I’m talking with my good friend David Pierce, Vergecast co-host and The Verge’s editor-at-large, about something he spends an ungodly amount of time thinking and writing about: software. Scores of new workplace apps are cropping with clever metaphors to try to make us work differently. Sometimes that works… and sometimes it really, really doesn’t. And it feels like the addition of AI to the mix will accelerate the pace of experimentation here in pretty radical ways. Links: Why software is eating the world | Wall Street Journal (2011) Mailchimp CEO Rania Succar on why email makes sense for Intuit | The Verge Why would anyone make a website in 2023? | The Verge Wix CEO Avishai Abrahami isn’t worried AI will kill the web | The Verge Figma CEO Dylan Field is optimistic about AI | The Verge We don’t sell saddles here | Stewart Butterfield (2014) The CEO of Zoom wants AI clones in meetings | The Verge Dropbox CEO Drew Houston wants you to embrace AI | The Verge Credits: Decoder is a production of The Verge, and part of the Vox Media Podcast Network. Our producers are Kate Cox and Nick Statt. This episode was edited by Xander Adams. Our supervising producer is Liam James. The Decoder music is by Breakmaster Cylinder. Learn more about your ad choices. Visit podcastchoices.com/adchoices
Support for this episode comes from AWS. AWS Generative AI gives you the tools to power your business forward with the security and speed of the world's most experienced cloud.
Hello, and welcome to Decoder. I'm Nilay Patel, editor-in-chief of The Verge, and Decoder is my show about big ideas and other problems. Today, I'm talking to my good friend, David Pearce, who's my co-host on The Verge cast and The Verge's editor at large. And we're talking about something that David spends, honestly, too much time thinking and writing about, software.
Specifically, we're talking about the software you use at work, the stuff you like or maybe just tolerate and use every day, the stuff you probably hate and try to avoid using at all costs, and the stuff you love and hate because your job revolves around using it all day long. It's fair to say that businesses of all kinds changed radically when software entered the office.
That's the foundation of the famous Marc Andreessen quote, software is eating the world. And now it seems like it's all about to change again as AI automates more and more of that software. At least, that's if you believe all the CEOs who have come onto Coder in the last year telling me that's what's about to happen.
These tools are all usually lumped together in a big bucket we call enterprise software, but there are often meaningful overlaps with the popular productivity tools many of us use in our regular lives as well. So first, I wanted David's help in just defining it all. Then I wanted to talk about how it's designed, and how that design shapes how we work every day in subtle and powerful ways.
That's everything from the big familiar bundles like Microsoft Office 365 and Google Workplace all the way down to familiar single tools like Slack. But as you'll hear David explain, we're starting to see scores of new apps crop up to handle very specific use cases built around clever metaphors and interesting new interfaces that try and rewire our brains to make us work differently.
hopefully faster, more efficiently, and lately, more remotely. Sometimes that works, and sometimes it really, really does not. Something is changing about software at work, and I often find the best way to understand the future is to take a moment and consider the present. Okay, David Pierce, Software at Work. Here we go. David Pierce, welcome to Decoder.
Thank you. It's an honor.
Is this your first time on Decoder?
Ironically, I hosted the show before I was a guest on the show. I did a thing with Sean Hollister a while ago when you were off, you know, gallivanting somewhere. But now I get to be here talking to you.
Well, it's interesting because you and I host another show together. But usually when we do other episodes of that show, I'm the guest and you're the host. And now the tables have finally turned. I know. I don't know how I feel about it. Only... The penalty is that you have to talk about enterprise software on this show. I've brought this on myself.
That's very much the difference between Decoder and the Verge cast is on this show, we're going to talk about enterprise software at great length. So you, in addition to hosting the Verge cast with me and being editor-at-large of the Verge, you also write Installer, which is our newsletter where you try all kinds of software.
I believe you do this as a form of therapy because you are addicted to software, which I'm worried about you as your friend. I'm worried that this is a problem that you have. But it means you have tried a lot of software in your life.
I have, I think I've come to see this as less of a procrastination technique of taking all of the tasks that I have in one place and putting them in another place rather than completing any of them because it still feels productive, even though I don't have to do anything as constantly rethinking the way that I do everything, which again, you could argue that's not productive at all.
It's probably not, but I, I have spent probably more time thinking about how I do stuff than than almost anyone I know. I don't know if it's useful. I don't know if it helps. I end up just back in the same apps 400 different times a year, but it is a thing I have spent way too much time and energy on over the years.
The way that I always think about enterprise software and software at work is the Marc Andreessen quote, software is eating the world. There's a lot of ways to interpret that quote. But what I specifically think about is that all of the things you do at work will go from being paper and pencil or sculpting clay models to using software on your laptop, right?
Like fundamentally, all of people's work will be done in software. And then we've had so many CEOs show up on the show and say, the second we start making an investment in software, now we've made a forever investment in software. We're just going to keep spending money on developing software, on software engineers, on maintaining the software, on figuring out our cloud storage bills.
Now we're just doing software all day long. Is that generally right? Is that how you see the growth of enterprise software that we used to have floors of accountants and now we've got floors of people using Excel?
Yeah, and I actually think that is the right way to think about it, right? That it's not that we replaced floors of accountants with nothing or with one magical computer that does all of our accounting for us. We just changed the tools that we use. And in so many ways, the idea of our...
productivity being replaced by machines turns out to not be true nearly as much as it is the tools of our productivity are replaced by machines. And I think that's what we've seen with software, where we all went from adding machines and paper to doing the same, arguably more work with computers. And the same thing is going to happen with AI, right?
Like it wasn't that long ago that everybody was like, oh, because of the industrial revolution and then computers, we won't have to work 40 hours a week anymore. We're all gonna be so efficient with this new technology that we're just gonna work eight hours a week and get the same amount done. Whoops, that didn't happen. But I think that transition is exactly right.
And I think the thing that comes after accountants using Excel is the question.
I'm going to hold you to that because later on we're definitely going to talk about whether AI will cause layoffs. But first, let's start at a high level. I want to try to define some categories of enterprise and productivity software with you. There's enterprise software, which is software that is sold to big businesses, and that can be a lot of different things.
Google Workspace, Microsoft Office, that's classic enterprise software. Everybody needs a word processor at their computer, at their desk. You can just pay per seat. Here's this productivity software. Email clients, whatever that is. Whatever that set of things is one kind of enterprise software. Then there's I don't know, whatever runs your industrial robots. Right. Procurement. There's that.
Then there's productivity software, which consumers use. People plan their weddings in Notion and Trello. You try all this stuff. You play with all this stuff. You cover this industry pretty extensively. Can you draw a definition for us here? How do you think about the different lines?
I think your categorization is almost right. I think I would shift the Microsoft office and Google workspace more into the productivity tool thing. I think the way I've come to see it basically is like there's software you use every day. There's the software you hope you never have to use. And then there's the software you definitely never have to use.
And the software you definitely never have to use is like the stuff that runs the robots, right? There's like a person at your company whose job it is to use that software and no one else will ever touch it. And there's a ton of that software out there. That's everything from like compliance software, which is like vastly big and powerful and important.
And most people should never once in their lives have to encounter compliance software. So that's one side of it. All the way at the other side, you have the productivity tools, right? So that's like the Google docs and Gmail, Microsoft word and Excel and the stuff that people do their jobs inside of.
And then in the middle, there's all the sort of messy stuff that everybody hates the most, like the HR software and the travel booking software and the way to set up IT tickets and all of the things that make a company work, but aren't technically how you do your job in most cases. is that kind of messy middle.
And people hate the software the most, I think in the middle, because it's not the stuff you use every day. So you don't build systems around it. You don't learn the ins and outs of it. You just encounter performance management software four times a year, and you are required to do it, but you don't have to do it long enough to actually care about it.
And so I think those are the three different buckets of it that I've come to see over time. And obviously all of them are gigantic businesses. All of them cater to completely different people. And all of them are at varying levels of invisible to how most people actually want to do their jobs all the time.
Yeah. I feel like a fair warning for the audience here is that David loves software. And my goal in life is to never use software at work. If I could get to the point in my career where people bring me printouts of things and I circle them with Sharpie and send them away, it's the dream. Truly the dream that I work towards every day. I'm farther from that dream with every passing minute.
Yeah. It feels like. Every day we get further from the light of printed things on paper.
It's not gonna happen for me, but it's what I think about because I am the person you're describing who encounters the software and thinks, why is this making me work this way? Why am I learning some metaphor that the expense software wants me to learn just to say I bought dinner for my team? I don't care about any of this. I just need you to pay me back for the dinner I bought for the team.
And it feels like those metaphors are like really what is getting sold over and over again, both in the productivity software zone and then in the your company needs to function zone. And it feels to me very much like you sell the metaphor because that is the promise. That's the exciting thing. Hey, if everybody just worked this way, you'd be so much more productive.
Yes.
Well, and I think if everybody worked this way, it would be great, is the great truism and mistake of the software industry, because it's just impossible. The idea that you're going to get an entire large group of people to all understand a single system and process and tool for getting things done is borderline impossible.
And so if you're a software company, you basically have one of two options. You can either build a piece of software that is so unbelievably specific and opinionated that you literally can't use it any other way than that. Seriously. And this is like a non-crazy way to build software that there is literally only one way to use it. It's not that there's even right or wrong ways.
There's literally only one way to use it. And you can actually get a pretty long way doing that because people will hate it, but they will use it correctly because it's the only way to use it. or what you end up trying to do, and this is, over time, the much more enticing and popular path, but also the path to ruin, is you try to be everything to everyone all the time.
And you say, okay, your boss wants to use this piece of software and they wanna use it this way. That's insane, but your boss is the one who pays for this. And so you have to do it because your boss said so. But what we're gonna give you is a million other buttons and knobs to press and twist so that maybe you'll find the thing that you like.
And the hope is you can build this sort of great denominator software across everything. And that's how you build Microsoft Word. And eventually it gets away from you, right? You build the thing that does everything to everybody. And then all of a sudden it becomes this like overwhelming mess of a piece of software and it all falls apart. but those are your only two moves.
And so like, I've really, I've come to feel for these companies over time because you have IT managers who buy a lot of the software, who want one thing. You have bosses who demand a lot of the software, who want something completely different.
And then you have the people who actually have to use this software all day, every day, who are having this, A, foisted upon them and B, dictated to them how they have to use it. I don't know that it's even possible to build something that makes all three of those groups happy. I certainly have not seen one that seems like it works for everybody.
Let's talk about this in the context of some pieces of software. So it doesn't matter if you run a newsroom like we run at The Verge or you run a small legal practice or you run a Fortune 500 corporation. You've got some set of needs, right? People need to talk to each other, whether it's an email or Slacks. You probably need to make something, hopefully, at your business.
Unless you're just making fraud, in which case, call me. I would love to do an episode of Decoder with you. But, you know, you've got to take something in. You've got to put something out, right? Most businesses operate on this principle.
So you need some software that helps you make the thing that you're making, whether that is the software that's running your 3D printer or, for us, it's the software that makes the website. If you work at a legal practice, it might be Microsoft Word. Some software where you're productive, where you're making something. And then there's just the, I'm operating my business piece of it. Right.
Which is usually Microsoft Excel, right? Like you've got your, I'm doing my financial modeling. I've. HR software, I'm tracking my employees, I'm running my payroll. All that is another set of software. There's a few ways to think about just buying that stuff, right? You can buy a big bundle from some enterprise provider or not.
We have a lot of CEOs come on the show and the CEO of MailChimp was like, what I do is I sell you the software to email your customers. And then I sell you the entire back office solution that runs your website and your billing and everything else as well. That's a CEO of Squarespace said that to us. CEO of Wix said that to us.
They want to set up the website for the yoga studio and then run the entire yoga studio's business so the yoga people can teach the yoga. That's one big bundle of software you can buy. Or you can piece together all of the best in class competitors, the ones that work best for you from all the independent providers and hope they work together. Which approach is winning right now do you think?
It's a strange overlap right now. Actually, there was a long period of time where the bundlers were winning. And you would start, like you said, with one need. And you would say, okay, if you're MailChimp, we're going to make it really easy for you to email a lot of people. And then...
You see the thing that people leave MailChimp for during their day and you're like, well, we could just build that thing. We can build better CRM software so that you don't have to go elsewhere to make the stuff in order to send it to your emails. And then you just sort of slowly build out from there.
And that stuff is starting to win because fundamentally all these companies would like to pay less money for their software and manage fewer things and have fewer contracts to deal with. And the bundling becomes very useful. What is also generally true is that it's very hard to do a lot of things really well simultaneously. Uh, and so, I mean, it's the, the truism of technology is right.
It's bundling and unbundling. And I think we, we were in a really intense bundling phase for a really long time. And then the pandemic happened and we unbundled so fast because all of a sudden everybody was at home. Everybody had new software needs and everybody needed them like tomorrow. And so everybody went like literally tomorrow.
It was like, oh, I'm not going to the office again for four years. I need new kinds of software. And so everybody just went out and looked for, okay, we need a new way to do collaborative design. We need a new way to do video calls. And we can't wait for someone else to attach it to the contract that we have. We're just going to go sign up for the thing that everybody likes that's the best.
That's why Zoom got huge, even though other video tools existed, right? That's why things like Figma started to take off, because people just needed these new workflows. And so there was this massive rush towards these so-called best-of-breed things, and Then what obviously started to happen is the big companies started to tack that stuff on Microsoft builds teams, Google invests in meat.
And so zoom becomes less competitive in its own way. But what's wild to me is the, this company Okta, which does basically like log in and a bunch of other stuff, like you're talking about, they started as a way to log into services. And now there's like a whole giant backend authentication stuff that they do. Uh, they do this really good serve way called businesses at work every year. And.
The wildest thing in, I believe it was the 2023 one, was that half the people who pay for Microsoft Office also pay for Google Workspace. And half the people who pay for Microsoft Office also pay for Zoom. And so we're in this place where...
We are very much in the sense of like, you sort of need the basic bundle, but then everybody is still in this moment, willing to spend the money on at least like a handful of the tools that work around it. So right now, Microsoft is trying to keep building the walls around its bundle.
And you're seeing the same from a company like Salesforce, which bought Slack so that you spend more time in the Salesforce universe. But these individual apps that exist around them, in many cases, are just so much better that no one else has been able to catch up to them yet. Meet is better. Teams is better. People still really like Zoom. And I don't think that's changing anytime soon.
One of the things that's really interesting about your description of the timeline there is all of the new apps are necessarily in the cloud.
Mm-hmm.
Everybody went home. We all have to work online together. Necessarily, we're going to work in web browsers in a collaborative environment together. And particularly on desktop, there's no action in native apps. There's still a little bit of action in native apps on mobile for a variety of reasons that we've talked to lots of CEOs about on the show. Dylan Field, CEO of Figma.
He's like, I got to build it natively on iOS because the browser isn't good enough to let me do Figma on the browser on iOS. The second you open your MacBook, the browsers are good enough. And he's like, this is where I should deploy an application. This is where I get immediate distribution.
The idea that all of these productivity apps work in the cloud so we can all collaborate together and all the data is stored centrally and you don't have to sync it, that seems like the end of the road for the desktop productivity application. You write installer.
Do you see any glimmers that some huge, powerful Mac app is going to show up or some huge, powerful Windows app is going to show up, run locally on your computer and revolutionize the business world?
Okay, I'm going to say this, and I should preface by saying I have very little quantitative evidence for it, but it is starting to be a thing I hear more and more in talking to folks who make this stuff. I think it's coming back.
You do?
I really do. And I think, again, this goes back to these sort of divergent needs of two group of people, right? You have... The stuff in the cloud is easier to manage in so many ways, right? It's easier to understand who has it where. It's easier to provision on different people's computers. It's easier to manage. You know where your data is. Like from a corporate management standpoint,
It's just better in so many ways to have stuff that is based in the cloud. Except, I don't know if you've noticed this, but everything seems to be down all the time now. Everybody is getting hacked all the time. And suddenly, the idea that the cloud is actually a safer, saner, simpler place for your data, I don't think feels true to people in the way that it once did.
And all of that, notwithstanding questions about, is my stuff being trained for AI purposes? Do I want the AI stuff that is coming to all of these things? I think... you're starting to see this shift towards what some people in this industry call local first software.
And basically it's this huge sort of re-architecting of the way that we think about files really, away from the idea that like a file lives in the cloud and you and I both open an app and act on that file together. That file can live on my computer as an offline local file that is mine in like a real meaningful way. But still, you and I can act on it together in a collaborative way.
That's possible to do. It's just much harder software, and we're only getting good at building that stuff now. But I think it's stuff that works offline, which, like, God help you if you're on a plane or in bad Wi-Fi or anything. Like, it's so hard to do anything online. without the internet now. And I think people notice that.
And we're coming back to this idea of like, I want something that feels fast, which means it has to be local. I want something that is mine in a real way, like it's a file I can see on my computer. I think that stuff is coming back and it matters to people. We're going to see that best of both worlds where we get the collaborative thing, we get the cloud access, we get the cross-platform stuff.
It's manageable by your organization, but But it has the same feeling of like, I can just open a thing on my computer and it works.
It feels like that was the core way of thinking about things before broadband internet. Yeah. Like Apple had entire software suite called iSync. Yeah, that's right. Do you remember? No one ever used it because it was so complicated and fiddly. And then the cloud existed. And then you could just sync everything over iCloud. And so iSync went away.
But the idea that you would maintain sync on files across devices in a fairly manual way was built because the internet connections were not fast enough. And now you're saying the people don't trust the cloud enough. So we're going back to that model.
You could sort of argue that we skipped the correct middle step there, right? We're like, okay, doing the thing where I email you a file that says David presentation one, and then you email it back and it says David presentation two, and then final, final, final dot PPT. Like that sucks. We can all agree that that is bad.
But we skipped over how do we make it easy for people to share things with each other all the way to what if you didn't own any of your data? And that we sort of left a good middle ground out there. And I think I would not... bet everything that it's going to win.
But I think there's going to be a push back towards the idea of like local on device, but still useful and in sync and collaborative stuff coming back.
That's a good place to take a break. I want to come back and I want to talk to you maybe just about Slack for the rest of the episode. Oh boy. And put all these ideas into that little case study. We'll be right back.
I'm back with Verge Editor-at-Large, David Pierce, talking about the evolution of workplace software.
David, we just talked about kind of a lot of ideas all at once. The idea that everything would be in the cloud, the idea that everybody on the team had to buy into a way of working to make a lot of this productivity software actually functional in an organization.
The idea that the big players were bundling capabilities, but you would still have some like renegade teams inside your company using the tool they want. And then, I don't know if people caught this, you brought up the idea that Salesforce bought Slack.
I was hoping nobody would notice.
So I just want to put this into focus around Slack in particular because it's just a good case study. Slack just turned 10 years old. You and I have covered the hell out of Slack. I've interviewed Stuart Butterfield, maybe not on the Coder but on the Vergecast for sure. He was the founder of Slack. Slack was supposed to be a revolution 10 years ago, right?
It was going to replace email and then they kind of walked back from the idea that they were replacing email and embraced Slack. replacing email again. Then Microsoft showed up with Teams, which is mostly a video conferencing platform, but they added some Slack-like features and that. Everybody has Microsoft Office, so Slack got out-competed in that way.
They filed antitrust lawsuits in Europe, and then Salesforce just buys Slack. And like I was saying, the idea is you're going to use Slack. Why not use our CRM solution as well? Why shouldn't Slack be the CRM solution that you are using? You're talking to your team about selling stuff. You're probably using Salesforce anyway.
Now you can just chat to it through whatever Mark Benioff AI system that is being integrated in Salesforce. Yeah. I watched Dreamforce. I couldn't tell you what's going on, but that's basically the idea. Inside of that is the first thing we talked about, which is the whole point of Slack was that everybody would buy into a new kind of metaphor for working.
And that would somehow change your company and make you more efficient. And actually it turns out that metaphor maybe didn't take and everyone just is like spamming each other with text messages.
Yeah, I think one of my great theories about Slack is that everybody uses Slack wrong. Like everybody uses Slack wrong. And I think if you rewind all the way back to that initial launch of Slack, Stuart Butterfield wrote this really great blog post. It's called something like we don't sell saddles here.
And he basically outlines like the whole vision for what Slack is going for and makes the argument that the biggest problem is that most people don't understand what they need Slack for. And so Slack's job is not only to convince you that you need this, but that you need something like this in the first place. And their whole idea was basically to be a search engine for all your stuff at work.
If you really want to boil it down, the idea behind Slack was not to be a chat app. It was to just be a place to put all your stuff. And all your stuff is communication, it's files, it's all these things that sort of you accumulate during the course of a workday. And by having that all in one place that's accessible, you could build something really powerful.
Super cool idea has nothing to do with what Slack is and how it's used, right? We use Slack as an email replacement, right? I think Stuart was serious when he said they weren't trying to kill email. They were trying to like subsume it inside of a different system, but they were always going to be like, look, if you want to send email to talk to each other, that's fine.
But like when you need to share files, do it in Slack. I think that is much closer to what Slack wanted to be than what it has become. And what it became is this awful system engagement bait app that we all spend way too much time in and never actually get any work done inside of.
So I will note that the background in for Slack is searchable log of all communications and knowledge. Yeah. Whether or not you start with the name or you end up with the name or you reverse engineer the name, that was the acronym. And the idea was that you would operate Slack at the pace of email. It wouldn't replace email, but you're talking, you're making decisions in Slack.
You have a history of all of the files and figures and conversations around those decisions. And then you could just go refer to it. You could find it. It's all happening there. And what ended up really happening is everyone was just chatting in real time in like a fractally expanding number of Slack channels.
In the service of this being a case study, where do you think the breakdown between the metaphors that Slack wanted people to use and the actual user behavior came from?
It's millennials and it's text boxes. There's this long sort of mythological story about Kinja, which was the old CMS that they used to run at Gawker.
And I have no idea if this is true, but it is a story I heard that I spiritually completely believe, which is that they ran an experiment at Gawker where if they changed the size of the default text box in Kinja, it would change how long people wrote. If you give people a lot of space, make the text box big, They're gonna write a lot in it. You fill the thing.
If you make it one line, people are gonna write one line and hit enter, right? Slack made it one line. It looks like a thing where you would send text messages. It doesn't look like an email inbox. You don't have subjects. You don't have two lines. You don't have a thing for a signature. Like, can you imagine if you had a signature at the bottom of every Slack message? You'd look like a lunatic.
And so all of these like product incentives taught you to do it really fast. The main thing there was a single line of text and everybody else uses a single line of text in messaging apps. And so we all just treated this thing like messaging apps.
And it was a bunch of millennials who came into the workforce and are used to the type of thought, press enter, type another thought, press enter, type another thought, press enter. If you get the text that is two paragraphs from somebody, either somebody died or you're being broken up with.
Those are the two options.
And instead, we just trained everyone to talk in work like they talk to their friends in text. And culturally, I think that is really interesting and complicated, but it immediately broke that paradigm that you're talking about, which is Slack even has always said that Slack uses Slack in a much more considered way. People write much longer things.
It's designed less for like minute to minute updates of what you're doing all day and more for like a, here's what I accomplished during the day. Again, so that someone can go find it later. You're not expected at Slack to be in lots of rooms. There is a norm that if you need somebody, you mention them. And Slack set up all these rules, but didn't bake any of it into the product.
They built a product that looks like text messaging. And so people used it like text messaging. And Slack has now spent the last decade being like, no, no, no, no, no, no. It's all about rules. You have to set up norms. You have to teach each other how to use the app. And it's like, no, stop building me a text messaging app.
When Stuart was still running Slack and he came on the show, he talked a lot about how they onboarded people into Slack and taught them how to use Slack. And they spent a lot of time teaching new Slack employees how to use Slack. I had the same reaction as you, which is why doesn't the product teach people how to use Slack? And it's kind of all the way back around to the first thing you said.
You can either build a very opinionated product that works extraordinarily well for a small number of people, or you can just take all the training wheels off and let people sort of build whatever product they want. and so that they can work however they want.
And that's how you get, even at Microsoft Excel, they can't take any features out of Microsoft Excel and point people to a simpler version of it because every 5% of users is millions upon millions of people. Slack is in that zone. A lot of other products are in that zone. Is it just we have to go through other kinds of fads?
We all used IRC and email, so then we used HipChat for five minutes, then we used Campfire, and now we use Slack, and maybe one day Slack will just fully give way to Teams or whatever?
Maybe. I mean, I think to be clear, the idea that you should spend a lot of time training people how to use your most important products when they join your company is really good and no one does it. Like you should have to spend a day learning how your company communicates when you join a company. That stuff matters. And we just like hand people Slack and they're like, we use Slack.
And that's as much as you get. Like, do you know how to use Slack? It's like, sure. But I don't know how to use Slack the way that you use Slack, right? And that stuff matters a lot. I talked to this woman, Laura Martin, who is like Google's productivity guru. And one of the things she recommends to everybody is like, just muck around in the settings.
Just go spend 10 minutes clicking all the buttons in the settings just to see how the thing works. And like even one extra tick of understanding of the software goes an incredibly long way. But if you're Slack and you both use this product and run your company on it, the one does not absolve the other, right? You can and should train people on your company's best practices for Slack.
That doesn't mean that everyone else should figure out your best practices just kind of by osmosis. That's not how it works. overwhelmingly the question is, are we ever going to go too far and swing the pendulum back? Because so far there's really no evidence that we're going to, right?
There are a lot of people who complain that everything is moving too fast and we're all too attached to these messaging systems and we can't keep up with everything and it's making us crazy and we all spend time in Slack instead of actually doing our jobs. And there's just no indication that that's going back.
You could actually make the case that email represents a much saner, more functional way to run a business than inside of Slack. But like, do you want to go back to email? I don't really want to go back to email. And so far, all we do is faster and bigger and more full of stuff. And we just kind of rely on people to solve it for themselves.
And either we're going to get a set of really clever products that does that for us without killing the idea that people are being productive because... You have to make it look like everyone is productive. And one thing Slack is very good at is making itself look really engaged.
And when it looks lively all the time, which makes managers and bosses feel like things are happening and that feels good. Whereas if you don't hear from anybody for eight hours, you're like, is everybody working? What's everybody doing?
And so squaring those two things is really hard, but I think the only way we get off of this road to insanity that we're on with a lot of these communication tools is to find a way to do both of those things. And I'm honestly not sure what that looks like.
Let me ask you about switching costs in that context for one second. We can keep picking on Slack, and I think both of us are motivated to do it. I would like to, yes. It's a little unfair, but we both use it all day long. But if I was to say, okay, the Verge isn't using Slack anymore. We're switching back to email. The switching cost of that would be almost impossibly high.
I think we would just have a straight mutiny. We would not know how to work, quite honestly. If I showed up and I said, we're going to stop using Concur. We're going to use some other enterprise provider for expenses and travel. I don't think our team would care a lot. but it almost would be harder to switch because no one cares a lot. Like I could bring down the dollars by some amount, 10%, 20%.
Great, that's a good reason to switch. But then everyone would, I would lose that in productivity because no one would pay attention to the email. that we're switching out Flims for Flams or whatever we're doing. And then they would try to file an expense. They would use the old app. And then we would just be sort of running the individual training all the time. Is that something you can fix?
Is that what you just get the bundle for? Like, screw it, Microsoft, just do it all for me. Is that the opportunity for any of these companies is to make an expense software that everyone loves so much that everyone actually pays attention to the emails?
It's one really good reason to keep building your bundle, right? If I have an expense product that you hate and someone else has an expense product that you hate that you actually already pay for, I can't win that fight, right? That is the great challenge. And so if you're trying to get into that world, you either need to be cheaper to make the people who actually pay for this stuff happy
You either need to be like vastly better. Like I think even being slightly better doesn't get you there. You have to be like orders of magnitude better. Or you have to do something else that the existing apps don't do. Like I think one really good example is in the expenses world.
I wouldn't swear to this, but I think it's true that Expensify was the first company to really do a good job of letting you take a picture of a receipt and turn it into an expense and upload it.
unbelievable like the greatest thing that ever happened to the product of tracking expenses was being able to take a picture of a receipt and have it like OCR out all the relevant stuff and turn it into expense I think it was Expensify who did it I don't know who it was for sure whoever it was congratulations you you win and they did and that like Expensify became hugely successful based on having like a good mobile app right that's a thing when you can add something meaningful to that that goes a really long way for those kind of
middle tier products you encounter but don't really care about all that much. There's just not that much surface area to do that. But then the stuff that people use every day, the switching cost is so much higher because it's so much more entrenched. Trying to get someone to stop using an app that they hate but have used for 10 years is so, so hard.
even if they hate it, they will tell you every single day that they hate it. And you'll say, here's another one. It's better. And they'll say, Oh God, because people don't care. Most people have jobs to do, right? Like most people do not use software for a living. They, they use software as little as possible so that they can go do the thing that they do for a living.
And I feel like that is the exact right balance for those people. And software gets that wrong where instead they're like, here's more stuff for you to do. Come switch to our software. And people are like,
Why haven't Microsoft and Google just run away with this? Is it antitrust litigation? Is it that they're not good at everything they try to do? Is it Google kills its products too fast?
I mean, to some extent, they have run away with this, right? I'm convinced that Excel is the single sickiest piece of software on earth. You see it and hear it all the time that... Like the companies I was mentioning that pay for Microsoft Office, but also pay for Zoom or Workspace or whatever. That's all Excel, man. That is all Excel. Like you can build good presentation software.
You can build good document editing. No one can beat Excel. It is just not possible. And who pays for this stuff is the people who use Excel all day. Yeah. Excel is so sticky. And the problem is, yes, it's very hard to do all of these things really well. It's also just a matter of focus and resources.
For Microsoft, I don't know that adding one more tiny piece of software makes your company more likely to sign up for it. Microsoft is going to keep doing that stuff. But at some point, Office is pretty sticky already. And I think for Microsoft to spend the resources and time and energy to add more stuff is actually a pretty... big bet for that company to make.
It's why teams was such a big deal, right? That Microsoft saw all of a sudden pandemic starts. The idea of, I think video chat in particular was like an existential shift in how we communicate at work. It had the potential to just like destroy outlook and become the center of everything. And zoom immediately sets out to build a whole office suite, right? That was the thing zoom was going to do.
They were like, we're going to make video into everything. We're going to start doing yoga classes over zoom, but we're also going to build like Zoom mail and Zoom docs. And that became the center of how people worked. And Microsoft goes, oh God, maybe that is going to be the center of how people work and built teams and crushed it because ultimately the free thing usually wins.
We need to take another quick break. We'll be right back.
Support for this episode comes from AWS. AWS Generative AI gives you the tools to power your business forward with the security and speed of the world's most experienced cloud.
I'm back with Verge Editor-at-Large David Pierce discussing the next big thing in enterprise software and if it will be as big of a change as software is eating the world. The last turn I want to talk about, and I'm very curious if you've actually seen any features here that work or are valuable. I already know what you're about to ask. It's obviously AI.
So the dream here, and you mentioned Zoom, the CEO of Zoom, Eric Yan, is on the show, and his dream is that we'll make an AI agent out of you with everything Zoom knows out of you, and your AI agents, potentially thousands of them, will go off and have meetings on your behalf while you sit on the beach. I want to be 100% clear. This is a real thing he said on the show.
This is the thing he wants to build. I don't know about that. That is as gentle of an evaluation of that idea as I can give. Somewhere underneath that is I work at a company where a lot of people have made a lot of decisions. I don't know what those decisions are. There might be some wiki and some horrible piece of enterprise wiki software that I don't really know how to access.
And my Okta's acting up and I just have, whatever, I'm not going to do it. This Slack room is too busy. I have no idea what decisions are going to be made here. I'm afraid of saying the wrong thing. All this is a mess. I'm just going to ask some all-knowing enterprise AI, hey, what's our history of sales in this region and who is our biggest client?
And it will spin through all the company's data and answer the question and tell me how to make the next move. That's the dream. Is any of that a reality yet?
Uh, no, I think it's, it is plausible in a way that I find a lot of AI stuff really implausible. Because in a real way, most of the data that you need exists, right? Like companies have that data somewhere. It is written. Company handbooks get written down. These things are placed somewhere. The problem is that they are in thousands of places, right?
Again, I was just rereading this Okta thing ahead of our recording here. And the average company in the United States pays for, I think it was 110 different pieces of enterprise software. That's too many. And what it means is if you want to find a thing, God help you. Like, where do you even start?
And so what you've seen, I think, is the beginning of what's about to be another big rebundling as a result of that. Because that thing you just described is super enticing. There are studies everywhere that say we spend a huge amount of our time at work just looking for stuff. And that actually easy access to information would be like the greatest productivity enabler in modern history. And
The idea that I could just say, what is our sales history with them? Or like, who's the contact over there and get that stuff quickly. Incredible. And so every company is after this, right? Like Dropbox built this thing dash that searches across all your stuff and all your different apps. You see companies like notion, which are trying to do more and more stuff.
They're building out new features just in service of getting all that information inside of the app. so that you can query it with AI. Microsoft is doing it with Copilot. Google is doing it with Gemini.
For the first time in a while, being the bundle is more valuable than the sum of its parts in a way that isn't just sort of purely about contract values and it's easier to have one relationship instead of several. Now, your products can all be better because you control more of them.
And I think what we're going to see as a result is this rush back to, I don't want to have 50 best in class apps that don't talk to each other. I want to have six things that do all 50 of those things, even if they're not as good, but because stitching them all together makes them more valuable to me. one of the things that AI is actually good at is summarization, right?
So you can say things like, not just summarize this email for me, but like in Slack's AI, the thing they're trying to build is you should just be able to be like, what happened today? And it'll be able to tell you like, here are the things people were talking about. Here's the file everybody was sharing around.
Like that stuff is meaningful and goes a long way towards making Slack not like a hell hole of one line communication. So this stuff is coming and I think it's gonna be really powerful. But it only works if it's all everywhere, because this is one of those things that if you solve 70% of my problem, you've solved none of my problem. And getting there is going to take a minute.
But I think the push towards that has already started in a pretty big way.
How do you square that with, hey, there's a big push to run more software locally? Hey, people are very cognizant of the amount of data they're giving up to centralized AI providers or centralized cloud providers.
Hey, I'm just generally uncomfortable with my law practice, my medical practice, having all of its data taken up into the cloud where I'm no longer in control of my client data or my patient data.
I'm hearing a lot of excitement from companies about open source models. I think right now you look at the open AIs and the anthropics of the world, and even like what Google is doing with Gemini. And again, we're in this phase of everything is humongous and in the cloud, but eventually like. these models will run on your device. They'll be able to run locally on your own instance.
Like what I don't think is going back is I think AWS and Azure and Google Cloud are gonna be fine. I think like on-prem servers are not coming back, but at least you're gonna be able to exert some control inside of that, right? So like the AI systems will be something run by your company inside of your company, more over time than they currently are.
And I think it sort of jives with that whole idea of like, I want to have all the conveniences of all of my stuff being online and accessible everywhere and sort of functionally managed by somebody else. But I also want that control that this data is mine. I know where it's going. I know how it's being used and I know where to find it.
And that balance has been really hard to strike over the years. But I think especially as these AI models from Meta with Lama and others get better and faster and especially like cheaper and simpler and more local, like my company approved laptop is going to be more important than ever.
The big idea that we started talking about is the Andreessen quote, software is eating the world. All these businesses are software businesses or they're going to operate on software. That is very clearly happened. Most people show up at work, they get handed a iPhone that runs their enterprise software. They get handed a laptop that runs their enterprise software.
No matter who you are, it's all just happening. It's all just happening inside of software. And then some other actual work happens somewhere else. I think we can agree for the purposes of this Dakota episode that is almost complete. Like that transition is complete in the workforce. Is the next turn AI that's actually going to redefine how we do work?
Because the AI companies are absolutely betting that they can pay off all this investment with that level of change and turnover and upgrade and investment.
Yeah. And not only are the AI companies betting on it, like how many CEOs do you think you've had on this show this year who have not mentioned AI once? I honestly, I would be shocked if it's more than one who isn't out here talking about AI.
Because again, A, you like have to have a strategy or everybody thinks you're an old fusty company that nobody cares about anymore and your board fires you and that's a whole thing. Yeah. But also, again, the promise is so huge, right? The idea that AI can remove a lot of the busy work that we have to do.
If you turn making a deck out of an Excel spreadsheet into just a command to an AI, like, that's a measurable productivity increase in the world. And the same thing with the finding of the information and even some of the generative stuff that's starting to happen, like... If that stuff works, and that is a bit, like I cannot emphasize the if enough there.
It's not when, it's if in a very real way. But if that stuff all works, it sincerely changes the way we do just about everything. And it means you and I, we probably won't go to the beach more often, but we get to spend more of our time doing things that are like interesting and rewarding and valuable rather than busy work. Like I spend a lot of my time doing busy work.
You don't because you just have people put printouts on your desk for you to circle things on. But like life is busy work. And if we can get rid of that, there are all kinds of really complicated societal implications to that because frankly, a lot of people's job is busy work and a lot of people do busy work for a living.
But the question of what could this mean if it hits, if it works, it's just too big for anybody to not try.
Is that how it's going to get sold? This will make your company more productive. So you might as well just let digital God make a bunch of decks for you.
Yeah. The alternative is this is going to make you more productive so you can fire half your staff. And like a lot of these companies don't want to say that even though that is both what they're buying and what they're selling is I can fire a bunch of people and replace them with this software.
I think in the near term, the actual real thing that's going to happen is it's just going to automate away a bunch of work, right? Like, I think what we're looking at is something much closer to the advent of computers when all of a sudden, instead of sitting at my desk writing out spreadsheets by hand, I could just sit there and
type one, and when I changed the number, the other numbers would change. Huge societal revolution probably cost a lot of jobs, but was not the like be all and end all of society forever, right? I think we're much more at that era now, where all of a sudden we are about to have this real sort of step change in what technology lets us do very quickly.
And then the question is, what are we going to do with the time that suddenly appears as a result? And with computers, it was we're going to make ourselves busier talking to each other and looking at TikTok. What it will be with AI, I think, remains to be seen.
All right, David, we're going to have to have you back soon. Thanks so much for being on the Cutter. Thank you. I'd like to thank David for taking the time to join Decoder. And thank you for listening. I hope you enjoyed it. If you'd like to let us know what you thought about this episode or really anything at all, drop us a line. You can email us at decoderattheverge.com.
You really do read all the emails. Or you can hit me up directly on threads. I'm at Reckless1280. We also have a TikTok. Check it out. It's at DecoderPod. It's a lot of fun. If you like Decoder, please share it with your friends and subscribe wherever you get your podcasts. Decoder is a production of The Verge and part of the Vox Media Podcast Network. Our producers are Kate Cox and Nick Statt.
This episode was edited by Xander Adams. Our supervising producer is Liam James. The decoder music is by Breakmaster Cylinder. We'll see you next time.
Support for this episode comes from AWS. With the power of AWS Generative AI, teams can get relevant, fast answers to pressing questions and use data to drive real results. Power your business and generate real impact with the most experienced cloud.