Salesforce, Inc. (CRM) September 2026 Conference Transcript

Hey, good afternoon, and welcome to the Salesforce session at Goldman Sachs Communicopia. I'm Gabriella Borges, delighted to have on stage with me Bill Patterson, President and Chief Commercial Officer, Rob Seaman, EVP and GM of Slack. Gentlemen, thank you for your time.

Thank you for having us today.

So, Bill, I want to start out with a little bit of the mic drop moment from two weeks ago now with Dario and Mark on CNBC talking about the launch of Quarterforce and it reminds me a little bit of when we were chatting with the management team earlier this year they were telling us about the Anthropic Saster video where Anthropic talks about the cornerstone of their front office stack being Salesforce and all of the different pieces that they built on top of it tell us a little bit about how that relationship came together and it sounds like you've essentially commercialized what would have taken a fair number of engineering resources to get off the ground. So we'd love to hear a little bit more.

Yeah, I think the – thank you for recognizing that it was two weeks ago, which seems like a decade ago now. But, you know, I think what's beautiful about the world of, you know, kind of Claude and the world of Salesforce is our customers were naturally sort of pulling us, you know, together to work together more effectively. And, you know, Salesforce is a huge Anthropic customer, and Anthropic is a huge Salesforce customer, So there's just natural synergies about these two technologies sort of working together in unison to drive ultimately better quality agents, better productivity for every employee, and sort of just better synergy on the product side. So it's always beautiful when, you know, like we always say, like the number one AI platform and the number one CRM platform can come together because we can actually, you know, really validate a lot of like the core use cases that we serve in the common roles with one another in that sense. So I think it was more than just an announcement. It was a product that was already being built, already being utilized, already being activated, already being deployed in almost every customer that we sort of see today, and it was just a natural fit for these two product organizations to sort of now work together and deliver product, which we now call Cloudforce. So I think this is just a classic example of kind of two magical things coming together and delivering great kind of impact for our users, and we're very, very excited about the early days there.

Was there anything that you've learned in the last two weeks since making the announcement that was unexpected or that you hadn't predicted?

Not unexpected, but I think that, you know, the market and sort of the customer community probably breaks down into three tranches. One, there's a lot of just, you know, heavy early adopters that were really doing a lot of innovation and building a lot of these connections and integrations already themselves that validated the pattern for why we wanted to create a product in this area. And so I think as part of that early adopter community, you got a lot of great feedback, a lot of validation, a lot of sort of conviction about the magic of these sort of two products becoming a strong partnership with one another. I think as you sort of fast forward to the second tranche or second community of users, there's still those users today that are trying to figure out how to make this work and how to make it scale and how to make it economically viable for organizations of their size. Not every small business has an endless token budget, if you will. And so really making sure that the economics work to drive the productivity and the performance of this offering is probably that second community of customer that now needs a little bit more knowledge and know-how to make it work and make it work right. And I think there's a third community that has a lot of skepticism around sort of what does AI mean for them. And as you travel around the world in different communities, different geographies, not here in San Francisco per se, but as you go to other countries around the world, there is a little bit more AI hesitancy. And so that is a final community that does need to be convinced that, you know, this is not a destructive force. It's actually a force to, you know, fundamentally transform work. So I think as you take a long view around what the cloud first opportunity means, the early adopters are going to deploy it fast. Those that are sort of trying to figure out the economics and the true math of how this works at scale I think are still needing some convincing. And then that last tranche is an area that we have more responsibility as leaders to show them that this is a strong sort of solution for their future and not, like I said, that cannibalistic force for eroding work as we know it.

I want to ask you about two of the bear cases that we've heard in the last two weeks. The first is the Fox and the Hen House bear case, which is, look, tech partnerships come and go. They break down all the time. And essentially through Cloudforce, Anthropic is now getting a foothold in the front office stacks of all of Salesforce's install base. How do you think about the risk that however many years from now Anthropic turns around and says, thank you, Salesforce. We're happy to take it from here.

Yeah, look, I think that the fundamental mission of both organizations is different from one another. You know, the mission of Anthropic is, you know, really to prove, you know, that, you know, kind of AI can have a really ubiquitous sort of, you know, kind of moment in our lives and our businesses. And the role of sort of Salesforce is fundamentally to, you know, kind of make businesses better businesses. So we have some relationship in that sense that we all want to help the customers that we serve. But, you know, in terms of the mission to, of what they innovate on and what we innovate on, they're very, very different technologies, very different pace of sort of innovation and also better, you know, sort of just different responsibilities of what we need customers to sort of know, learn how to do and manage, you know, to be successful with these offerings. There's been all kinds of new innovation that, you know, has always been the death of a CRM system at some point in time. The first one was like a website, you know, oh my God, that's going to, you know, eliminate how we do, you know, kind of marketing and marketing outreach. No, it didn't transform, it didn't, you know, do away with CRM, it just, you know, transformed CRM to be a world of, you know, outside-in engagement, just as much as it's been about inside-out sort of operational responsibilities. So I think it's, there's a long list of companies and a long list of technologies that were trying to kill the CRM business. We're still here. We're doing well.

The second bare case, and we can bring Robin to the conversation naturally with this one, is the idea that Salesforce is essentially giving up the user interface. So the end customer now interacts with Claude. Claude is their first port of call. And Salesforce essentially becomes such an adela CRUD database argument from 18 months ago now. How would you respond to that particular UI bear case?

I'd go about it from the perspective, one, I don't agree with it. I'd start with that. So I think minimizing Salesforce or CRM in general to a database, I think, is somewhat glib and a bit of an oversight into what's actually happened in CRM systems. Like, so much of what people have done is express how their company operates in the form of metadata in Salesforce and their CRM. And that, candidly, just isn't very replicable. And that's what people are excited about accessing through other services like Cloudforce or Slack or whatever it might be. Now, on the argument about giving up the user interface, while you may be giving up a bit on the amount of time people spend in the interface, you're actually increasing the amount of work that's happening through the services that support the interfaces. And to me, that's the bigger bet, is that ultimately more work gets done through a CRM system. And, like, if you look at our own internal data on our usage of Salesforce, our usage of Salesforce has skyrocketed within Salesforce. But the overwhelming majority of it's happening through Slack and through Clot. And so the amount of work that's actually happening and the quality of work that's happening has dramatically skyrocketed, and our productivity has improved, as you've seen through our financial metrics. But it's all because we've moved to this headless manner in a way, I think, that tilts the umbrella and makes it a lot more convenient for users to get the work done.

Tell us a little bit more. You've talked about Slack as the operating system for work. What does that look like? What does it mean for your R&D roadmap to now be investing in Slack, not just as a communications tool, but as this essential gateway to all of the underlying system of record?

So the biggest asset we have at Slack is the aggregated attention of knowledge workers. So the average Slack user actually has the Slack app open 10 hours a day and actively uses it two hours a day. So if it were a consumer app, it would be a great place to put ads, but we will never put ads in Slack, so we do not expect that to come from us. But it becomes an amazing place to distribute your software. So B2B software is increasingly being distributed through Slack as agents are. Now, when you increase the amount of stuff that are coming at humans, humans can become distracted, and that defeats the purpose of that attention aggregation. And so what we think, the role that we can play, if you go back to original operating systems and what they did they basically abstracted or hid the complexity of the hardware from the end user and typically had a user interface paradigm about it that's effectively what slack does for a litany of apps and agents that are now coming at people in addition to the humans that are reaching out to them so from an operating system perspective we think we can basically provide that overall design guideline and system that absent agents can adhere to we can abstract the complexity of the fact that there are thousands of them now competing for people's attention and make it a single place that people can kind of orchestrate their day and be a cockpit of and so I think we ultimately

become where the AI stack starts and stops you gave an example there of where it's been successful within the own Salesforce organization and I imagine you have customers that you would classify its top quartile in terms of early adopters of slack functionality my question for you is what is the alternative for the customers that maybe we separate out the cohort that's still hesitant. So the customers that are leaning into AI that are not using Slack as one of their primary user interfaces, what does that look like? What's the alternative?

I mean, the alternative is alternatives from competing products that I think are functionally deficient. And so I think that a key thing about Slack, and I think you see this play out in all the major AI companies that use Slack, and I apologize in advance if I'm saying something to whoever if I'm saying something that I'm not supposed to. But if you look at Anthropic, you look at OpenAI, you look at Vercel, all of these frontier labs and AI-first companies operate their Slack in a very unique way, which is they operate totally in public. I was talking to Boris from Anthropic yesterday. It was so fascinating. Every single Anthropic employee has a notebook channel. That's basically a channel with them, themselves, where they just think out loud through the course of the day. They do this. OpenAI does this. Vercel does this. We do it at Slack. And what that does is that radically flattens an organization and immediately disseminates all information from that person to every other person in the company, but importantly to all the other agents. And that model in Slack and the way that the software actually encourages people to participate in a very social way and a very public way is not something you see happen in the alternatives that are out there today. Full stop. I would challenge anybody to go talk to a customer that's using an alternative product and ask them how they do that. and they'll just stare at you like you have ten heads.

Rob, you've got to stay on this. This is a really interesting rabbit hole. So what is Anthropic doing with Slack that allows them to power this army of agents behind the interface? Tell us a little bit more about this.

They do as much as humanly possible in public. So every single product channel, feature channel is public for the entire company to access and all the agents to access. Every single person then has what we call a notebook or brain channel. Vercel does a similar thing. if you talk to G from Bursell, they have a brain channel and they just think out loud in these brain channels. And so everybody in the company can see what G is thinking. And every agent that's in your Slack can see what G is thinking because it's public. And I think increasingly, if you look at what Toby Luka, I don't know if I'm pronouncing that right, the CEO of Shopify, he wrote a piece on Twitter called Learning on the Shop Floor. And it was about using their river agent in Slack but only doing so in public. And what's interesting is they've deployed their coding agent, River, which is built on top of another harness, but they've deployed it in Slack and only allow it to be used in public channels in Slack. And so what you're doing there is you're effectively immediately disseminating to the entire company every single thing that's being built. So everybody can just search and be like, is somebody else working on this yet? So you radically reduce duplication of effort, but then you also learn from each other. Like the bear case to me on these individual contributor AI tools is that the, and Toby says this as well, is that the AI learns and the person using it learns, but their peer doesn't learn and then the company doesn't learn from it, right? But you move that into public, like in a channel in Slack, and then everybody else can learn from it. And it immediately becomes a knowledge artifact for your employees, but also for other agents in your Slack.

Okay, so you've already given us one of the answers to this question on why Slack is the natural place for that activity to happen, which is it's already open 10 hours a day, it's active two hours a day. What else is in the IP of Slack that you think allows that use case to flourish?

I mean, something comes down to how we work. People are, so one of the things we've traditionally said about Slack behind the scenes is our enduring competitive advantage is going to be the quality of our user experience and the customer love that we generate from that user experience. And the fact of that is, like, we have a really high NPS. Customers love to use our stuff. They tell their friends about it. It's viral. It takes off within a market. And so when people are asking their company for access to software, they're like, hey, can you put this thing in Slack? So we've been able to develop this kind of ground, swollen, grassroots support and pull from our users. I think that it's really, really helping us right now. You saw it play out at a large social networking company last week, right?

This opens up an interesting thread here on monetization, and I'll ask it in a couple of different ways. So first with Slack, historically it's been a fee count model.

Per user per month, yep.

Per user per month, and now you have any given user getting boatloads more value, especially for the AI-pillot examples that you gave there. How does the pricing model change for the Shopify users of the world?

Yeah, so it's interesting. So we have, and Bill obviously, feel free to chime in here, we have a series of plans. So we, I think, have been one of the more successful premium products. so we still have the overwhelming majority of what we call our invoice business, which is our sales business, our enterprise sales business, initiates is what we call self-serve. So you start as a free team. You grow up through our plans, through nudges in the product to use more features of the product, start paying for it, and then eventually convert over into what we call an invoice enterprise plan that our salespeople help you with. From a monetization perspective, I think there's a couple things that we can do here. So, one, like, overall, like, the total TAM for knowledge workers is around a billion at this point, so I think we still have a ton of upsides as it relates to user acquisition. But increasingly, I think you're going to see more agents in Slack than you see humans, and that creates a fascinating monetization opportunity that we haven't totally cracked the nut yet on, but I think it creates incredible upside for Salesforce. And so you can think of agents potentially as users. You can think of consumption-based mechanisms for monetization as agents are actually the access patterns for information in Slack by agents. It's very different than that from humans. And so we're in the early stages of that, but I could not be more excited about the upside for Slack.

Yeah, and I'll build on that, Rob. I think as sort of Rob mentioned, what started as largely per user per month is now expanding into the agents that are sort of exploding in the workforce, the usage that's exploding like never before. And even into the future, you're going to start to see, you know, kind of more things like outcomes that, like, organizations are sort of ruminating on or delivering with Slack at the sort of cortex of that, you know, work that really become, you know, kind of opportunities for participating in more value exchange between us and the customers that we serve. So as you think about sort of the monetization opportunities, they're highly elastic and they're highly sort of variable based on sort of what companies do with the software and what is really fascinating about slack is how much people do with the software yep and how you know in in we all embrace it in sort of our core work and workflow um and it just becomes almost like the place where i prefer to do my work as opposed to all these other systems i might have had to interface to do my job with prior so i think the the monetization sort of opportunity you know what started sort of very very small around the user-based world now expands into lots of different sort of upside potential for for many of our businesses black included and that I think goes back to

something you asked earlier that I didn't perfectly answer when you were talking about our differentiation or how we might maintain this is I think that's our moat honestly is like the distribution that we've had the compounding context flywheel that sits in slack and just the more and more you use it the more and more valuable it becomes and I think that's that's ultimately the most.

Bill, I want to ask you a couple of broader monetization questions. The data point we got from Miguel on the earnings call was 5% roughly of the Salesforce installed base is on premium SKUs. And premium SKU can sometimes be a 60 to 80% uplift. Could you just break this down for us a little bit? If we're a Salesforce customer and we're not one of the ones that have already invested engineering resources into making CloudForce happen without CloudForce. And we say, CloudForce sounds amazing. How soon can we onboard? Explain the 60% to 80% a little bit more. What is that uplift?

Yeah, let me give you, it's sort of in the case of our traditional two core cloud, sales cloud and service cloud. But when you think about all the software that's required to make a salesperson sort of productive, You know, not only is oftentimes your CRM system for managing your sales pipeline, your contacts, your accounts, et cetera, but that often sort of expands into your sales performance management system where my quota, where my commission, where my compensation sort of management systems are aligned. Maybe my sales enablement software, which has sort of helped me kind of become enabled or effective as a seller. Maybe that's where all my content exists. I can come out and deliver pitches or proposals to my customers. And there's also sort of all these extremities, like, you know, sometimes it's call recording software or coaching software that helps me sort of in moments of working with customers. Well, the average sort of spend on a customer with all of those other technologies in the United States is about $1,300 for an average sales professional. And what we've been able to do is put a lot of this value into one price edition called our Agent Force One Edition, or we may have a new name for that next week, but I'll call it Agent Force 1 Edition today, that Agent Force 1 Edition, you know, is retails for $558 U.S. So the goal here is to really, you know, kind of take money out of the extremities where we know every sales organization that has to recruit, hire, activate, and make salespeople productive and performant and now put it into sort of one offering that's there. And it's early days for us. You know, we only started really thinking about premium levels of monetization really the last couple of years. And so as Miguel was sort of rightfully pointing out, it's just early days for us to put these premium additions into our install base. And as they go through the lifecycle of a contract with our company, what we're seeing is more and more of our customers are sort of trading up to these additions as they come up for renewal. So it's not something that will be a big bang immediately overnight, is everyone on a premium addition. It is something that as companies have other contracts that they're rolling off from other vendors and now consolidating onto Salesforce, course. That's the motion that we're kind of running with the premium additions that are there. And I think it really allows our customers to sort of save money in the process while driving higher performance and higher productivity for every salesperson that they do. By the way, the same is true on the service cloud side. You know, you have to have a case management solution, a telephony solution, sometimes a chat or channel solution that's out there. And so the same sort of, you know, kind of play is what we're engineering on that side, which is more consolidation of extremity with spend into more of a consolidated Salesforce platform.

So there theoretically will be an upgrade cycle that we could guess at for each of the Salesforce clouds over time. That is essentially the strategy you talked about. The thing that I want to better understand, so this can get pretty confusing pretty quickly, because you've got the agent for a SKU. You've got what will be the core for a SKU. You've got Headless, which is sort of a pick-and-mix independent of Anthropics. Maybe just help – you had a beautiful cohort description earlier. Help us put this into cohorts so we can understand what the different classes are.

You're doing a pretty good job with all the names. And so, like I said, the naming is something that, you know, we're definitely working through as a team. But let me kind of make it really simple. With our Agent Force 1 edition, you get Claude Force. You get Agent Force. you get data cloud, and you get the best of all the sales club that's there. So we want to have one edition that they all come into. Just a super tier. Exactly, a premium tier of all of those offerings. Now, can you buy all of those items discreetly? Yes, you can buy all those items discreetly, but we want to make it an economically viable and attractive offering for companies to get the best of Salesforce in this one edition because what we know is that for those organizations that run that Agent Force One Edition, their salespeople are more productive, They're using the software more because they have Slack also into that addition. They have our Tableau offering, so they have all the analytics that they need to be productive in their role. And so this is really not just a strategy of sort of loading someone up with all the sales force that, you know, they could have. It's really we've engineered this from the start to make it the best addition for workers to have the best performance of their lives.

And that is the 60% to 80% or map it back.

That would be what that number would correlate to.

Now let's introduce spin and qualified into the equation. One of the nuances that we struggle with a little bit is spin in practice sounds a little bit like agent force in practice. But it's not. Explain the difference.

So I think as sort of the world of agents and agentic platforms are starting to clarify for all of us, It's become very clear that there are many customers that want to have, you know, purpose-built, specialized agents to perform discrete tasks on behalf of their business. FIN is a great example. FIN is a solution that was engineered for the customer service domain, really to help, you know, kind of streamline and automate a lot of those routine and interactions that a company has with its customers. And so FIN is a purpose-built agent that is hyper-specialized and focused on that area. Piper, which is an agent from Qualified, is actually a hyper-focused, purpose-built agent for the marketing domain, where Piper's job is to sort of sit as part of your website, enroll someone through your sales and lead qualification process, and ultimately try and book meetings with sales professionals. So these are two agents and two very different tasks on top of kind of a commonly what might be a website experience from a company. Now, what Agent Force does is it sort of orchestrates all of them in the back end. It understands and knows what Finn is doing. It knows and understands what Piper is doing. It knows and understands what agents that are built by developers are doing on this platform so that we know that one agent can sort of start an experience and then pick it up, you know, as a conversation traverses from the marketing world to the selling world to the servicing world. And so this is where, again, as sort of the world between platforms that orchestrate the journey and agents that sort of power the front end of those experiences, that's how they sort of differentiate from one another. But it's overly probably reductive to say, is it qualified versus agent force or Piper versus Finn? We don't see customers sort of say, I deploy one or the other. Oftentimes what we see are companies deploy all of them. They start with the top of the funnel with a Piper, the bottom of the funnel with a Finn, and agent force that sort of orchestrates in between.

And then I also have to budget for Anthropic tokens fund?

Not on those offerings.

When I'm on Agent Force 1, sorry.

Yeah, no, not on those offerings. If you're buying, you know, kind of Claude or Anthropic as a, you know, kind of customer, you do that with them, with Anthropic. As part of sort of Agent Force and as part of Piper and as part of Fin, you're not paying sort of for Anthropic tokens to make those interactions power.

Okay, that makes sense. Let me ask the broader question, which is there's an evolution in software here that goes from being C-count-based to selling outcomes, selling a unit of work, which we think is quite expansionary for the software term. Where does the budget come from when you're having these conversations?

Yeah, well, I think maybe I'll start, and Rob, you can kind of add in from your productivity lens. But, you know, I think as terms of budget goes, it depends on where the outcomes are. You know, sometimes outcomes are actually aligned towards savings. So the actual savings become a self-filling proposition for, you know, for all of your dollar we save, we can reinvest those dollars in the software that powers those savings. And in other cases, you know, some of the outcomes that we're working on now, like in the case of our sales and sales agent domain, you know, kind of paying for qualified leads or paying for orders that are actually booked. And so that's where organizations are happy to invest in those outcome-based agents because the more that those agents perform, the higher revenue that they're actually performing for their organization. So I would say that the outcome-based agents actually allow us to align value realized with budgetary sort of planning altogether. I've not yet met a CFO that says I'm not willing to give you a dollar if you're actually going to help me make three. So this is an opportunity for us to, like you said, get into more of that expansionary opportunity, especially as our benefits are more jointly aligned with customers on those domains specifically.

I mean I would layer in we continue to see more and more of the premium for slack because of the additional work that we're taking on and the economics we're helping our customers with so we're a little bit different so when you were talking about in general the move to headless and cloud force as an example earlier within slack we have something called slack bot which is an out-of-the-box agent for knowledge workers that is built on top of anthropic and it's a broader Salesforce level we think it's important that you be able to accomplish the work through whatever your tool of choice might be so we want to be able to tie into any harness that you use but we are going to have an opinionated harness out of the box and that is slack bot and so when we go to when we look at slack bot one of the bets that we make is we actually take on the burden of token economics ourselves and so it's part of the per user per month see for our highest grade plan our knowledge workers that our customers are able to go to Slack bot and get done much of what they might have done in an open-ended cost tool in another way and does it mean those other things can go away no it doesn't it means it's an easier way to budget and some of the more specific you know higher order like deep knowledge work that needs to happen as other tools will happen over there but this is just like a very easy reliable way for you to to give a knowledge work agent to the the majority of your employees at a reliable cost, and then really target like, yeah, they need cloud code, or they need co-work, or they need an unlimited token spend, et cetera. So, but we see often the premium that we're getting coming at the expense of utilization of other tools, frankly, and it's not necessarily elimination of heads. Like when you look at Slack, like we'll have a business leader bring Slack into an organization, like a sales organization, they'll be like, okay, I wanna connect Slack to Salesforce, and I have a goal of getting my first sales reps, my new sales reps too productive in 90 days or less. It used to take six months. And you connect Salesforce and Slack together and they can do that and they're effectively accelerating revenue rather than that and taking some cost out because they aren't using third parties to come in and train those people. They're actually having the software do it for them. Or it's like, okay, we're doing a percentage of our internal meetings through huddles now as opposed to these other things that we're doing. So we're shaking 30% of the workload off of some of those other tools, similar things with documents, similar things with agents for knowledge workers, right? So we shave off by being more convenient economics in other places.

I have two big-picture software questions I want to pick your brain on. So the first is, Bill, you've talked in the past about back-office versus front-office convergence and how agents can perhaps blur the lines. What do you think is the future of what has historically been silos between back and front office?

Yeah, I started my career as a production scheduler for Motorola as an intern, and so I was like a back office dude. And then I learned about this magical thing called CRM, which is about like actually working with customers. And so I've always had this like inner rift in my own brain between back office and front office. And what was sort of fascinating is the limits of the technology systems have always been really about the roles that were kind of specialized to perform discrete tasks and functions. And when you learn about, like, a lot of the systems in use for many businesses, those systems themselves were operationalized or built for the constraint of what human labor could do. And so I think it's so fascinating right now, and, you know, Rob, you talked about it and alluded to it with the Slack world. We'll see more agents in our workplace than employees and human employees sometimes in our workplace to perform work and tasks. And so we have to go back to the ability to reinvent a lot of these fundamental systems and processes and activities that people do because now the technology is going to be always on, always running in our lives, really to help us become more productive and to drive higher quality output in the work that we sort of perform. So if anyone here is sort of a back office and front office sort of maven, I'm sorry to say I think they'll blend. And I think that they will blend in ways that really break down a lot of categories and a lot of sort of opportunities. But at the end of the day, if we use software to make companies perform better, that's where value systems like Salesforce are really being engineered for from day one, which is help make our companies be better companies, help them sort of engage their customers in better ways and help them sort of transform their business, not to recreate yesterday's legacy, but really to define what tomorrow sort of looks like. So I'm very, very bullish about just workplace and workforce transformation in that regard. I never want to have to kind of do MRP scheduling again in my life. And so I'm very, very excited about this ability to fundamentally hand some of that work to agents and to spend energy where I like to spend energy, which is with people and relating in new and exciting ways.

Okay. So the second one is horizontal versus vertical. And you all had some pretty spicy comments from David Friedberg. it gets to the beginning of the earnings call on horizontal eating vertical. You've announced some really interesting life sciences wins. For either of you, what do you think is the future of vertical software?

Maybe I can start. Because I think Slack's inherently more horizontal, I think, than the rest of the Salesforce products. But I think there are opportunities for us to get more vertical, and I think maybe that'll be a good segue to Bill. But again, coming back to the total available market for knowledge workers, like it's an excess of a billion people I think globally and so we very much look at the horizontal opportunity for slack because what's been very interesting is if you look at from the inception of slack the the challenge that was set forth for the team was is there a way that we can build software that will make coordination easier so to make people the hardest part of people's working lives coordination more simple pleasant and productive and that's a problem space that fortunately has endured pretty tectonic shifts and underlying technologies whether that be you know the the cloud substrates to now the new models you know like coordination is still a pain in the ass I don't think about the Red Komarovsky slime molds article like go read that and I think there is just tremendous remains tremendous horizontal opportunity for coordination and orchestration of work for knowledge workers despite everything that's happening with with agents that are out there today. I think there are domains that we can verticalize in that we're interested in.

You're starting to see us do a slack code which is we can that span industry verticals and so I think you'll see us go in the stove pipes of LOBs rather than necessarily industry verticals but I think you can expect this to be expect to see us go long on on tech and dev and coding in particular from a verticalization perspective but yeah in the spicy commentary I think really the message was about finding out where really your true differentiation you're able to create unique values sort of represents and you know if you keep going down that spectrum there's horizontal and then there's vertical and then there's regional and then there's local and then there's personal sort of opportunities to create value along that sort of spectrum and I don't think that value stops on any one of those sort of you know kind of stops along the journey I do think there's immense opportunity for us to utilize the technology you know and I think I take a lot of inspiration for what Slack does for me in my life where I get out of my own way and I work with my team and I work with my group and I work with my organization in ways that I never could do before. But that just sort of goes on that same spectrum where I'm now working less in a local way but more in a communal way. And I think that as we see this vertical opportunity maybe more clearly, there is great value that we can create making hyper-relevant computing for companies in certain industries that maybe We have regulatory requirements and uniqueness that allows us to maybe sort of deliver higher value services to them or not. So I'm not sort of a big believer in the versus. I think there's a big spectrum of opportunity there, and we're excited to sort of play in that spectrum, broadly speaking.

Hey, it's really good stuff. Please join me in thanking Bill and Rob for their time. Thank you, John. Thank you, everybody.