← The Directions on Microsoft Briefing Podcast
The Directions on Microsoft Briefing Podcast
Three Smart Guys: Can AI Save Collaboration Software?
00:34:00
Loading summary
Transcript34 lines
- [00:12]
A
Online collaboration. It's essential for teams to work together and get things done. Yet is collaboration the most despised or at least the, you know, most underperforming category in enterprise software software? And if so, why? If not, why do some people think it is? And lastly, how can AI perhaps come to the rescue? Hi, I'm Barry Briggs at Directions on Microsoft, your independent source of information on all things Microsoft enterprise software, cost management and licensing. I'm joined today, as always, by Peter o'. Kelly. Wave, Peter. And George Gilbert. Wave, George. Okay, so collaboration, that's our topic for today, but let me warn you in advance, three of us have all been in and around enterprise collaboration platforms for literally decades. So stand by. Peter, I'm going to start with you. Is collaboration a dirty word in enterprise software? It strikes me that this is a timely topic, as you reminded me, given that we just passed the 30th anniversary of the launch of one of the premier collaboration platforms on the planet, Microsoft Exchange. So what do you think? Is it a dirty word? Is it loathe with justification?
- [01:33]
B
Yeah, so it's a long story. I do think it is very much a stigmatized word today. Although I think over the course of this discussion, I would like to make the case that I think it's about to be revitalized, maybe as collaborative intelligence, but as you suggest, with the integration of AI, I think that what historically was known as collaboration is going to be redefined, reconsidered. So I think one thing, if we stand back for just a minute and say, making sure we're on the same page for terminology, just a couple quick things through this and then maybe I'll jump into some brief comments about both notes and SharePoint and why they have both been not on the top 10 for end user loved products at any point in their histories. So, but just for some terms, it's important to distinguish between communication, which is just transmission of information from point A to point B, and collaboration, which is joint purposeful activity, usually in a workspace, or today we might say a project or, or in a notebook. But it's important to understand they're different things and they have different goals. So historically we could think about both synchronous and asynchronous communication. Collaboration.
- [02:51]
A
When we think asynchronous, there is a. The line is blurry right between the two. I mean, you can have collaborate on projects within a team's conversation and so on. So I think that's part of the problem. And let me ask you this, do you think that's Part of the problem that align.
- [03:10]
B
Yeah, so, and I think historically. So first, you know, the software industry has a fun tendency to redefine things, you know, so if we go to the dictionary and say, what are the definitions of these terms? They existed a long time before software was out there. And so the distinction, one important distinction for me between communication and collaboration, as I said, is collaboration is really sustained, joint purposeful activity in a workspace with shared artifacts, with audit, with security, with everything else on top of it. And I won't subject you to the whole asynchronous versus synchronous part of it, but just to say where I was going to go with that is to say the part that has been most missing historically is asynchronous collaboration. So it's the content based collaboration that is not predicated on us being online in real time. And so that was Notes, databases, Notes applications, to a certain extent, SharePoint, sites, and other things as well. But I would argue that that is both the most hated part of the story historically and also arguably the most productive or the most potentially productive for organizations that don't have teams, all of the participants in the same place. So I think that's the part that's been stigmatized and I think the reason it's been stigmatized categorically is it's hard, it's really hard to get people aligned on goals. It's hard when the incentive systems are saying, don't do deep work, you know, do agile, be agile and just punch things out. It's difficult when people are living more in communication tools than collaboration tools to get people to actually engage. So I think historically, you know, if you look at it and say, why did you know, why is the reputation of Lotus Notes so tarnished? It's because a lot of people just wanted to do email and calendaring and scheduling and they saw these asynchronous content based collaboration document databases and thought, yeah, I don't get paid to do that. That's, you know, and other people are not playing along with me. I can contribute a lot of content and I don't see people engaging with it. So I think Notes was stigmatized because people really wanted the email and calendaring and scheduling. And often there wasn't an organizational case for saying, here's why you should do document based collaboration as well. And then just briefly, on SharePoint, I think one of the reasons that SharePoint's had an uneven reputation historically is because SharePoint was supposed to. So SharePoint was to complement Exchange. So Exchange was going to do the email and calendaring and scheduling parts of notes and then SharePoint was supposed to do everything else that Notes did. So that workspace based document centric collaboration and it just never really came together for that. It got the basics okay, but then there was a whole series of companion products with it that came and went and also just a lot of confusion about where it goes. One thing I will say, I promise I'm going to stop talking in a minute here. One thing I will say about the SharePoint part of the story is that perversely one of the challenges with it is that the product teams within Microsoft actually don't collaborate very well or historically have not collaborated together very well. In some cases they're competitive. So if you go back a decade ago, speaking of Microsoft and anniversaries, there was a moment in time where GigJam, Delve and other the Microsoft 365 groups were introduced at a Microsoft event. And that kind of tacitly pushed SharePoint off the stage. And the Exchange group once again was taking the charter of the communication and collaboration part of the story. And then the organizational dynamics changed again. It went back to like the pendulum swung back and SharePoint was the answer. What was your question again for everything collaboration related? And then other things factored into it like the acquisition of groove networks that ultimately ended as SharePoint workspaces. So anyway, I think the short version of this is these have been stigmatized in the past because a lot of people just want give me email calendaring and scheduling and now with teams or slack, give me IRC style conversations and then don't tell me I have to change the way I work for everything else because it's hard, it's hard to be able to support that.
- [07:38]
A
It seems to me though that part of it is that the vendors are stuck between different design points. But if we take both lotus notes and SharePoint as an example of one poll, which is really to do anything beyond file sharing, you have to write an application of some sort. You have to have app developers on site to write your applications. And you know, people, a lot of people didn't want that, still don't want that. And you know, the overwhelming majority of SharePoint sites today are again still just, you know, glorified file shares, you know, so that's sort of on one side of the pendulum. On the other side of the pendulum, pendulum set sides anyway is are the canned Apple canned collaboration apps. And you think of things like Asana, for example, you know, which are pro, you know, they advertise themselves as project management applications. Right. But they're collaboration applications, but they support one application, which is project management. Oversimplifying a little bit, but you know, so, you know, you could do that in SharePoint or Notes, you'd have to write a very substantial application in them to do it and it'd be difficult to handle all the edge conditions and exceptions and so on, which Asana and others have and notion and so forth. But it seems like you can't have it both ways. If you want the pure collaboration platform, then you have to write applications, which nobody wants to do. If you want the specialized applications like project management, then you have to buy a lot of them, one to support each different application model that you have. So I think maybe that's part of the reason why collaboration is still kind of getting a little bit of a dirty word, which is it's hard to describe what the category actually is in 30 words or less.
- [09:47]
B
Yes, I definitely just proved that. So I agree with you on that part. I mean, we did try with notes with the nifty 50 to have a set of application templates that were just out of the box that you could leverage, but again and make those readily customizable. But I agree with you that when you cross over into another very long term stigmatized word of workflow, when you start talking about applications and we say it's workflow in some cases means like it's document centric and or more human intervention is required at various steps to move the process forward. That's another one that is just, you know, it's a third rail for software companies.
- [10:27]
C
I have a question because I'm listening to this and I'm wondering whether part of the trade off between these general purpose collaboration platforms like, like Notes and sharepoint on one side and then the sort of point solutions like Asana on the other side, when you have something general purpose, there's a lot of overhead in the user interface to allow customers to specialize it. And that like one of the reasons Notes was challenging or was challenged as an email environment was because it had so much general purpose overhead. And you know, Asana and, and, and the others can then have a very special purpose experience that's just tailored for that one.
- [11:19]
A
Exactly.
- [11:20]
C
You know, that's just a fundamental tension.
- [11:24]
A
Yeah. And you know, but Asana, Asana does one thing really well and arguably notes and SharePoint don't do anything really well out of the box, but can be tailored, customized and developed to become, you know, a specialized application for that Enterprise. You know, George, you said something I want to, I want to come back to because you said the W word, you know, workflow. And you know, we were talking before this and it feels to me that, you know, workflow and collaboration is another one of those lines that gets blurred an awful lot. And I have to tell you, I have to tell you, I wrote a Think Week paper back in the days when Bill was the chairman of Microsoft. And he had this tradition, just to digress for half a second, he had this tradition of every year anybody could write a paper on any topic that they wanted. It was called Think Week. And he would go off that week and read them all. You know, I, I always told people that it was a double edged sword because he could either say that it was really good, in which case it could help your career or you could say that it was that it sucked, in which case it would not help your career. Anyway, I wrote a paper on, about business processes that fortunately Bill liked. And I talked about the spectrum of them. There's very deterministic business processes like Swift, for example, the financial interchange, where it's very clear what the steps are, what the success criteria of each step are and what, what success at the end, you know, constitutes, that is the funds get transferred, you know, and then I looked at other kinds of processes like, you know, we could, we could argue our collaboration. So thing, you know, I think there are two examples I'll give. One is an rfp, for example. So an rfp, somebody writes a draft, it goes through multiple rounds of review and it could be a number of n number of rounds of review before it finally gets declared. You know, it may just be that it's, it's declared done on the day when it has to be shipped. You know, and that involves a lot of people collaborating. Is that a process? Is that a workflow? Is that a collaboration? But I'll tell you the other one that, that, that Bill particularly liked was I, I found there was a, one of the US States, there was a process describing how the act of dying is handled and it actually turns out to be quite complicated. You know, when somebody passes away then you, all these things have to happen. For example, you know, the CDC has to be, has to be notified if it's a, if it's a disease. The VA has to be notified if it was a veteran. Social Security has to be notified if the person was receiving Social Security benefits. And there's a whole bunch of other things that have to happen. And I actually found this documented in, I Can't remember which state, but one of the state, one of the US States. Again, is it a process or is it a workflow or is it a collaboration? And you know, I think George, you know, your friends, Salonis would look at this and say, well, all of these things leave traces in a database someplace which we can mine, so therefore we can do process mining on it and therefore it must be a process and it's not collaboration. And others might say, well, it's collaborate collaboration. What do you think?
- [14:51]
C
This is sort of near and dear to my heart because this is what I've been working on, you know, for like a couple of years now, which is like you, you talk to, you talk to someone like, like Sridhar, the CEO of Snowflake. And he, and when you ask him about enterprise applications, he's like, well those are just workflow rails that no one really uses anyway, you know, because they can't handle all the edge cases, you know, and then you're like, you go to the Salonis folks who do mine all the edge cases and they're like, yeah, but we don't yet handle the, the, the task contest context, you know, like the, the personal productivity stuff or maybe the sort of interpersonal collaboration outside the rails of the multiple enterprise apps. So I think we've had these artificial barriers that are going to have to start coming down and I think this is bringing me back to my favorite topic. When we start taking the hundreds of silos of the analytic sort of data aggregates and the line of business applications and we start turning them into a human and agent readable kind of substrate, some sort of, you know, something that encodes all these semantics of the applications, of the collaborations again in a human and agent readable form. I think then we start to capture, you know, the rules that must be followed by. And then there's a place for the context that needs to be presented when judgment is required and where the rules break down. And then I think these categories start to melt away.
- [16:43]
A
Yeah, that's interesting. It makes me think, I'm curious what either of you think. Someday in the distant future, do you see an agent coming along and looking at, you know, collaboration resources like SharePoint sites and you know, all the logs as well as all the content and the Microsoft graph and you know, all the different forms of communication from Slack to teams to whatever and sort of putting together some sort of aggregate, this is what the hell's going on in my enterprise. Does that make sense?
- [17:19]
B
I mean, from, from my perspective we could, you know, to be fully terminology updated for this. I think today we would say that is an organizational world mod. So you want everything you know about the organization and the relevant parts of the ecosystem within where it's operating within. And you don't want to make a hard distinction between this is stuff that's in my traditional databases or my enterprise apps, and this is stuff as you just suggested, that's more sort of fluid and it's in teams or it's in web conferencing or other things. But you're creating artifacts that have information about the organization and you would love to be able to stand back and say I want the totality of it. When I say what do we know about customer Acme widgets? We want to know everything. Whether or not.
- [18:06]
A
Is that what you mean by collaborative intelligence?
- [18:10]
B
It's part of it, it's not all of it, but yeah. So on the collaborative intelligence part, just to maybe put a pin in it for a second, we can come back to it is basically collaborative intelligence to me is taking the things that have been really drudgery in the past for that sort of document centric workspace based collaboration and make it easier. So take the parts that nobody really wanted to do like distilling and synthesizing and harvesting all of these documents that get generated in different places or in email or instant messaging, whatever, and then also take advantage of AI assistance within the collaborative endeavors to say bring in the bigger context window. So like humans don't have that, that level of scope that they can go look for surprises and things that if you don't know to try and connect these things, you know, so that's, that's complementary. So I'd say collaborative intelligence to me is try and do what we were doing before, but help to automate the parts that were unpleasant before and then bring in new capabilities to complement the things that humans are doing. But with that, I'd be remiss if I didn't also say there has to be the organizational change because if everybody is still doing the information worker equivalent of agile and it is tacitly telling people you shouldn't do deep work because we're not incenting you to do deep work, then none of it's going to be successful.
- [19:41]
C
I want to key off something you said, this world model. You know, I've been doing some research on my project and it turns out that the definition of management and the definition of a firm is a shared model of reality and that the functions of management are planning, coordination, resource allocation. And I think there's there's one more that I'm, that I'm missing, but it's based on that shared model of reality. And so what you're describing and what we're talking about is if you start breaking down all those silos, not just the applications, not just the analytic data, but the collaborative intelligence. And it's all one substrate. Some are hard rules, some are context for judgment that's attached to the rules. So, you know, when the rules break down, like, then you redefine what a firm is and what it can do. And you know, there's this popular myth going around that we're because of agents, we're going to have, you know, the first billion dollar company run by one person. But that's personal productivity. That's about task productivity. And it leaves aside the fact that, you know, there's this, there's this world model of a firm that separates a bunch of individuals as contractors in a marketplace from people who are organized towards collective action by planning coordination, resource allocation.
- [21:15]
A
You know, I love the way you put that, a shared model of reality. And you know, it makes me think of all the various dysfunctional organizations that I worked for over the years. And you know, the one thing that they all have in common is that the people within that organization did not share the same vision, the same view of quote, unquote, reality, reality, or at least what the purpose of the firm was. I mean, that if you were to boil it down to a single thing, it's always that you had some people who are going this way and some people who are going that way and they all think they're going the right way and the other guy is an idiot and so on. So, you know, I wonder to what extent you guys think that, you know, the whole purpose of collaboration platform in the limit, right, is to drive that organizational alignment or to at least, you know, provide, use your word, George. Substrate for organizational alignment.
- [22:09]
C
I, I couldn't agree more. And I, you know, I, I was going back to read that the, the birth of the OR chart, the hierarchical or OR chart, it was inspired by the Prussian military, but it was first really implemented by railroads because you needed to manage this very large scale fixed asset that was geographically distributed and you needed to have like clear lines of authority and responsibility. But we've not hugely updated that concept. We have like, you know, matrix organizations and like network organization, but we don't have the substrate for alignment. And like, if you, if we have KPIs now because we have enough digital technology to measure things, but if we have KPIs all the way from little activity level things that we measure all the way up to the North Star objectives for the organization and they're learned, these connections that becomes the quantifiable measures of this shared reality and can help orient people. It's, it doesn't, it doesn't answer. Barry, your, your notion about, well, are we on the same page in terms of like, are we building the same platform? You know, but that, that becomes part of the substrate. Like, like a digital twin is a platform that organizes work or it can be anyway, that's my two cents.
- [23:49]
A
Yeah. And you know, what strikes me about this conversation is that we have a shared vision, I'm going to assert, of what a collaboration platform is. And yet that shared vision that we just described is light years from where the technology is today.
- [24:08]
B
So I think it's getting closer, you know, and maybe that's a segue into talking about some of the players that are really driving this today. Both the small specialized players. Barry, you mentioned notion which is probably the most successful pure play, asynchronous collaboration platform today. And it's interesting for a lot of different reasons, but you pull it back to the bigger Microsoft picture and the ongoing competition between Microsoft and Google. I think there is a case that this is now moving to the center of. There's all the stuff that's commoditized. So yes, you must have email, it must be auditable, you must, must support documents, they must have, facilitate real time collaboration conversations inside the documents and stuff. But now we hear the word notebook a lot. So we've got Google Notebooks in Gemini and Google Notebook LM and now we've got Microsoft Copilot Notebooks. And it really is a pivot because it's saying to your point a moment ago, this is basically a way to organize comprehensive collection of resources that we need to accomplish this task, this process or you know, to bring people up to speed. And it really is a new take on a lot of the things that historically we were trying to do with Notes.
- [25:26]
C
One quick, one quick comment on that. Which is you're talking now about the tool layer, but the substrate. You know, Barry, when you say like this is, I mean, in fact, in effect we're talking science fiction. But Microsoft has announced the substrate for this, which is Work IQ based on the Microsoft graph fabric iq, which is, you know, their ontology like Palantir. And there's, there's one other, which I
- [25:55]
A
forgot, but Foundry iq.
- [25:59]
C
Foundry iq. Oh, okay. But you put all those Together and you know, you let them mature and over time that becomes their answer to Palantir.
- [26:12]
A
Yeah. So that leads me into my next question and last question, which is our discussion point, which is as we think about the evolution of the Microsoft ecosystem, Peter, you mentioned copilot notebooks and I think now I use Google's notebook Lm a lot of Claude code a lot and you know, I have evolved to the point now where I think of them as my assistant and to some extent I think of myself as assisting them rather than the other way around, you know, and you know, and I think, you know, it's possible that the IQ brands of platform software that Microsoft is coming up with, combined with Copilot Pages, Copilot Notebooks in the limit, could become that kind of enterprise platform upon which you can, but you can collaborate at all times with anybody and with the entire base of enterprise knowledge. Does that make sense? And so you know, you are sort of much more tightly integrated with everything that the enterprise has ever done. You're much more, you're much more bound up with all that, with all the enterprise knowledge that that has been accumulated as well as being connected to all the people that are, that are part of that part of that enterprise. Does that make sense?
- [27:33]
B
Yeah, it definitely makes sense to me and I think one way to frame that would be to say you will go use a copilot notebook or you will use a Google notebook in either workspace and or in NotebookLM. When you say I want to engage a group of people to collaborate on something that we haven't done before or you know, it's a one off thing like hey, we have a new competitor, we need to rally everybody and figure out how we're going to address this. Or we just got this huge new regulatory thing loaded on us, we haven't done it before. How are we going to bring all of the domain experts together and figure out how to assess this and respond to it and then figure out how to put it into our systems. But I think the notebooks are the places where you'd say I have a new collaborative endeavor that I need to do with other people and that's going to be the starting point and at the end of it, if it's successful, you will create resources, process definitions, as you say George, other artifacts that are going to be part of the digital twin of the organization then.
- [28:39]
A
So where do you think we'll be in five years? And if you're a Microsoft customer and you're looking ahead, is the future copilot notebooks or is that just a step along the way?
- [28:51]
B
So from my perspective, and this has been kind of a recent change in perspective for me, because I will say, I will acknowledge that I am a very long term OneNote customer and I have been frustrated. You guys, you guys have both heard me lament this before, so I won't subject you to it again. But I thought OneNote going back, going back 20 years, you know, was way ahead of where a lot of the rest of the market was. But it is, it is time. If you're a Microsoft Enterprise planner, you're a Microsoft Enterprise customer and you're trying to figure out where you should place your bets going forward, to cut to the chase, it is exactly what you just said. It's teams for communication and it's going to be for dynamic collaboration or one off kind of collaboration. It is absolutely, if they are successful, going to be copilot notebooks and then all sorts of intermediate, all sorts of layers that need to be invisible, like SharePoint needs to be invisible, as invisible as the Azure services that are enabling it. Let people focus on the work that they're trying to do together and not 100 different copilot tools.
- [29:58]
C
Let me ask both of you a question. If you are in this new copilot notebook and that's your work environment, that's your collaboration kind of container and now business processes and the looser collaboration logic is now part of the harmonized substrate so that you don't have all these silos. Is this your front end to everything?
- [30:30]
B
I think so. And you know, you think about it from engaging assistance, whether they're human or AI in context. I don't want to go to a separate app. I want to move this customer through the pipeline. I want to do this requisition, I want to hire this individual, whatever. I want to do that in my preferred context. I don't want to have to go all tabbing around to different apps all the time.
- [30:52]
A
So it's a front end to the enterprise, right?
- [30:55]
C
Yes.
- [30:56]
A
You know, that's the point you're trying to make. So if you're the CEO of the firm, and I've used this example before, if you're the CEO of a firm, of a multinational firm and you want to push a button and see and lower the prices of a product in Japan by 5%, say, and then check to find out if the next day, if sales improved, you should be able to do that. I mean, as I've said, the purpose of it is to support the business to the extent that we can make it as frictionless as possible for the business to be able to make decisions and take actions based on changing market conditions, so much the better. And so to the extent that we can make it this virtual front end to the entire enterprise, you know, I think that is, I think that is nirvana, man. We've come a long way, we've come a long way from, you know, it's, it's collaboration is loath and despised to there's hope for the future.
- [31:56]
C
But this is, was one wall that had to fall because there were walls between every enterprise application. There were walls between these collaboration apps and then the general purpose collaboration apps really did, didn't connect all that well to everything else. But to get to that future that we keep talking about, you know, from different perspectives now we're talking about the end user environment for, for, for collaborating, collaborating with agents, for collaborating with other human expertise and you know, and then the more structured processes underneath that.
- [32:39]
A
And I think you had the last word, George. Gentlemen, once again, it's been a real pleasure chatting with you. Thank you and thanks again to our sponsor, Directions on Microsoft. If you need a reliable, independent information source about anything Microsoft Enterprise, whether it's software, services, costs, licensing, cost negotiation, come visit us at directions on Microsoft.com and if you like this podcast from the three of us, hit that thumbs up button. Also, if you have feedback on anything we discussed today, or if there are topics upon which you'd like us to focus in a future episode, please drop a comment in this session's YouTube page and don't forget to subscribe. Thank you all very much and until next time, Sam.