves more than two people uh would should appreciate what you've done and use this stuff so a great work okay i'm now i'm gonna switch gears and uh present some of my own work so uh professor tucci has talked about strategy and big questions of like which type of crowds one should involve and professor valentine talked about structure and i'm going to bring it back to management now my own work in the last um sort of 10 years has focused on crowdsourcing but over the last 20 years i was interested in issues of boundary spanning and as many of us are switching uh to remote work in the last year and a half i realized that some of the stuff i've been thinking about for 20 years might be actually irrelevant um above and beyond with what i i've been studying and crowdsourcing so i'm going to try to go back in history to some of the work i've been doing on crowdsourcing so um to go back to this ideas that my colleagues already shared which is the idea that innovative ideas uh don't just fall from the sky not necessarily sourced by a lone genius sitting in a single room but they're rather coming from the integration of diverse things from different spheres of life as for example is the case um if you look at the books that were published on the topic of innovation you see things like borrowing the brands or managing fact uh with medicines being famous for integrating diverse types of art and science to come with for innovative outcomes now uh to bring this a little bit closer to business um you you might have heard the story of how the uh ford assembly line was invented henry ford visited plans especially in nitpicking district meat plants in chicago saw the idea of a moving assembly line and introduced it into the production of the model t ford vehicle which resulted in the time being saved from 12 hours to assemble the vehicle to 93 minutes so you know one could kind of say on the surface that ford invented the assembly line but as many of my colleagues already argued in this conference i would argue that ford brought in the idea of assembly line into the automobile production and i had to do quite a lot of work to make it happen and that's the point of my talk like what is that work of boundary spanning that needs to take place to make the things happen so when you think about like what happens today with the idea of boundary spinning i'm sorry uh technology is basically enabling us to combine these diverse ideas across the globe right so we have enterprise social media and zoom that's supposed to help us like dial in anybody anytime right to tap into diverse expertise we have the idea of uh global work right if we don't have expertise locally in our team we can you know call japan and get that expertise from there we have online communities like wikipedia that professor tucci talked about where people from very diverse backgrounds work together to innovate and we have crowdsourcing has been mentioned many times where you often pick people you innovate with or they some self-assemble or decide to contribute to your problem right so technology is enabling this at scale and why why is this uh an important piece of it well because my colleague professor schilling lead in this conference wrote an amazing book about how this type of integration of diverse ideas for innovation can happen inside an individual genius right so she writes about elon musk and about einstein nurses but i i'm going to argue that in the day and age in which we live there's too much deep expertise you need for a lawn genius to integrate at all and as a result you need to build the team around the project and you do need technology unfortunately technology does not have to make this happen it only enables it it still takes management and skill to make this integration happen and even more unfortunately we have this ideal of integrating diverse expertise across boundary but the reality is and that's based on research of many people studying organizations that the ideal does not happen and instead teams with more diversity tend to perform worse in studies than teams is less diversity and here i don't mean racial diversity i mean all types of diversity professional diversity uh tenure and organization belonging to multiple organizations so when you think about having a lot of diversity to deal with it's more likely that people will fight and as professor neely who will give her keynote later today writes about in her book create fault lines where people align across the aisle and kind of stop talking to each other so with that said like what can we do realistically well one thing we can start thinking about is the rewards of actually making them happen organizations that are able to build capabilities that consistently integrate diverse expertise across this expertise boundaries tapping both on internal expertise and as professor tucci said external expertise tend to gain competitive advantage so while this is very hard it is worth doing so i'm using apple here as an example of a companies that have been successfully integrating design and technology at the very least if you want to add to it marketing as well you can but two very difficult things to integrate well it has been doing it for years since the very first computer to today right and doing it well gives them this long-term advantage that is very hard to replicate now um with that said like let's look deeper into what it takes to be effective in boundary spanning so one way in which organizations tend to address this is by relying on individuals people like us who end up being boundary spanners who end up doing the work of connecting translating and integrating diverse ideas so as you see these individuals become critical to organizational success and i'm sure many of you listening to me now i think i am one of those people right i am a person who is able to integrate connect and translate this ideas now um give the one thing we know about this people in this roles is that actually very few are able to perform these roles well more often than not people who are nominated to be boundary spanish like sort of they're seen as somebody who is put in position in position as a team leader or project manager or even senior leader to integrate this diverse expertise for innovation well the number one thing that happens in this people become very stressed and psychologically uncomfortable because what they are asked to do is to basically be um belong to multiple cameras and to none deeply enough right so to be an effective boundary spanner one has to belong at least partially to multiple groups of experts and uh but at the same time to never take one side fully right to keep negotiating you know yes i agree with you but let's look at this side of it maybe designers are right here maybe engineers maybe sales people and um in fact they talked about six people sales people tend to be one of those boundary spanners and it's well known that their their job stress is extremely high now on top of needing to do this sort of being in between person these people also need to have sufficient expertise and credibility to speak to multiple groups of experts credibly in their language and with expertise right so it's often we are good at two things right maybe three but in modern organizations we tend to be in this position of boundary spanners where we need to have at least some expertise in three four or five domains and not only that but credibility to to speak to people in this domains so finally let's imagine you achieved it all you you work through the stress you have developed this expertise well now you become this primary communication channel and sometimes even the bottleneck right there is a ton of stress in this position when you do it well because everybody relies on you to do this knowledge integration good news is so that's the bad news the good news is that there are that once you're able to cope with this stress there are actually rewards and a research literature documents that those who do who are able to integrate diverse expertise and practice broken knowledge connect people and do it well basically enjoy the sort of innovation innovative outcomes that i talked about like the example of give a ford um like people do say you know addison was able to integrate his diverse ideas and then for you know regular organizational members not edison's of this world there are usually career advancements associated with being able to fulfill this role so this is my most important slide and um i am going to slow down and talk about what i learned over my 20 years of studying people doing this work and you can think about your own organizational context your own job to some degree to think about what this translates to i am going to use examples of things i study most often which is new technology development projects so the first thing that people in this position need to do and is to figure out at any given point in time which difference in knowledge or what i call boundary is the key one to deal with to span at this point in time now this seems obvious right like but if you slow down and think about it i'm going to use an example of a outsourced strategic innovation technology development project in this case the first boundary that was most prominent was most relevant for people to deal with what's the boundary between clients and the consulting firm right people come from different backgrounds they don't share a lot of knowledge so they they sort of start struggling with who's who whose expertise should be done should be the main one leading the project obviously clients have the money and the authority but the consultants are bringing the expertise that they're paid to bring in so it's important that they have this leading role well that's clear enough but if you as i'll tell you that as the project progressed this boundary between the clients and consultants became less important like people were able to establish shared knowledge and shared understandings and work together and the one between designers and technologists uh became more prominent one right and as they work through and you know work across this boundary the next one that became more prominent was the one between newcomers to the project and the old-timers and these things are constantly in fact so the number one question uh i recommend this people who are trying to do this well ask is like what is the current boundary no difference that i'm dealing with that's causing misunderstanding and tensions the next thing is um about cultivating shared identity and avoiding us versus them you know it's easier said than done but focusing on what we are doing together we're building this system it's very helpful to actually put a little um excitement into this like we are doing something new and exciting we're talking about innovation projects uh to speak the language of we actually using the word we as much as possible even if people are not collocated um is very important uh focusing on shared interest if we do it well this is what will we will kind of the rewards would be we will you know revolutionize the industry or we will just finish the project and build our portfolio etc um creating shared space is particularly hard in day and age when we're working on zoom well you can argue the virtual shared space is not so bad right you have your project repositories and you have your um you know githubs and shared spaces and maybe the physical space is not there anymore and that may actually be a problematic because the physical space is necessary to build a social space so professor neely in her book which i will introduce later talks a lot about sort of building the socials that we're now increasingly missing in this online collaboration world why is this important well it's important because it helps us overcomes us with exam right we both see each other as human the next uh critical thing is to build shared practices um by which i mean routine sort of things you do together such as daily standing meetings or weekly meetings and sometimes not just meetings but just check-ins and sharing what you've done for the day and shared language and by shared language i mean not english or spanish or russian in my case but rather the specific language in the projects that only people on the project understand and others don't now i think uh working in organizations you may be frustrated by okay these people in this project speak this language nobody understands they have abbreviations and they refer to certain power points but that's exactly what needs to happen to integrate across this university to have this language nobody else understands um it may be a bit counter-intuitive but what it takes to innovate so to build something new right you have to have a new language um and la and the last one that is sort of important by itself is to work to overcome power differences now um what do i mean by that you might if you are managing a team that is not about focused on innovation all of the things i said above still apply but sharing power is a critical thing for producing innovation because what needs to happen is that people who are not necessarily in their high status or high power positions need to be able to be comfortable enough to challenge people in the higher status position and by that i don't mean like organizational power so i'll give an example again from my own work so the project was lab um that i studied for for a year was led by strategy team and the strategy team sort of sold the project and defined the requirements with a client for three months right so even though the strategy person on the team wasn't the most senior person in fact it was very young person this person became a de facto high status person right this knew everything about the project uh later on uh as the designers were designing the product they were actually more senior and often won awards in their area but for this project their status was lower they joined later and they were not heard at some point they basically gave up challenging as a strategist and uh eventually the technologists joined like the software people those were even less powerful right because they were behind in time and and they basically had hard time understanding and challenging everybody else because they would thought look this is what we already decided so the critical part here is to enable folks who are at this point in time in the low status position to challenge um what's been done already to undo what's been done already and to challenge people in high power position and finally um the last two steps are sort of uh meta steps which is in order to deal with the stress i outlined and to stop being a bottleneck it's critical that the number of people doing boundaries spending on the project gets larger you have more people who are who are like you i recently was hanging out with a friend who was leading a last a large team and she was saying i wish i cloned myself right so this is the idea try to clone yourself spend some effort cloning yourself uh in the team like that and finally you have to constantly reevaluate which boundary is the most pressing one now because as as i mentioned earlier the organizational boundary was a pressing one at the time when the project started but it was um sort of other boundaries as other differences between people that became problematic later on so i um wanted to share this one quote from one of my later studies on offshoring um this is a project where everybody was a software engineer there was no expertise diversity but as in many projects you guys working on with global workforce right people in um other countries in this case india were contributing to this project but the point again is this is a project focus on innovation complex development in in production of a brand new system so this is a quote from a bank manager uh who i interviewed who was working with the vendors in india where everybody around this person told me they are the true leaders through boundary spanners enabling us to innovate and by everybody i mean both people in the united states and in india where i went to collect data so what he said is doing this stuff complex development in india is difficult because there are huge cultural differences and not in the ways that you think when about it when you go to classes on cultural differences a lot of the people want to treat indians as second-class citizens he's talking about american workers here my project manager and i have gone through a lot to make sure that that does not occur if it occurs you are shot in the foot as far as accounting and innovation with this offshore guys it just won't happen innovation is something we're constantly looking for but you do not find it unless you create unless you sow the seeds so with that said uh the last sad part is that it's very hard to do in virtual work so this project while it had virtual component right people were still around the notices this isn't all the project and it is particularly difficult to do in virtual work not only because you know you're more stressed because the family is there right like in this picture but because we don't have the social fiber to help us overcome these differences and see each other's actions as coming from good will as opposed to sort of other places like arrogance and a lack of professionalism um and i am going to come back to the stuff and on crowdsourcing to to integrate with my friends here um professor tuccin professor valentine to say that these things are particularly difficult when you're dealing with crowds it's difficult within the firm we already deal with a lot of diversity but at least you know what expertise the experts are coming from it's difficult with inter-organizational projects and alliances because you know which differences you're coming from but usually on top of technological and expertise differences you are also imprinting organizational differences but it's even harder with crowdsourcing because you have no idea what differences you're dealing with so if you guys remember the beautiful landscapes that professor tucci showed we shared with you right you're kind of working around this landscape you have no idea where this peak in the landscape is and what kind of difference is causing the the high mountain there to get the answer and you are not seeing the answer because that's the idea of the graph and with that said i'm opening it up for questions to everybody and um i hope um you guys have saved some good questions for us so i stopped sharing the screen uh and melissa if you can turn on your camera so we can handle some questions and let me start with the question i think the first question in our chat is for professor valentine um so melissa fascinating material right and the question is uh do you have any thoughts or work on flash flash teams in the matrix organization so with dual reporting structures because right you describe the one with small a single reporting structure it's a little bit overlaps with my talk right so you kind of it's one thing to do with one difference what if you're dealing with multiple differences so i'm curious to melissa what do you think yeah that's a great question um and it made me laugh a little bit because um it's some of my work is in like large bureaucratic organizations and matrix structures are known to be so complex they're so useful and they're so valuable and they're so complex and then flash organizations are complex because you end up with like adding i.t and not being able to see each other which just also makes it more complex um so i appreciate natalia your sort of noticing that like what what's desired i think with this question is basically integrating in different ways across a lot of differences um but it certainly underscores what's really complex and what's really hard about innovation and about um structuring for innovation um so i think what i can say based on the empirical data that we collected from all the times that we've run flash organizations um is that currently um the boundary spanning happens in private messages like it's all interpersonal dms so what i like about your question is it's the opportunity to sort of formalize some of that backstage interpersonal boundary spanning that's happening so i think it's a cool idea i've not tested it thanks melissa yes interesting um next two questions i think are for me i'll kind of try to answer them almost at the same time um the the question i see here is whether using facilitators and teams uh that have very strong expertise and project management helps so i uh have thought about this question for many years so thank you for asking because i have observed uh teams with and without such facilitators or sort of hired project managers in fact the very team the very situation i told you about at the beginning had such a person and the problem you know if you think about my long list of bullets um and requirements for a boundary spanner which relates to that second question is that such people rarely have expert domain knowledge right so they usually know how to run a project not the domain knowledge and when it comes to innovation on a level of in integrating domain knowledge sort of diverse knowledge they tend to not be very effective in my experience because you really want the key sort of to boundary spanning for innovation is relating diverse expertise it's not just training project well so i find that in fact introducing such a person creates a new boundary so now you kind of have a boundary that already existed and yet another one between somebody with project management expertise that's been my experience uh i think some of these facilitators know how to deal with that new boundaries they just created but in general i'm not a fan of this idea when it comes to innovation type of projects um now the second question is how do you know so whom do you nominate and how do you know these people are effective it's a million dollar question when they teach cases about this um where you have a kind of a real case we tend to argue out who would be a good one right like so here is a protagonist of the case and here is a maybe an expert and this and that and we we sort of argued out in the reality is there aren't perfect ones so what you end up with is extremely long list of potential criterias that such a person need to meet no human being can do that as a result you start compromising to say okay well maybe they're not going to know as much about software we'll just focus on hardware and design or something like that but the number one thing i learned from looking in the series you have to check in you have to you know the senior manager keep checking if they're doing the job and how do you do that is by talking to the team you basically have to check in and when you talk about global projects it's critical to talk to this remote people who are not in senior positions to see if they're heard right if the manager is enabling them to be heard and if they're teaching the last piece we're going to say we for example looked at whether immigrants are good for this job when uh doing projects with the country from their country of origin and in a nutshell is not very often they're good if you're interested i could refer you to our people in this topic or some short summaries of it for practitioners because uh immigrants bring in a lot of that skill right like they can talk multiple languages but there is a lot of emotional damage coming with that as well um a question for professor tucci i came in so i haven't had the chance to read it yet but i will now so the findings remind me of the impact of formal versus informal firm structure on their innovation output the more formal procedures the less rate of innovation output your idea seems to me that as an extension of this relation beyond about borders of the firm what do you think chris i am going to take an opportunity and actually ask you to speak about your other research on the as well if you don't mind on the level of formal structure for innovation the research you've been doing with professor shapiro i think comes right in if you don't mind yeah actually it's funny um okay so there are different ways of thinking about this indeed it's a really interesting question i think there could be a kind of analogy so let me just talk about the analogy part where this is coming from and then i'll talk about some empirical work that i've done you know in this area so um so the analogy is you know if you have kind of a formal structure that you know people may be less willing to do things because they don't like the bureaucracy or they feel like they're being too controlled um and where i feel like in the crowdsourcing exercise uh i don't think that there's a lack of willingness you know i think that when you have kind of a an open crowd let's just say um that day people would will suggest things that are that are crazy that might be crazier than they would have otherwise so it's not it's not this lack of i think if you put if you get a linkedin group and you say brainstorm give me a hundred ideas you know and this is kind of related to what hilar was saying at the beginning too you know if you say give me 100 ideas you know they'll give you 100 ideas so i think that you know it's not it's it's not a question of them hold holding back they're doing everything they can but but maybe that the very organization of this thing is kind of preventing them from [Music] deviant behavior or completely crazy you know ideas that might be interesting for other people now um thanks also for the prompt on the empirical work because on the empirical work actually what i was doing is studying a little bit about the levels of hierarchy and how how many so-called big ideas come out of these you know these things and looking at cross-sectionally and longitudinally um but mainly cross-sectionally you know the firms that happen to have more deep hierarchies tend to have more important ideas and one of the arguments that we were making there was that there were you know that there was more of a chance of knowledge recombination if you move up a hierarchy so there could be a role of hierarchy and and we're just speculating at the end of that paper by saying well you know maybe there's an it analog for this that flattens it allows you to have a flatter structure but that also allows for more knowledge recombination so you know what we were looking at was from several years ago decades even um and you know back then maybe they didn't have the same kind of i.t platforms that we have now to do knowledge recombination so maybe there's a role for hierarchy you know in knowledge recombination so that um that would be my you know sort of uh reaction to this question thank you very much for that thank you chris i wanted to make a comment i think for the last 12 years i've been studying crowdsourcing i think we're increasingly accepting that there is a role for hierarchy in general right if you look at melissa's talk right you're basically creating a hierarchy on top of a crowd right and a lot of we see actually related to the question of ideas uh dying right the crowd produces ideas that eventually die the questions that he loved earlier right it is really if you have a good manager part of their job is to organize it in a way that they don't die instead like i heard and integrated and of course i pretty much i'm talking about managers so the in reality so i think we are kind of acknowledging now that crowds cannot do everything they are super powerful but when organized yeah so uh question for melissa an actually fascinating question so melissa your work tends to be on flash team in kind of software virtual space the question uh is how do you translate that into the things as a software development yeah that is a great question um let's see so the industry examples that i'm thinking through right now um are not let me look at the question again i think are not hardware or manufacturing um so when you have when you have an industry that has a lot of like physical equipment um i'm i'm sure there's examples of that i don't have one i can find one um the ones that um come to mind are like artella which has the platform and the like community of

2021-10-05 15:29

Show Video

Other news