AI, Creators & Agentic Commerce — A Conversation with Cloudflare CSO Stephanie Cohen

AI has this ability to get it to be a golden age of content, because right now, I think we all agree, there's a bunch of clickbait on the Internet, there's a bunch of repetitive information on the Internet, and AI is really searching out unique information, so things that are additive to human knowledge. And so if we live in a world where that's the stuff that is valued, that that's the stuff that there's real demand for and therefore gets created, I think that's a pretty interesting world. And so you need, as a publisher, to figure out where are you playing, how are you winning in this new and different world. Hello everyone and welcome to This Week in NET. It's the December 8th, 2025 edition. We're coming early this week because we have a special guest with us. Stephanie Cohen is Cloudflare's Chief Strategy Officer, the first Chief Strategy Officer of Cloudflare, and is sharing with us how strategy is determined in a tech company, and what Cloudflare is doing in AI, the financial sector that she knows so well with over 20 years of experience at Goldman Sachs, and also what to expect from Cloudflare about the intersection between content creators and AI companies in the next year. We talked with Stephanie a few weeks ago when she was in Lisbon for Web Summit. Hello Stephanie and welcome to This Week in NET. Great to be here. Thanks for having me. So you're in Lisbon right now. Can you tell us a bit of what you're doing in town? Well, I love coming to Lisbon. It's a beautiful city, but we also have such an amazing team here. So I'm here seeing the team, but it's also Web Summit. So I got the chance to speak at Web Summit, talking about Agentech Commerce. I was on a panel with MasterCard. We recently partnered with Visa, MasterCard, and American Express on some protocols for Agentech Commerce, and we also have a bunch of dinners with all the Web Summit participants. It's an amazing event. It is a very international event in Lisbon, a cool one. Your first time at the event? First time at the event, not my first time in Lisbon. Specifically, for those who don't know, can you explain to us a bit of your path before and during Colfer? Just a quick bite for those who want to… So I grew up in the suburbs of Chicago. I was a competitive figure skater. I went to the University of Illinois, and after college I went to Goldman Sachs. I was an analyst in investment banking, and I spent a bunch of time in investment banking at Goldman Sachs. I was Goldman Sachs' chief strategy officer, and then I ran two of Goldman's reportable segments. And then about a year and a half ago, I came to Cloudflare as our first ever chief strategy officer. Exactly. So about the first ever chief strategy officer, what can you tell us about the role in particular? Were you trying to make the role your own, even because you were the first there as well? So I think it's one of those roles that you do make your own. When I became the chief strategy officer at Goldman Sachs, I got handed a blank sheet of paper, and they said, go figure out your job. And that was basically the same thing at Cloudflare. And so one of my first jobs at Cloudflare was to learn about Cloudflare, learn the basis of our competitive advantage, learn our people, learn our customers. And from that, we figured out what were the most important things for me to work on. And so I try to target things that are key initiatives that are on our path to becoming an iconic technology company. So think about that as the future of our network and infrastructure, things at the intersection of AI and content, our developer platform, or where we bring go -to-market and R&D together to make sure we're building products for specific sectors like financial services or media or retail. And then for fun, I get to work on our U.S. Ski and Snowboard Partnership for the skiers and snowboarders out there. You know a lot about that as well. And you live in Park City, right? So you are close to where the action in terms of ski happens as well. I'm the worst skier in my family by a material margin. I learned to ski as an adult. But yes, Park City is a beautiful place. Regarding your full, more than a year now, almost two years at Klaufler, were you surprised with the shift from a company with a lot of legacy, a lot of traditions from the 1900s? Actually, the university went to, I think it's two years apart, Goldman Sachs and the University of Illinois, so old institutions. But now you're at a very recent company. You know the founders because it's a young company. What was like the transition? What have you learned? So I was really running towards technology and everything happening at the intersection of cloud and cyber and AI. And I really, I wanted to learn it, but I want to be a part of it. I have young children, and I wanted to work on something that was going to affect the world that they live in. And I knew that the innovation velocity was really rapid inside of Klaufler, but it's hard to imagine what it's like from the outside versus being inside. So in many ways, it's exactly what I expected, that people are amazing, the technology is great, the customers are super interesting, what we're doing is really interesting. I find myself in conversations where I know things about cutting-edge technology that I know I would have never known if I was not sitting inside of Klaufler. But actually seeing it up close, one of the things that I spent a bunch of time thinking about when I was inside of Goldman Sachs was how do you maintain a culture that's innovative? How do you have a really large business but then continue to innovate? And Goldman Sachs has clearly done that over a long period of time, given how long Goldman Sachs has been around. It's one of the things that Klaufler is so good at, which is keeping that innovation engine going, despite the fact that we have very large businesses with lots of customers and lots of products, and there's lots of work that has to be done to maintain and grow those businesses. But we still manage to figure out how to continue to innovate faster than any company I've ever seen. Regarding the strategy part, and we talked about this in the show lengthily with Will Allen and others, all of this talk about AI and creators, also as a new area for Klaufler, being in the middle of content creators and AI providers. How do you see that came about in terms of strategy, of pivoting, of changing that part of the strategy? One of the great things about Klaufler is Klaufler is so mission-driven. So our mission is that we help build a better Internet. One of the most important words in that sentence, of course, being help, which means that we can't do it alone. And it's a company where that's not just written on a wall, it's written on a bunch of walls, but it really shows up in every meeting and shows up in every discussion. And so if you think about what we're doing at the intersection of AI and content, it's completely aligned with the mission. Because there is no need for an Internet if you don't have high-quality, diverse, original content. If you're just going to live in a bunch of walled gardens, you don't really need the Internet. And the Internet's obviously been great for Klaufler, but it's been great for society, it's been great for innovation. And so all of this stems from this idea of how do we keep a thriving Internet? And so we saw a lot of things that were happening on the Internet. We see trends at Internet scale, given more than 20% of the Internet sits behind Klaufler. And we started to see that people were reading derivatives rather than reading original content, and that was changing traffic patterns on the Internet. And fundamentally, people weren't going back to the original source. And when that happens, there's no incentive to create that original source because you need ads or subscriptions or maybe you just get ego boost from someone reading something that you wrote. And so we realized this was happening. We could see it because we see trends on our network, and we said we've got to do something about that. And then that will intersect with another one of the things that we had started to really work on, which is how do we make sure that we're speaking at companies of all sizes to the thought leaders inside of those companies? And so there's the practitioners who are buying our product who are really important, and the product-led growth inside of Klaufler is really important. But it's also important as we sell more and more products and we're more of a platform that you're speaking to the senior leadership. And so it was in those conversations, the combination of our data and what they were seeing in their business, there was clearly a need for what we were going to do. And we view ourselves as an enabler. And when we first started having conversations with content creators over a year ago, they felt lost. They felt a little helpless. They said there's nothing we can do about this. AI is scraping all of our stuff. No one's coming back to read it. I guess our business is over. And we could show them that they could have the tools. They could have the control. They could have the transparency to decide what to do and therefore create a business model for themselves. And so I think that is one of the reasons why the Klaufler culture is so great. It's because we're trying to ultimately solve a customer's problem. And we do that by listening but also bringing to bear the large Klaufler network both from a product perspective and a data perspective. You've been in panels even today, more that one about, of course, Mastercard, Visa, financial players, but in others with publishers as well. Probably you already said this, but the main question they do ask, but also from the open AIs of the world, the AI companies of the world, they're not asking the same question for sure, but are there some similarities in the way they approach this? I think AI companies and AI innovators more broadly plus domain owners, site owners, content creators, e -commerce sites, they want to figure out what the business model is on the other side of this. I think everyone is excited about the innovation. Everyone has the experience of it used to take me hours to research something that I wanted to go buy and I no longer have to do the two hours of hunting through sites. I can actually just get the answer right there. Everyone as a consumer likes this experience, but they realize that in order for this experience to be sustainable, we have to build business models that work. I think everyone is trying to do that. I don't think that's just the content creator. I think the AI companies know that in order for their models to keep getting better, they need high-quality, original, human -generated content. In order for there to be commerce on their site, unless they're going to go start manufacturing everything, they're going to need e -commerce companies and retailers to exist. I think everyone's trying to figure it out. The things that the creators really want is, one, transparency. They want to know who is on their site. That's why it was actually the first product that we released in this initiative. We called it AI Audit, and it allowed you to audit. It allowed you to see who was scraping your site. Then control, which is the ability to decide which bots you want on your site. Once you have those two capabilities, then the question is, what are you doing with that? Are you blocking as a business model? Meaning, is your view that you don't want your information in AI and you want people only to go back to the original source? Is your view that I may need to create some scarcity in my business? I may not give my stuff away for free for some period of time, because if I don't give it away for free, the models will realize that they need it, and then they'll come back to me and have that conversation. People, Inc., the CEO of People, Inc., recently on the earnings call, talked about how they were using the Cloudflare system to create scarcity as it relates to their content. That had resulted in a bunch of deals for them. There's a bunch of people creating scarcity. Then there's ones who already have deals, and so they're sending their information through. Or their view is that they're getting enough referral traffic that they're happy for you to scrape their site, whether that's because you're selling something and so someone's ultimately coming back to buy, or because for whatever reason there's enough click -through or that quality of the click -through rate is different. I think once you have transparency and control, you can create and experiment with business models. If you get underneath the tech of this a little bit, what's going on is likely unsustainable, meaning we have a human-optimized Internet that machines are scraping. We're very likely to move to a world where there's a human Internet and there is a more machine-optimized Internet. We recently launched a product called AI Search, which is this ability to make the information on your site easily readable by AI models, which helps the site owner, but it also helps the model because it makes it significantly more efficient. That's a very cool one, actually, and it was built here in Lisbon, the engineering team. But one of the things, and you spoke about this, we can see is that the final model is not there yet. We're seeing, but it's very cool, even me, and potentially you'll think the same, it's very cool that Cloudflare can actually, for example, block even the blocking part and create the scarcities. That's quite a cool perspective of we can contribute, we can pivot, we can change, iterate. That's a very tech company thing. But also, being on the side of both content creators and AI companies that want the good content as well, it's kind of interesting as an intermediary perspective, really. I think the ability for Cloudflare to be an enabler to what's going on is really important. The Cloudflare network has been built over 15 years and really is amazing. And the fact that we're in more than 335 cities and more than 125 countries with CPUs and GPUs and storage and bringing that level of technological capability to companies of all sizes. And I think it's one of the most important things we need to think about in this AI era, which is, like in a lot of these transformations, we have the risk of lots and lots of concentration. And I think we like living in a world where there's diversity, diversity of content, there's small businesses. And so one of the things that we talk a lot about is how do we, as the Cloudflare network, make sure we're enabler of the long tail. By the way, the long tail on both sides of this equation, the long tail on the innovation side, all the startups, all the new models, how do we make sure that they can continue to innovate and we don't have this significant concentration of power that could happen in a time and place where there's so much capital investment required. And so the Cloudflare network does a bunch of amazing things, but I think one of the greatest things it does is it provides Internet scale to companies and individuals. Were you surprised with how much AI you're working on these days? The questions that people ask, it's on everyone's mind, publishers in particular. But were you surprised that you are not dealing more with the financial sector, as you had so much experience, and also with AI, potentially even more AI than financial these days? Yeah, so when I started at Cloudflare, if you had told me that I'd be working on our infrastructure network and that I'd be spending all this time at the intersection of AI and content with a bunch of media companies, I wouldn't have believed that, but I think that's just a testament to how fast everything is moving. Having said that, I do spend a bunch of time with customers in the financial services sector, both on normal resiliency, cybersecurity, all the things that Cloudflare does. But on top of that, they're also having this problem, right? Meaning, if you're a financial services company and you have a direct relationship with an end customer, if you look at the different chatbots, one of the biggest uses of it is people asking investing and wealth management questions. And so that is a situation where you'd be sitting on a chatbot rather than going back to the original writer. In many cases, financial services companies write lots and lots of articles about wealth management or what's going on with the economy in order to get you to go back into their environment. And so it's very interesting to see that, yes, for a little while, they're spending lots and lots of time with media companies. I still do. But now all the things that they were dealing with are happening in all of these other sectors. And then now we've gotten to agentic commerce, which is absolutely at the intersection of payments and our network infrastructure. And so I still get to spend some time in financial services. Of course. And we had in the show Thibault from our research team talking about WebBot Auth, a new protocol that actually is enabling payments. So financial sector with agents, AI agents, which is a very cool thing. And on the financial sector, what are the, and you already spoke about the AI perspective, but even outside of that, there's the Zero Trust VPN type of thing. There's the make their online experience better for customers. What are the main concerns in the financial sector these days? Yeah. So one is resiliency, right? This idea that you need to be available to your end customer. I mean, if you think about financial services, depending on which piece of it, but some part of it is part of your everyday life. It's literally how you are able to get into something like an Uber or you need to go pick up lunch, like all of these things we all take for granted. But that's the financial services sector working in the background. So resiliency is really important. And so we're spending a lot of time in that sector where companies are figuring out whether or not they want to add another provider. Many of those companies had a different provider, but they basically made a choice before we were a company. And so now we've been added in a bunch of places as a resiliency provider. And so I think that depending on the size of company, depending on the geography, some of it's regulatory driven and some of it's customer experience driven. But there's significant focus on that. The second thing is the Cloudflare network. Because of our millions of free customers and locations all over the world, we have enormous access to information around what's going on on the Internet. And so we also spend lots of time with particularly the banking sector around threat intelligence. And we're part of an organization called The Arc. It's the largest banks in the U.S. that come together and work together around cybersecurity. And there's basically two tech companies that are part of it, and we're one of them. Similarly with FSISAC, which is a larger group of financial institutions. So trying to make the overall sector safer by sharing threat intel across. And also things like suppliers, vendors. Like how do you work together so that you can create resiliency across an entire sector? Sharing information, but also through a bunch of our products, whether it's the Zero Trust products, our network services products, or our reverse proxy products. And then you mentioned it, which is the how do you create wonderful, beautiful experiences that have security built in by design. And so spending time in financial services with our developer platform and making sure that as they're thinking about things like MCP servers, that they have cybersecurity built in by design. And then maybe last, just because you mentioned WebBotAuth. The great thing about WebBotAuth is that in an agentic world, you not only want to know who's on your site, you want to know that they are who they say that they are. And that they're not something being spoofed from an IP address. And so we created WebBotAuth as a protocol that's open to anyone where you can do cryptographic verification. And while that's important in general, like you don't want bots that are not who you think they are on your site, it becomes increasingly important when commerce is at play. One of the things that I think is quite interesting is, first, being a tech company that tries to be on the edge, even creating protocols, that's how on the edge it is, also helps those that are navigating a new AI era with new vulnerabilities as well, because those are appearing. So even from the financial sector, with a lot of legacy, usually the financial sector moves slowly because they need to move securely and with everything in place and always sure about things. How do you see that legacy part of moving slowly because of regulation, because of all that, with Cloudflare helping some of those in the financial sector? I think that's one of the amazing things about having Cloudflare as a partner to you, because we bring this security-first mindset. If you think about how we've adopted AI internally at Cloudflare, how we build our product, we build our products with security built in by design. And so while we tend to move fast, we don't move fast and break things. And so I think a lot of the financial services sector has liked joining with us as a partner because we can bring them to this cutting edge of technology, but bring it to them in a secure and reliable way. And if you think about the financial services sector and how they think about innovating, there are obviously regulation and other reasons why they have to make sure everything works exactly right. But many of them have sandboxes. Many of them have parts of their organization that can move quicker. And so you find that because they're so tech-focused, most financial services companies are like technology companies embedded in them. So it was one of the reasons that actually at Goldman Sachs where I learned all about engineering and technology, and it was because so much of my team were engineers that I ultimately wanted to be in the technology sector. And so they're so forward-thinking around technology that they're able to engage with us in a very sophisticated way. So they're absolutely careful and prescriptive. But you see actually a lot of innovation happening in the financial services sector. I was in New York at an event. The event was entirely about AI encoding. And most of the people there were actually financial services companies because there's so much engineering work that has to get done. And there's a lot of what they do that is kind of perfect for AI because it's less interesting. It's a little bit more mundane. It can be checked. But it's something that humans would actually rather not do. There are companies like Robinhood, like those are the new bank in Brazil actually were quite popular. And I'm talking about that one in particular because I do our Internet services popularity for radar. And new bank is quite popular in Latin America. And that's a very online bank. How do you see the process? So we were able to see Stripe and others showing that you can do a lot very quickly in the financial sector. Some of those are now starting to have more typical banking systems after a few years. There's a big path still in terms of customer improvement, more online interaction, easiness. Where do you feel like the older perspectives of banking going to the more online fast moving tech sector? There's a big path to cut through. The financial service sector is a big broad sector. And so there's the part of the financial sector that's heavily regulated. It tends to be the part of the sector that takes in deposits and lends out money. And I think that part has generally been slower to move. It's also the hardest one to disrupt because of the regulatory moat that sits around that. But I do think in many ways that's some of the place where there's the most benefit from things like A.I. Because again there's a lot of rewriting of their systems. There's a lot of moving it to more modern architecture. There's creating better front ends. All of that stuff lends itself to A.I. enabled engineering. And so I think you're going to continue to see companies particularly of scale invest in technology in order to do that. Interestingly there's then a bunch of platform technology companies that help smaller banks to enable the right level of technology. And this will just get to my broad view of the world. The largest companies in the world have the money and capability to invest in technology and CapEx and kind of everything else. The key is whether or not there's enough platforms that get created in order to enable a more diverse long tail. And so you know Cloudflare is one of those platforms. A platform like Shopify is one of those platforms as it relates to merchants. And so I think that in this age of A.I. again where we started we were here before and we'll come back here again. And I guess we'll keep coming back here which is that in many ways you can say because of everything that's going on because the level of investment required just the largest companies are going to continue to accrue power or we can all collectively work together to figure out ways to enable these broader ecosystems through platforms. So yes the biggest companies are successful by the way. The largest companies in the world sit on Cloudflare but we also do our part to enable that longer tail with Internet scale with infrastructure with the world's best technology so that they can compete. Because I don't think any of us want to live in a world where it's just a bunch just a few really large companies. When someone from the financial sector maybe with more legacy like a European very style type of banking old bank ask you for some advice where should we start like moving forward to a more online path securely. What are your drive your main advice there. Before I tell them to like start by partnering with Cloudflare before I bet that we'll assume that that's a given. Where should they start. Yeah exactly. You know it's what it's one of those things where if you look at something you say this is just too big it's too hard. There's too much to do. There's too much legacy but there's always a part that you can find and figure out a way to innovate there. And I actually think the way that Cloudflare is structured is really interesting in this regard. We have no emerging technology and incubation which is a part of our R&D team that focuses on the newest things that that group behaves in a different way. It tests and iterates more. There are more things that they start that maybe they don't finish because we decide we're not going to do it versus the part of the R&D or that's been doing products that we've been in for a really long period of time. And while every company will figure out their own version of that model that is a good model like that idea of creating a safe place for experimentation for for iteration. One of the biggest issues legacy companies will have is that people are afraid to fail. Like so if you take an existing business and you do a bunch of stuff to fix it you may actually destroy the business in the process. And so it's really important if you're going to get on this path of transformation and innovation that you take a part of the business. It's safe in that environment to test and iterate and then you make a bunch of mistakes and you figure out what works and you take those learnings on to the next one. It's almost impossible when you look at the whole thing and say like from the ground up we're going to fix everything all at once. Makes sense. The thing that I find interesting is the perspective of people companies really are trying to make the best of the reality of these days. And you've worked in a very old company that had a lot of legacy but as you said was trying to innovate. And you've contacted with other companies even for partnerships mergers and acquisitions things like that. How do you see companies in the future creating partnerships. You speak about Koffler helping being one of those that is helping. How do you see companies nowadays need to trust more in good partnerships than potentially even in the past. One of the times of change I do think human beings can try to band together and figure out how to work together. So it's not surprising that in kind of this age of A .I. which is a transformative period that you're seeing lots of companies try to figure out how to work together. And some of that is security. Some of that is it's hard to tell exactly what's going on. And if you're working with someone else you can figure that out. Also in a higher risk investment period maybe doing that amongst a group of people is safer than doing it on your own. One of the things though just that's a risk around any sort of partnership is that the announcement is great. There's a press release. There's like some articles. But then the work really begins and you need to make sure that if you're doing partnerships that there's a real reason for it. The framework I tend to use kind of across lots of different types of decision making is called playing to win. I will give Roger Martin credit for the for playing to win. It's his framework. But I think it's a really important way to think which is that you've got to figure out where you're playing and then how you're going to win and then what the management systems and capabilities are to support that. And on top of all that you have a winning aspiration. But like the two tidbits around that are one, don't play to play. You've got to play to win. If you're an adult and you're like playing soccer with your friends maybe just play to play because you don't want to get hurt. But like here you need to play to win not play to play. And that something that where the opposite is stupid on its face is not strategy. Like saying that you're going to treat your customers really well is not strategy. That's just a to do list. You've got to figure out how you're going to treat your customers really well because the opposite is obviously ridiculous that you're going to like treat your customers poorly. So you've got to make sure that you're making real decisions that distinguish what you're doing versus others. And bringing that to a partnership situation makes a ton of sense because you could have you may have a where to play how to win where that how to win requires some other capability. So I'll give you an example. So we announced a partnership with Visa and MasterCard American Express and a bunch of other payment companies around it. And it's this idea that we want to enable agentic commerce. And the problem with there's two sides of agentic commerce. We have people who are building agents and we have a developer platform and an agent SDK where we make it easy to create agents. But you then need to make sure that those agents have access to Web sites. And then we sit in front of a lot of Web sites and we need to make sure that good agents get through and bad agents don't get through. And so that's like an ecosystem in order to make all of that work. And so we had the Web bot off protocol which we talked about. But then there's a whole host of payment protocols that need to be added into that. So you know where the agent's coming from. You can verify the identity. You can put like a payment token in there or you can do. There are also other things you can do around stable coins. And so there's this whole ecosystem that's required around that. And so if we're in a world where we say we want to enable agentic commerce and we think that Cloudflare, best place in the world for you to build and deploy and run your agents given the big network latency and cost advantages related to that, there are a whole host of other things that had to happen. And so those types of partnerships make a ton of sense because we can't do that on our own. Makes sense. We spoke about if I'm in the financial sector and want to start to be a bit more online ready, what should I do? And you get this a lot. If I'm a publisher and I want to learn what Cloudflare is saying about this, where should I start? Well, one, we have a free and pay-as-you -go product. So you can just go to Cloudflare.com and you can start there. AI crawl control is a super easy way for you to see once your traffic is coming through Cloudflare for you to easily see who is on your site. And then there's a bunch of buttons where you can decide to block or allow or you can ask to be part of our beta to charge. And so it's completely straightforward as it relates to getting started. Having said that, there's a whole host of business decisions that need to get made around that. And so our advice is that your traffic's running through Cloudflare. You can see what's happening on the site. And spend a lot of time looking at that, like which bots are scraping you, which bots are not, why, which specific articles. There's a whole host of information that exists, not just at the this bot or that bot, but when, how much, what type. And so learn from that. And then you need to figure out what your own strategy is around that. Do you already have deals? Do you want to block? Do you want to block in aim of getting a deal? Would you like to do something more programmatic, which is our paper crawl function? And then there's this question of, well, maybe I want to create AI experiences in my own environment. So we have a partnership with Microsoft around NLWeb, which is Natural Language Web, combined with our developer platform. So you can create more of an easy, agentic AI experience inside of your environment. And then, of course, there is this question of maybe the idea is I don't want to actually my site scraped at all. I just want to feed you the information. And that's the AI search product. And so there's a whole host of technology solutions that are pretty easy for you to get access to and play around with. But then there's a whole host of business solutions. And my guess for most people is the world's changing too rapidly to do anything long term. You can make decisions pretty quickly, but make sure they're one-way door, they're two-way door decisions, and you can easily reverse them. Because I think it's a pretty interesting time. But AI has this ability to get it to be a golden age of content. Because right now, I think we can all agree, there's a bunch of clickbait on the Internet. There's a bunch of repetitive information on the Internet. And AI is really searching out unique information. So things that are additive to human knowledge. And so if we live in a world where that's the stuff that is valued, that that's the stuff that there's real demand for and therefore gets created, I think that's a pretty interesting world. And so you need, as a publisher, to figure out where are you playing, how are you winning in this new and different world. Makes sense. Last but not least, for those who want to get your sense of leadership, do you have an advice, even a book that you recommend or someone that people should be aware of, like a talk? So I talked about Roger Martin because I think I talked about it more as business, meaning in your job. There's actually a whole host of personal things around that, meaning where are you playing? How are you winning? What are your management systems and capabilities? Where do you need to invest in order to learn new skills? And so I actually find it an enormously helpful framework in my own life. One of the things that I learned really early at Goldman Sachs was to make the people around you better. And so obviously there's a whole host of things you do to make yourself better, but you're only one person. And probably now my advice is going to be, like, how do you make other people and probably agents also better? It's probably both. But this idea of being a real enabler of other people, one, like they do great work and that's amazing, but actually that's true. But the other thing that happens is people really want to work with you. If you're someone who they know if they come work with you that they're going to be better, that's what your focus is going to be on is actually making them better than they tell all their friends. And so it's one of the best pieces of advice I got. I learned it very early. I was an analyst in investment banking at Goldman Sachs. I had an intern added. So very early in my career someone was working for me because we had an intern. And I just immediately saw the benefits of instead of just handing them the work and not explaining it, actually taking the time to explain the reason for it meant that I only had to do that once. And then my life got a little bit better versus when I had to do everything on my own. I bet. That's great advice. Thank you so much, Stephanie. Thank you for having me. And that's a wrap.