James Cadwallader. Profound.

James Cadwallader:

When ChatGPT recommends a product, it doesn't just give you a link, it opines. On average, ChatGPT will use about 3,000 characters to answer a commercial question. And in this new world, the question is not just do I show up, but how do I show up. Chat or Claude know more about Audacious maybe than you do, but there are things that you know about Audacious right now that Chat and Claude simply don't know. You're an API between reality and AI at this point. We've got an absurd number that we're trying to hit this year. It's just a ridiculous ARR number. For us to hit that number, there are just laws of physics. That is all a function of people.

Nakul Mandan:

Marketing, as we know, is about to change more than it has in the last 25 years. The old internet marketing playbook was built for humans browsing the web. Search, social, feeds, creators, and paid distribution. The next one is being built for AI agents interpreting the web. James Cadwallader is building the infrastructure for this shift. In under two years, his company, Profound, has gone from zero to a billion-dollar company with 700 enterprise customers and more than 13% of the Fortune 500 already on the platform. James is especially interesting, because he saw the last major platform shift from inside the room. Building Kyra and running creative for Nike, Apple, Beats, Unilever, Prada and others as social and TikTok were eating the old marketing playbook in real-time. Now, he's building through the next shift. Today, we get into the new marketing operating system, the future of marketing orgs, and what an AI native company actually looks from the inside. Knuckle Up. James, welcome to the show.

James Cadwallader:

Thanks for having me.

Nakul Mandan:

Great. So you built Kyra through the last big shift when marketing shifted from search to a social media era and TikTok was taking over, and now you're building it for this new world. What's actually changing in marketing today compared to the prior world you built at Kyra?

James Cadwallader:

I think this platform shift is orders of magnitude bigger than what we've seen in the past. I think this shift from blue link search to the generative internet, internet where you can talk to it and it talks back to you, puts the prior platform shifts of social, mobile, cloud. Yeah, I think they pale in comparison. Yeah, I personally believe that how AI talks about your brand products and services will be the biggest function of marketing. This will be the highest priority item on every CMO's list by the end of this year.

Nakul Mandan:

But what's actually shifting in the sense that is it how people are discovering brands now or being recommended brands that is changing in your mind? Because the internet already exists. There's a lot of ways brands are marketed. Is this just an add-on to how the world of marketing plays out, or is this a massive paradigm shift?

James Cadwallader:

No, I think this is a total rethink on information retrieval. The human being's ability to exchange ideas has always been underpinned by technology shifts. So if that was sharing ideas around a campfire or via a newspaper, or television, or the Internet, or social media, now our ability to access information through just having a conversation, this is upon us now and how AI talks about your brand is already a really big deal. I think that the failure mode on this journey, we launched the business 21 months ago and the thing that's been most shocking to me is how much people in tech, in particular, seek heuristic when trying to understand the future. I think people want to use the pattern matching of search and SEO to understand this new world, but there's just new dimensions to this. When ChatGPT recommends a product, it doesn't just give you a link, it opines. On average, ChatGPT will use about 3,000 characters to answer a commercial question. It's like 11 tweets. It's like an entire tweet storm and it will offer all kinds of supporting vernacular. It will give you comparisons. It will compare a competitor's product to you. It will give all kinds of detail that is essentially unsolicited by the person asking the question. So really, in the old world of search, it was just this question of, "How do I rank?" And in this new world, the question is not just do I show up, but how do I show up. And especially for software businesses, we work with some of the biggest software companies on the planet and this is no longer a question of, "Do I show up?" And the answer when someone's asking Claude to build something for them, does Claude recommend AWS or GCP? And then will it go, "How does it use AWS? How does it troubleshoot AWS?" Claude is the user.

Nakul Mandan:

It's more definitive, because it is not 15 blue links, it's like one name, maybe a second name, and then the opinion that goes with it in a pretty definitive way or declarative way from ChatGPT.

James Cadwallader:

Yeah. And you see all kinds of misinformation being pulled through. As an example, Southwest Airlines. We're seeing tons of instances right now where if you tell ChatGPT that you're flying Southwest, it will append unsolicited advice of like, "Hey, Nakul, you might want to show up early because they have unassigned seating." But Southwest ditched unassigned seating in January of this year. So ChatGPT, one of the most trusted sources of information on the planet is giving this pretty detrimental advice to consumers right now and that's on the Southwest marketing team to fix that today in 2026. They have to build great marketing that is designed to be consumed by agents to fix the narrative and allow AI systems like ChatGPT to understand that, "Hey, Southwest no longer offers unassigned seating."

Nakul Mandan:

So maybe let's take a little bit of a step back. At Kyra, you saw pretty large brands, Nike, Apple, Beats, Prada. What did you learn about how these best-in-class brands think about brand marketing?

James Cadwallader:

At Kyra, that was really my opportunity to cut my teeth and understand what it is like to partner with and frankly sell to some of the biggest companies on the planet. I feel like at Kyra, we never experienced the level of product market fit that frankly we're experiencing here at Profound, but what we did do very well is we understood the ICP of the marketer and we spent thousands of hours at the offices of some of the biggest marketing teams on the planet. The brands you just mentioned, we would go and do offsites and understand the psyche of marketers. And I think going into this business now with Profound, I think I'd maybe take it for granted, but yeah, I understand the marketer ICP pretty thoroughly.

Nakul Mandan:

Did you have some learnings around how the most successful marketers play the game?

James Cadwallader:

I think the most successful marketers, it's almost like ... Do you know the midwit curve meme where it's like you have the dumb person, then the overthinker, and then it returns to ...

Nakul Mandan:

Yeah.

James Cadwallader:

Yeah. I think it's almost like, "Let's just try things." Then it's like, "No, attribution, measurement, everything's a science, a math problem." And then you go back around to the most sophisticated marketers have the confidence. They use their gut a lot and they have the confidence to just try things. Yeah, you won't always know in advance if something is going to work or not. Marketing is this fascinating category in business, because I mean it's a top three line item for every company on the planet, right? If me and you started a sneaker store, the first thing we think about is marketing and distribution.

Nakul Mandan:

I mean, you and I are doing marketing right now.

James Cadwallader:

This is marketing right now.

Nakul Mandan:

This is marketing right now.

James Cadwallader:

That's right. Yeah. It's a top three line item for every business on the planet pretty much, yet so much of it doesn't work and is very difficult to attribute. It's still very mercurial as a category. And we're seeing a change in marketing, but really the interesting thing here is a change in the consumer. Because as a marketer now, you're not thinking about how you access John Smith, the consumer. You're thinking about how you access John Smith, the consumer's agent. So you build marketing for machines. And what does that mean? Machines are way more objective. They pay a lot less attention to the platitudes and the emotional vernacular that marketers have spent so long perfecting. I mean, right now, AI systems like ChatGPT don't pay any attention to visuals. Visual marketing has been the thing for the last 20 or 30 years. So, yeah, everything is changing right now.

Nakul Mandan:

So in this new shift, every era of marketing has a unit of attention. So for instance, in the search world, PageRank was that unit that everybody optimized for. In social, maybe it was the impression. What is the unit around which AI or AEO is built around? What needs to be optimized as people use your platform?

James Cadwallader:

I think optimizing is a weird word. I think it all sounds, again, more mathematical. I think you are just building marketing that is designed for an agent to consume rather than a human. And the way that you do that is you think about how does an agent browse the internet right now. So if you ask ChatGPT what's the best running sneaker for a marathon, it's going to send a bunch of crawlers to the web. And then if you think about it, how do these agents navigate information on the web? Firstly, they are extremely thorough. I use ChatGPT to research a showerhead and it visited 65 websites to find a showerhead for me. A human would never do that. The second is they really pay attention to the atomic details. They'll go into, let's use that showerhead example just to continue. They will look at the water pressure, they will look at the weight, they will look at the material that it was made out of. They will look at slight pricing discrepancies in a way with an intent that a human would never be capable of. Then the third thing is they're looking in real time, so that they pay a lot more attention to up-to-date information. It's almost like AI has a chip on its shoulder. ChatGPT knows that it has to find more recent information, because if it relies on out-of-date information, it can send the results in a direction that's wrong. So they really index on up-to-the-minute information.

Nakul Mandan:

I mean, how is this that different than SEO?

James Cadwallader:

I think the philosophical difference between what we're talking about here with agent-led growth and SEO is in SEO, if you're an SEO marketer, fundamentally, you are creating marketing that is designed to rank in the Google algorithm but fundamentally be consumed by a human being. Whereas in this new world, we're creating content that quite literally may never be read or consumed by a human.

Nakul Mandan:

But ultimately, even AI is going through those 65 websites and looking through what third-parties are claiming about your product. So ultimately, whether it's an agent or a human, it is kind of surfacing all the insights that live on the internet. And so in the prior world, the SEO agencies would try to make sure that whether it's Reddit or other places, it's all ranked in a favorable way for you. So I understand that now you're saying it's more definitive, more declarative, it presents the brand with an opinion also. But ultimately in the background, is it still ultimately ... From a brand marketer's perspective, ultimately they still need to have a good product with good reviews across the internet?

James Cadwallader:

Both things are true that it's the same muscle group, right? You are still marketing effectively. There's no silver bullet here. The onus on marketers now is to do quality marketing and more of it and think about agents rather than the consumer, but fundamentally, you are still producing information just as you and I are now as we sit in this room on camera and distributing it with the hope that it will influence opinions, which is fundamentally marketing, which I think is what you're getting at. I would say the thing that's different though is that, hey, if I want to show up on ChatGPT, I will pay attention to a very different set of signals and information sources than if I cared about Gemini. And if I care about ChatGPT in this specific topic, in this specific language, in this specific territory with this specific user group. So you're being very intentional about creating marketing that's designed to target where ChatGPT goes to get that information.

Nakul Mandan:

Let's then talk about how marketing orgs are changing with this shift. So let's say I'm the CMO of a Fortune 500 company. I know you guys serve a bunch of them. What are today my various marketing sub-functions in this new world and where am I using AI across the board?

James Cadwallader:

Yeah. Well, I'd give a precursor here, which is marketing technology has traditionally sucked. It's always been about doing the work around the work, capital T, capital W. What is the work? The work is very human capital intensive. It is groups of people who would sit and go through a dashboard, look at some changes, scratch their heads, come up with an insight, then write a first draft, get their boss to look over it. Once it's approved, they would publish it or distribute it or post it on social media. They would then check the stats. They would look at how the stats are performing. If it's working, do more of it. If it's not working, do less of it. And what I pose to you today is because of generative AI. A lot of those processes can now be done with agents. And really now, the marketer is becoming a conductor of agents. And if you want to analyze the data, you'll speak to Profound and you'll say, "Hey, can you go and check my citations over the last seven days versus my next biggest competitor?" And if you want to improve your presence on Reddit, you would build an agent and you'd say, "I want you to go and monitor all of the Reddit threads that are influencing this topic that I care about and this language that I care about, extract all the core themes and then use that to write a brief for my team so we can go and focus on Reddit." So I think it's this idea of a marketing engineer is what we've coined and it's a new role within the marketing org. And it's really taking off, because it's this idea of someone who is technical enough to navigate, orchestrate and customize and deploy agents at scale, but also human enough to have that taste and judgment that's required to know where to focus the calories, where to focus the energy. We announced the idea of the marketing engineer maybe three or four weeks ago. We've seen hundreds of job posts crop up. Google just hired a marketing engineer. Yeah, we've seen incredible companies around the world now posting to find marketing engineers. Yeah, we believe that Profound is becoming the home of the marketing engineer.

Nakul Mandan:

But also, as you think about the various sub-functions, brand marketing, paid marketing, maybe content marketing and maybe there's many others, it depends on the nature of the company also. Where is AI being used the most? Where is AI not penetrated? So I'll give you an example, meaning copy generation. One could argue AI is literally generative. So why hasn't AI penetrated deeply into all kinds of copy generation from multimodal media, meaning like videos, images and all of that? Or the other side is also you don't want AI slop, it should still feel like a human connection, taste and all of that. So can you talk about various sub-functions and where AI is being used in which way?

James Cadwallader:

Yeah. So we believe, to kind of extend your analogy, if you think of all these different pillars within the marketing stack. So, yeah, content marketing, brand marketing, performance marketing, growth marketing, social, SEO, demand gen, everything in between. So these are all pillars that represent hundreds of billions of dollars worth of TAM per pillar. We believe the marketing engineer sits above all of those pillars. I think if you're in PR and comms, you will be building agents that will identify media opportunities and create those media lists and do the outreach for you. I think if you're in product marketing, you will build agents that will take a PRD and transform it into maybe objection battle cards that you could give straight to your sales team that pulls from a knowledge base of super brand specific information. So I think this goes across everything. The nuance is that the new headwind or the new tailwind is this concept of slop. Slop is the idea that it sounds like AI. It's regurgitative by nature. There's no original thinking there. The antidote to slop from my experience is more context. I think right now, we could write a thousand blog posts for your company's website right now using ChatGPT. We can say, "Hey, write a thousand blog posts." And they'd happily do it, be like, "Cool, great idea. Here's number one, here's number two." And go and spit out a bunch of crap. The solution there is not to say, "Hey, AI sucks. Let's use pen and paper from now on." And the solution is to say, "Okay, cool. Here's a much more intricate system prompt. Here's a database of super specific information about my brand and our tone of voice. Here's dos and don'ts. Here's a bunch of objections that I've pulled from Gong calls about my business that we can pipe in." Data and context is how you solve for slop.

Nakul Mandan:

Are there areas in marketing that are still not automatable with AI where you would say, "Hey, it'll happen, but right now AI is not as good"?

James Cadwallader:

I think we're going to see a bit of a dumbbell effect start to form where I think brand marketing, and celebrities, and these events, and these very human aspects of the marketing stack will become more and more important. And there will also be this kind of high frequency marketing that takes place where if you're Nike ... I was going to say five years ago, but maybe even three years ago, you'd release a new sneaker and maybe have a few blog posts for it. And now, I believe if you release a new sneaker, you're going to have 250 separate pieces of content and it's going to be localized into 50 different languages and then we're going to speak to why this sneaker is great for people in New York City who walk a lot and also great for people who like hikes in Switzerland or something. And you're going to need to highlight the atomic units of that sneaker. What is it specifically? Not they look good. What is it specifically about this sneaker that makes it good? Is it the weight? Is it the material? Is it the fact that it was designed by some big brain who designed something else very good? You need to give quantifiable, variable objective facts that can be used and interpreted by AI so that when Nakul asks ChatGPT a question like, "Hey, what's the best new sneaker? I'm visiting New York." ChatGPT has verifiable claims that it can use to build that response.

Nakul Mandan:

Where is human judgment still required? In the world that goes in the direction that you are talking about, is there a role for human taste in marketing?

James Cadwallader:

Yeah, there will always be a human role. It's just higher up the stack.

Nakul Mandan:

But is it taste? Is it something else? Is it review?

James Cadwallader:

Yeah, taste, ideas, suggestions, steering. I don't see a future where marketing teams are hunched over a laptop writing that content and analyzing data and having to come up with an insight based on the data. That doesn't seem like something that will exist in five years.

Nakul Mandan:

If the production cost of marketing copy generation goes to zero, what's the new bottleneck? Like you just talked about how there can be a lot of high frequency. Audacious could write a thousand blog posts today about recruiting and all the other things that we do as a venture firm for our portfolio. So what's the new bottleneck in marketing any company?

James Cadwallader:

Ideas

Nakul Mandan:

Ideas?

James Cadwallader:

Yeah, ideas. Original ideas. How do you tell the thing that knows everything something it doesn't know? ChatGPT, I would say Chat or Claude know more about Audacious maybe than you do.

Nakul Mandan:

Yeah, yeah.

James Cadwallader:

I bet if I asked Chat or Claude a bunch of questions, it would give me a better answer about your own business than you would. But there are things that you know about Audacious right now that Chat and Claude simply don't know, because they have not experienced what it's like to step into your office. They've not experienced what it's like to do a deal with you. You're an API between reality and AI at this point. You need to be the way in which AI understands your business comes through you. So it's like ideas are this kind of ... It sounds a bit trite when I say ideas are the limiting factor, but it's really your ability to communicate reality or reality as you want it to be with a very sophisticated intelligence.

Nakul Mandan:

What's the new marketing stack, marketing tech stack? So in a prior world, people would have all these multichannel systems, whether it's Marketo, Braze, and others, email marketing systems, and Adobe, and NetSuite, and Canva for SMBs maybe for all kinds of image creation and marketing copy creation. What's the current marketing stack that you're seeing in parallel to Profound that is forming up in Fortune 500s at least?

James Cadwallader:

Yeah, I'd say ... I mean, eventually, I want it to be just Profound.

Nakul Mandan:

Yeah.

James Cadwallader:

The idea is that we become the place that every marketer spends their day. For now, I think we still see an important role for the CRM and the CMS. I think we're seeing more focus from marketing teams on the CDN now because when agents visit your website, often, a JavaScript Pixel, which has been what we've used to monitor human usage is pretty redundant when agents are visiting, because a lot of them, it won't trigger, won't be pulled through. So yeah, CDNs like Akamai, Cloudflare are becoming the way in which marketers identify, "Cool, I created this piece of content." How many times did ChatGPT's user agent visit that page? That's like the new proxy for a human visit. So, yeah, I think that the thing that I would say about the big difference in the marketing stack now is that point around that you can now do the work inside a piece of software. Prior to this, marketing technology has all been about managing the work or doing the work around the work. Kanban boards or social media planning platforms. But now, what we are saying is, "Cool, you want to create content, you do that in Profound. You want to reach out to a hundred publications, you do that in Profound." Eventually, we'll get to a place where you're running your ad campaigns through Profound, you're running out of home campaigns even through Profound, but the whole thing we believe can live in a piece of software now because of generative AI.

Nakul Mandan:

Got it. So your vision is to go not just ... You started with GEO, generative engine optimization, but you're going for the entire marketing operating system.

James Cadwallader:

I mean, my ambition would be to build a company bigger than Adobe or bigger than Salesforce. I think that would be the most fun. And that outcome, if I could rub a genie in a bottle right now, I think building a company as big as Adobe or as big as Salesforce would be the most fun outcome here. I'm never going to have another opportunity like this. This is it. This is my shot. So, yeah, I'm kind of planning on trying to take it all the way.

Nakul Mandan:

So what's next then? You started with GEO, answer insights, analytics around that. Can you share some of the product roadmap that is coming in terms of the next big pillar that you're knocking down?

James Cadwallader:

We began with analytics, then we shipped agents and we now see it's about 60% of the orgs on Profound are using agents regularly now, which is crazy. That's people using agents to build content, do things.

Nakul Mandan:

And it's multimodal content, not just text?

James Cadwallader:

Yeah.

Nakul Mandan:

Wow.

James Cadwallader:

And you can create reporting. We have integrations with like Gamma now. So if you can query Profound and say, "Hey, build me an agent that builds a report and spits out a weekly report on our SEO data," and it spits it out as a beautiful PDF.

Nakul Mandan:

Can you do ad creation already?

James Cadwallader:

Yeah, we've introduced Google Ads and we've just introduced-

Nakul Mandan:

Like display ads and even video ads yet, or not?

James Cadwallader:

No, just display, just like SEM basically. But then we've just introduced ChatGPT, OpenAI ads into our agents as well. The big thing that you're going to see, frankly, over the next, we're talking single digit weeks, is the recommendation system beginning to really kick in where we have long-running background agents or cloud agents as they're being referred to that will go through the data in loops and analyze the data in the background while you sleep and then you wake up to, "Hey, Nakul, this is a recommendation. Hey, have you seen this? Hey, your competitor just did this." Which sounds trivial and there's a lot of companies in the space like that. That idea in marketing is played out, but no one's really done it with these long-running background agents that are looping and looping and looping. You run a lot of inference through that process.

Nakul Mandan:

Have you guys already gotten into lifecycle marketing also, meaning integrating with things like Segment and others to collect what this user may have done 10 days ago on an e-commerce website or something, or if as logged in or not logged in and stuff and then when is the appropriate time to send them the next follow-up email?

James Cadwallader:

Yeah, we have CRM integrations now and CMS integrations to sort of push out, but yeah, that's all kind of like ... I'd say it's a place that we're putting more calories into now. Yeah. The problem with marketing, it's such a big stack. There's so many angles to it that you can kind of drown if you try and do everything all at once. So, yeah, as you mentioned, we began very focused on AI visibility. How does my brand show up when ChatGPT responds? There's lots of angles there as well. It's not just as simple as the responses. We also plug into the CDN and capture network logs. We also have one of the biggest consumer panels on the planet of real user prompts to help you understand what people are typing in. But then the next big step was to say, "Okay, cool. We've got the richest data. We have the most breadth and depth in our data and insights. Now, let's give that to agents that can do the work for you." So now, it's just a case of like, yeah, just expand and expanding.

Nakul Mandan:

How does a marketing team look like even at a company like Profound? What are the human roles still? So you talked about marketing engineer. What are the other human roles in your marketing stack when agents are doing so much work? Is it just evaluation? Maybe just to touch on this, I think your head of growth, Nick, he mentioned that when he was at Loom, his marketing team had 30 people at a similar stage and now you guys have only eight people. So, yeah, can you talk about what does the marketing team of today and the future look like?

James Cadwallader:

Yeah. I mean, we dog food our own product as you'd hope and expect, so we produce a lot of content. There's some of my favorite agents that we've created are like competitive insights. So we use agents a lot to maniacally monitor every competitor in our space. We always know what competition is doing and that's effortless now and wonderful as an experience, because it just spits out into Slack. Again, to reference the dumbbell at the other end of that dumbbell though, we do some very human-centric stuff. So we get tons of value out of events. We have a big user conference that we've now done three times. We did the first one in New York last year, then we took it to London. The first one I think was a couple of hundred people, then the second one was 450 people. Then we did our most recent one in SF that was 700 people full house. We had OpenAI there. We had the folks from Figma, and LinkedIn, and Reddit. It was a really big event. Arguably, I'd say maybe one of the biggest AI marketing events of the year. Yeah, we invest into that dumbbell basically. We do very human-centric stuff and then we create lots of content that's aimed at agents.

Nakul Mandan:

In your organization, what is the role of the marketing engine and what is your marketing engineer doing today?

James Cadwallader:

It's a lot of what I said. So building and deploying agents that can help us create content, a lot of monitoring, all the stuff that humans have done, particularly in that kind of growth, demand gen, sort of domain where we don't spend much time now creating written content. It's very easy for us to do that, but what we do spend time doing is providing these agents with context.

Nakul Mandan:

A lot of marketing dollars end up going towards agencies, all kinds of things. Ad agencies, other kinds of content marketing agencies. What of that world survives this shift with agents doing a lot more work? What timeline do you think that shift happens? Does that industry even survive if agents do all the work?

James Cadwallader:

I think people underestimate the relationships that large enterprise brands have with these big agencies. I believe at least for the foreseeable, we're going to see marketing teams rely on agencies to build and deploy these agents that we're talking about. With business, so much of this is timing and I think it would be a big mistake to believe that agencies will move out of the picture or should move out of the picture, I think is a big mistake. I think you play alongside agencies and I think ... Profound is used by the biggest agencies on the planet right now, because agencies need the best technology. Their space, their category is the most competitive it's ever been. They want to show up to their clients and say, "Hey, we're using the best technology right now. We can help you understand and control how you show up in these new systems. We can build agents that can do all kinds of wonderful things for you marketing that was not possible before." So I think the dynamic hasn't changed that much.

Nakul Mandan:

Are you seeing a difference in which B2C marketing organizations versus B2B marketing organizations are adopting AI?

James Cadwallader:

Our demand is orthogonal. We have demand from every brand under the sun. There's no rhyme or reason. We work with the biggest banks in the world. I mean, we work with the biggest bank in the world. We work with the biggest airlines. We work with some of the biggest B2B software companies, pharma, you name it, fashion and apparel.

Nakul Mandan:

But is there a difference in the way they are adopting it or it's pretty similar in your experience?

James Cadwallader:

I'd say it's more tied to the culture of the business than the category of the product. I mean, maybe sure, if it's a very regulated category, finance or healthcare or something, they naturally move a little slower. But yeah, I mean, we work with giant Fortune 10 brands who are a lot faster than you'd expect. They get this image, and especially here in Silicon Valley, there's this idea that the enterprise is slow and done. And my experience is, it sounds like sycophancy, but my experience is quite the opposite. They're very leaned in.

Nakul Mandan:

Is there almost a moment of existential thread that they are feeling that, "Hey, if we don't move to this new world, somebody else will write our story on ChatGPT's answers and we need to make the shift as soon"? Is there a higher degree of urgency that you're sensing there?

James Cadwallader:

Yeah. I think if you're a leader at a giant Fortune 100 company, let's face it, you're probably in your 40s or 50s, which means you probably saw dot-com, you probably saw mobile, you probably saw social and they know that this one is a really big deal, and they're not going to miss it. And they understand that for these brands, the level of investment we're talking about pales in comparison to the downside of missing the biggest platform shift in the history of marketing.

Nakul Mandan:

So actually, yeah, it kind of is a segue to my next question was going to be more about Profound as a company. You guys have scaled absurdly fast, right? Under two years, these many customers, some of the biggest brands using you. You're a unicorn already within 18 months I think of starting. What did you get so right in this story? Was it a product market fit thing? Was it just better category creation, better sales execution? What did you get right? What explains the speed at which this is happening for you?

James Cadwallader:

Got very lucky, an amazing co-founder. I've worked with co-founders across various businesses in my career. My background is, it sounds kind of cringey, but serial entrepreneur. I've always started my own businesses. I've predominantly had co-founders and there's a co-founder ... What do you call that? Like a co-founder market fit or something where we click very well with each other. Dylan is good at everything, I'm not. I'm good at everything, he's not, and it's perfectly complementary. And then you combine that where we have this founder market fit with the problem space. I've worked with the marketer ICP my entire career. Dylan is one of the most ferocious product minds that I've ever met. He's got this velocity that is rare to see in engineering. Yeah, he moves too fast sometimes, but in our space in this category, it's a very competitive, ferocious category. So I believe that we're the most ferocious founders in our category and we've always been like that. When you're coming up with an idea, it is the most lonely, awful experience because it's like being in the deep blue ocean, there's no island in sight. You're just swimming around trying to figure out what you're trying to do. And even then, Dylan and I were working six days a week in office before we had an idea. It's like hardcore from day zero or hardcore from day minus one. And what that meant is we attracted a founding engineer, Charles, who's also very hardcore. And then we attracted the next person and we built this culture from day zero of people that ... I'd say that the common theme, particularly in the founding team of Profound is like folks with a bit of a chip on their shoulder. I'm not a GSB guy. I didn't go to university. Dylan's not some MIT research brain, so I think we've got something to prove. It creates this fire in us that just mean that we push super hard all the time and that exists in our culture today. We've attracted brilliant people like that. Yeah, Profound's a nice place to work, because if you're ambitious and you're smart and you're good at what you do, you want to be around people that really want it.

Nakul Mandan:

Where does that translate most? Is it from the speed at which the demand has shown up, did it translate more to iteration in the product? Did you always start with this current idea, or it was more like you had such a great culture of iteration and speed that you always knew you were going to attract to where product market fit would end up happening?

James Cadwallader:

No, I didn't always know anything. Definitely not. We've been extremely lucky finding this moment in time and finding this market.

Nakul Mandan:

But was iterative, or you knew what you were building at least?

James Cadwallader:

No, we would sit in front. I would drag Dylan into South Park Commons, which is where there's an accelerator where we're working out. So it's like basically an empty loft studio in New York City. I'd drag him in on a Saturday and he's got that engineering gene where all he wants to do is like pull his laptop out and write code. I'd say like no laptops. We would sit painstakingly in front of a whiteboard with Post-it notes and Post-it note, Post-it note, Post-it note, Post-it note, Post-it note. And all we're thinking about was tailwinds and we'd go off and research the craziest ideas. We were exploring lithium mining for ages because Dylan had this fascination with maps, he still does, so we are exploring that. And as we were exploring these ideas, the penny started to drop around, "Well, hang on." As we are researching these ideas, I've stopped using Google search, I'm using Perplexity all the time, but this is different to how I'm using ChatGPT. Why? Why am I using Perplexity and not ChatGPT? It's because Perplexity is not using the large language model to generate the response. It's using the large language model to synthesize from the web. So we just went all in on that. Yeah, it was one of these things where it's ... Our space is obvious today. When you tell people, they're like, "Yeah, that makes so much sense." It felt like that at the beginning of 2024 as well, but it was just very strange because people weren't talking about it.

Nakul Mandan:

Did you feel like you had that early moment where you suddenly had a light bulb moment that, "Okay, we've been writing on ideas, this is the idea we've been iterating on last month or so or last six months." This is bigger than everything else. This is the moment. Was there a light bulb moment, or you grew into it? You grew into the current company, or not?

James Cadwallader:

I mean, there were people that told us the idea sucked. I remember this one guy whose name I can't name, but Dylan, my co-founder, would laugh because we've had a private joke about this guy. He was just the most negative dude at SPC. And yeah, of course, when we had the idea, he came over and he had like a hundred reasons why this wouldn't work. It's really hard in the early days. Ideas are fragile. I think it's Jony Ive that's got a very good saying about ideas are like babies. When they're just born, they're very fragile and you have to be very careful and nurturing with an idea. And if you drop it, these things can shatter. If you try and poke holes in them, they will easily ... You can always pull apart an idea. It requires this very specific balance of optimism, almost you might say blind optimism, but with a bit of self-awareness to know that, "Okay, this is just a stupid idea." And yeah, we had a few people telling us in the early days that Profound was a really stupid idea and I nearly listened to them. I've actually never said this before, this is interesting. So my co-founder, Dev, so my co-founder of my previous company was visiting New York. He lives in London, which is where I'm from. That's where we ran the previous company out of, and he was visiting New York and we'd gone for a coffee or something just to hang out and he was interested in what I was working on. And I remember I said to him, I pitched him basically the idea for Profound and he said, "That's a really good idea." And it gave me, I don't know why, maybe it's because he was my former co-founder or something, but also Dev knew me very well and I trusted him that he would be critical and tell me if the idea sucked. Yeah, his vote of confidence kind of gave me a little pep up.

Nakul Mandan:

You're also an AI native or you're an AI native company. It's not just an AI native product. So outside of marketing, are you using AI in every function? Can you talk about how AI native your customer support, sales? And engineering is easy, it's probably all AI now, but yeah, can you talk about how you guys are using AI across the organization?

James Cadwallader:

Engineering is interesting honestly as well, because six months ago, we were almost exclusively a Cursor shop. Now, I'd say it's predominantly Claude Code, Codex and we use Factory as well, which is really interesting. The fact that a lot of our team uses Factory, which for anyone who's not familiar is like a Series B startup who haven't trained their own foundational model, they built a very good harness. It kind of breaks my mental model because I'm like, "Oh my God, how is this Series B company competing with the most insanely scaled companies on the planet right now?" So, yeah, we use Anthropic, use Claude Code, Codex and Factory on the engineering front. The rest of our business, yeah we're incredibly AI build. It's kind of like not a big deal though. We launched the company in August of 2024, so everything-

Nakul Mandan:

It was born in that way.

James Cadwallader:

Yes. Do you know there's the Bane ... Batman, Bane-

Nakul Mandan:

Yeah. "I was born in darkness."

James Cadwallader:

Yeah. Yeah, I was born in AI. Yeah.

Nakul Mandan:

It was in The Dark Knight Rises movie. Yeah, yeah.

James Cadwallader:

Yeah. We were born in this world of AI and using AI in the workplace. With that said, we lean in heavily. I judge people that don't use it.

Nakul Mandan:

Does the recruiting process have almost a test for AI nativeness or AI fluency?

James Cadwallader:

Well, no, but the engineering challenges, they use-

Nakul Mandan:

No, beyond even engineering.

James Cadwallader:

Oh, right. We will ask about it. I think we'll ask how people use these systems. It's kind of just expected. I don't know. If you're hiring ambitious, hungry people, it's kind of weird if you're not using generative AI. We've built an internal platform, so internal tooling that we've called Profit. It's like a classic generic internal tool name, but it pulls from all of our Granola recordings. It pulls from all of our Gong recordings. It plumbs in from Salesforce. We use Dust, which is a great tool to pull everything together, but it all pipes into this one monolithic platform called Profit. And our CS team uses that, our sales team uses that, management can use it. It essentially is becoming our central intelligence.

Nakul Mandan:

You were a co-founder at Kyra. How is your life as an exec different today using AI than when you were running Kyra?

James Cadwallader:

You just get more shit done. I take a customer call, I get back to my laptop, I ask Claude to draft the follow-up and it pulls from a knowledge base that I've created. It pulls from the Granola transcript, from that call. Six months ago, I would spend maybe five minutes editing that email. Now, it's like five seconds. I'd say seven times out of 10, I don't even need to edit the follow-up. I literally say, "Draft the follow-up to John Smith," who I just had a phone call with and I go into my Google Drafts and the email is there and it's perfect and I click send. And in my last company, I shit you not, if you're meeting a big enterprise customer, you might spend an hour on that follow-up, and now it's five seconds. If I want to know what's going on in the business, I can query profit. If I want to write a memo, I can ... All of this stuff is just an order of magnitude more time.

Nakul Mandan:

Has your sales team also started adopting it pretty aggressively in the way they follow up or even do pipeline generation stuff? Do you guys use AiSDR kind of stuff?

James Cadwallader:

It's weird. I think the companies ... And it's funny, because via Sequoia, we work closely with Brian Halligan, who's the founder of HubSpot.

Nakul Mandan:

Yeah, of course. Yeah.

James Cadwallader:

And Brian's become this bastion of founders, especially on the East Coast, but all over. All the hot AI startup founders all love Brian. He's the kind of founder, mentor at this point and he's got this very interesting take, which is the younger, more upcoming hot AI startups are a little more skeptical. In the same way that we will judge you if you don't use AI, if you come in and it's just AI slop, you get lots of judgment here. We did our quarterly planning, just with the leadership team last week and the prerequisite was no AI slop.

Nakul Mandan:

Yeah. So how do you maintain quality? It's just the manager in each department is responsible for, "Hey, I'm not going to tolerate AI slop"? Is that how quality is controlled?

James Cadwallader:

Yeah, kind of. And you call people out. I will call people out if someone sends me something that's clearly ... It's not about not using AI, but you know the type of slop that I'm talking about where it's just this rambling, unstructured nonsense.

Nakul Mandan:

Yeah.

James Cadwallader:

Yeah, I'll call someone out. I'll be like, "Hey, don't send this slop." It's not useful.

Nakul Mandan:

It's an interesting thing, because it kind of also begs the question of, what does great human work look like going forward? Because in the prior world, we all had our definition of that, which could have been thoughtfulness in those responses and emails, follow-up times and urgency and other things. Maybe do you have a better answer for what does a great human marketer look like going forward?

James Cadwallader:

I think a great human marketer is always going to need to take risks and have original ideas. Our ability to generate original thinking that really is sort of a function of our human experience of planet Earth, it will still continue to be very important. There's a smell in this building that I like. I don't know if it's the ... It's like a woody smell, right?

Nakul Mandan:

It might be, yeah.

James Cadwallader:

Because there's all this timber everywhere. It's got a nice smell in this room. And if I were to go and create a piece of content now, I'd say, "Hey, I had a great interview with Nakul. It was really thoughtful and it had this nice woody smell in the building that I appreciated." ChatGPT would never know that. So it's our ability to take insights from our human experience and transmit them to the world through the internet or through AI.

Nakul Mandan:

Let's talk about the culture at an AI native company like Profound. You've touched on some aspects, but maybe what are the cultural pillars of Profound as a workplace?

James Cadwallader:

I've said a few times that our culture is the best product we've built. It's the product that I'm most proud of. I think culture at the extremes is very easy. I can do the rah-rah 9-9-6, let's all go kill ourselves thing. I'm very capable of doing that, and it sounds like bravado, but with the caveat of if you're not Elon, most people end up imploding their company that way. On the other end of the spectrum, there's the warm, fuzzy, we're all a happy family, work-life balance thing and that's very easy as well. So the extremes are easy. It's easy to be at this end of the spectrum and it's easy to be at that end of the spectrum. The difficult thing about building culture is to capture nuance. And we've built this culture which is insanely hardworking. It would be a light to say otherwise. A lot of us are in the office six days a week. You go into the office at 10:00 PM, 11:00 PM there will be people there basically every night. At the same time, we've built a culture where you can leave at 6:00 PM and no one bats an eyelid. The way I think about it is I'm default hardworking. If I were stacking shelves at Walmart, I would work really hard. That is just in my nature. And I think about the type of place where I would want to work as a default hardworking person, nothing would piss me off more than having a manager or a boss kind of like second guessing why I had to leave. "Hey, dude, my mom's in town. I've got to go for dinner tonight." Fuck off. I don't need someone breathing down my neck. And we're trying to create this culture of people that they're default hardworking, but they can work hard in their own ways. People in the office have families and kids and they go offline from 5:00 till 8:00 and then you see them pop back online and they're hardcore, but they're in their own way. So it's like working hard without it being performative and stupid. I think I'm able to do that now because I'm 36, I'm a little bit older. When I ran my last company when I was in my mid 20s, I didn't get that nuance. I was just like pedal to the metal. It was a bit more performative. So, yeah, we live in this nuance now. Our culture is very specific to us and it's something I'm very proud of. And I think frankly, I think there's a saying that said culture is how you manage people when you're not in the room.

Nakul Mandan:

What kind of person thrives at Profound?

James Cadwallader:

Yeah, people that are default hardworking. Yeah, I probably value, or frankly, I do value hard work or people who work harder over people that are super intelligent.

Nakul Mandan:

Are there other criteria though that go into beyond work ethic, for sure, that makes a person very easily a Profound person or ...

James Cadwallader:

Yeah, like curiosity, like domain mastery. We look for people who are very certain about what they want to do. A big red flag for me when hiring is when people are kind of generalists and they're like, "Hey, I want to do a bit of this, but also a bit of this. I'll start here, but maybe go here." What we're really looking for is like, "Cool, if you're a front end engineer, you want to be the best in the world." And you are learning from the greats, you are researching, you're doing your homework and you are obsessed with the idea of total mastery of craft. And that's not just an engineering thing. You could say the same thing for sales, for customer success, for design. So, yeah, I think we look for masters of craft and people that are default hardworking, that's what they do if left to their own devices. This sounds super trite and throw away, but I don't like to work with assholes. Neither does Dylan, my co-founder. There was someone that we had in ... I mean, I go and timestamp picks, it might be obvious, over the last month. This guy was super credentialed, been to all the fancy schools, everything checked out. I just thought he was obnoxious. Yeah, he wasn't curious. There's a certain energy that we try to ... We want people that are kind of humble.

Nakul Mandan:

The other piece about you guys is when a company scales this fast, as fast as you have scaled, things also have a risk of breaking, right? Communication breaks down, recruiting speed leads to suddenly the quality of people can drop because there's always pressure on hiring more people. How are you ensuring, as a CEO, that the quality doesn't drop as Profound still meets its ambitious growth goals?

James Cadwallader:

I'm having to grow myself. I'd say it's like I've had to be a different type of CEO multiple times on this journey already and considering it's only been 21 months. When we first launched this thing, I was hand-to-hand combat in the trenches doing this out, doing the things. And then that starts to change maybe around 30, 40, 50 employees. It's like, okay, now I need to inspire a bit more or lead a bit more. And then you get to a hundred plus and it's like, "Okay, now we need a leadership team. We need a team of leaders." And then you're looking at how do you motivate your leaders and ensure that your leaders are culturally and morally aligned with me and Dylan. And then now, as we're crossing the threshold of 200 people, it's kind of thinking of the business as we scale. We've relied and been very fortunate, especially because we're very much an in-person culture. We've been very fortunate and then we could rely on the sort of tacit transfer of ideas and you sit next to Sally and then Sally tells you something and that's kind of how we've learned and grown as a business. But now it's like, "Oh, shit, that's breaking because there's people in different offices now. There's people that ..." Our offices are bigger, so sometimes you don't even get that water cooler conversation. So it's like actually building the systems that can help us scale to maybe 500 or 1,000 people. I was with one of our investors, Keith Rabois at Khosla Ventures. We were with him last Thursday and he said in classic Keith, kind of look you square in the eye, "We're about to hit a very awkward phase." He's like, "Yeah, 250 tends to be where things get really awkward." The OGs start to get unhappy because they're like, "Oh, it's not like it was when we started. Now, there's systems in place." But then you have new people who are joining and they're like, "There's not enough systems in place." And you're trying to kind of like both are right and wrong in the same vein. It's like the OGs are right. We want to maintain that original violence and aggression that we had from day zero. But then the new people are right in that, yeah, we're not a 20-person company now. We do need systems to orchestrate what is becoming an army of people. So reconciling those two things I think is like the new challenge.

Nakul Mandan:

Are the CEOs that you go to ... You have an amazing investor base also of ex-operator CEOs, but are there founder CEOs you find who are your go-to if you're stuck on an operating or leadership issue?

James Cadwallader:

Listen, there's no shortage of content on the internet. I am like a pretty ferocious consumer of podcast content. I've watched both of Winston at Harvey's podcasts. He did one with Halligan, he did one with Harry Stebbings. I just watched one with the CRO of Legora while we're on the subjects of legal tech. Brian Halligan's got a great podcast, Long Strange Trip. He interviewed Jack Dorsey recently. So I'll munch up all of that content. I'm pretty insatiable when it comes to that type of learning. I've never met Winston at Harvey. Maybe there's things that he would tell me and there's secrets that I would glean from a private off-record conversation with him. Sure. But I reckon the 80/20 I can get from listening to two and a half hours of him talking about his business.

Nakul Mandan:

Are there elements of you from Kyra where you were ... You were a brand marketing exec, let's call it that. That actually there were instincts that you carried from there to Profound that you had to curb and say, "Hey, that worked for Kyra. This doesn't work for an enterprise software company, especially in the age of AI."

James Cadwallader:

One of our investors, Ben Braverman, who was the CRO of Flexport, he now runs a firm called Saga VC. Yeah, he told me a story about Flexport with him and Ryan. He said, "Prior to Flexport, we never knew what it was like to be at an A++ company." And how it's very surprising the quality of talent you can attract. And I think being in this moment with Profound where we're, as you said, it's gone so fast, we've gone to unicorn status in 18 months, we're backed by Sequoia, and Kleiner, and Lightspeed, and everything. The revenue numbers, which I won't share, have all gone vertical, everything's crazy. And that has required a recalibration from me on understanding the caliber of people you can attract and pull into a business like this. And it was very interesting having that conversation with Ben because he said they had exactly the same thing at Flexport where there's this awakening where you're like, "Oh, shit. We can attract quite literally the best in the world." There's people who work at Profound now. I mean, like Mark Ebert, who's our SVP of revenue is a known quantity, one of the best sales leaders on the market right now. He ran revenue at a company called 6sense and he did a very good job, built a reputation for himself and people follow Mark. And it was a bit of an eye-opening experience for me being able to build this company, which to me is still just a baby and have people like Mark. Arguably, Mark's most precious asset is his career, right? He's spent so much time and energy building his career. To have Mark come and take a bet on his career with Profound is like a wild, required like a big recalibration for me.

Nakul Mandan:

So from the outside, it does look like Profound is a straight up into the right journey, but you and I also know almost every company has a thousand fires it had to dowse. Have there been tough weeks, tough days at Profound?

James Cadwallader:

Listen, we've been very lucky and I say this to the team unfiltered. I'm like, "Hey, I will stand ..." We do a weekly all hands every Monday. It used to be at 9:00 AM Eastern, but it's now at midday Eastern because we need people on the West Coast to wake up and get into the office. But I've said on a few occasions looking around the room to a team of like 200 people now and saying, "Hey, I'm very curious to see who in this room sticks out when things actually get hard," because right now we're playing the game on easy mode. At some point, we will face adversity that we probably can't even fathom right now, and you have these tests. And in this weird fucked-up masochistic kind of way, I'm actually excited for those tests to see what kind of CEO I can be. Be the man you want to be. And we're going to go out ... It's inevitable. I know that's going to come for us and I'm trying to kind of harden the team now and help them understand like, "Hey, this is the easy days. This is when we have the tailwind of all tailwinds blowing in our direction and we built this category-defining brand that everyone's excited about and interested in and we're the hot girl at the dance right now." And that might not last forever, and then at that point, it comes down to our discipline and our strategy and how we carry ourselves through adversity. But to answer your question, I mean, it's all hard. There's sometimes, like weeks on end, where I literally don't see my fiance or my wife now. I just got married.

Nakul Mandan:

Congratulations.

James Cadwallader:

The inside joke with her is that some weeks I will go an entire week and only see my wife in her pajamas because she's in her pajamas in the morning when I leave the house. And then by the time I get home, she's back in her pajamas and we both had a full day. I'll quite literally go an entire week and I've only ever seen her in her pajamas. Sometimes it'll be a Sunday and I see her in jeans and an outfit. I'm like, "Wow, you look great. I haven't seen you look like this." And she's like, "Yes, because you're never here." So it's like super taxing on some fronts. I'm spending way too much time on airplanes right now. Things go wrong all the time. Our platform went down today because I won't say who. You can bleep this out. (beeps) went down today, which is one of our upstream providers and fucked up everything, and the whole platform tanked for an hour and a half and we've got some of our biggest customers messaging us. "What do we do? It's down." We haven't built the redundancies in, but you tend not to at this stage of company.

Nakul Mandan:

Yeah. What are you most paranoid about right now?

James Cadwallader:

I'll give you what I'm not paranoid about and then what you might think I should be paranoid about. Everyone's slightly jittery about Anthropic and Claude and software going to zero. I personally, as we've spoken about, I've worked with the enterprise my entire career. I don't think every company on the planet is all going to use one piece of software called Claude. I just don't think that's the way things are going to shake out. I don't think it's wise for enterprises to anchor themselves to one lab and say, "Hey, we are," insert biggest bank in the world, "we are now tied to Anthropic." We are an Anthropic shop, because we've seen how fast this space evolves. In three months, it'll be OpenAI is now the best intelligence. So you need interoperability with the intelligence if you're an enterprise brand. There's this necessary optionality that enterprises will need. And I also think at the application lane, and obviously this is a very biased take, but I think legal teams will have Harvey or Legora. I think marketing teams will have Profound. I think finance teams will have Rogo. I think we're going to see application layer winners across the stack and that will be powered by the intelligence. And do I think that a giant enterprise brand is going to vibe code everything with Claude and do everything in-house? I personally don't see it, but maybe I'm wrong.

Nakul Mandan:

But there's something about future competition or anything that makes you feel this is what you're watching.

James Cadwallader:

The thing that I'm most paranoid about is a boring answer. It's not new or AI related. It's just, "Hey, as we ..." I mean, we had a hundred people in January. We're now at like 210 and it's May. When you add that many new people so fast, the culture gets diluted. I was just in our San Francisco office and most of the people that haven't been there for longer than six weeks. I'm like, "You don't really understand our culture. I've not hung out with you." And our ability to scale our culture, yeah, that worries me. I think that's hard. And you're in this kind of constant place of tension where it's like, "Cool, we have to scale." If we want to hit ... We've got an absurd number that we're trying to hit this year. It's just a ridiculous ARR number by end of year. I'll tell you the number, and you can bleep it out. We're trying to hit (beeps) ARR by the end of the year.

Nakul Mandan:

Okay. Wow.

James Cadwallader:

For us to hit that number, there are just laws of physics. You can have all the AiSDRs in the world, but there is just laws of physics that surround our ability to hit that number, that predicate our ability to hit that number. You have to have X number of ramped AEs in seat. And then if we're going to add X many logos, you have to have the CS team to support it and the product has to be here, here, here. And that is all a function of people. So we have to add X many people.

Nakul Mandan:

Yeah, that's why culture in recruiting are the risks with such hyperscaling.

James Cadwallader:

Yeah. I read the Twitter slop where people were like, "Oh, you need to do it with a team of 10 now." Sure, if you're running like a PLG thing and then it's all run rate nonsense anyway. When I'm talking numbers, I'm talking MSA enterprise contracts is what we go after. And for that, you need humans. And sure, they're more efficient humans. Sure, we use Gong to review every single one of our sales calls and plumb everything into profit and this and that. But fundamentally, you need the humans who are ramped in seat who can take the calls.

Nakul Mandan:

So next up is our quick fire round. One marketing workflow you'd kill at every Fortune 500 company now.

James Cadwallader:

I'd say building reporting decks. I think marketers spend way too much time building decks that they can use to update their managers and now you can do that with AI, with Profound.

Nakul Mandan:

Biggest lie marketer tells themselves about AI.

James Cadwallader:

That it's one silver bullet that's going to solve everything, that you just press a big red button and your marketing is done.

Nakul Mandan:

A brand whose AI era marketing has been excellent.

James Cadwallader:

I think the Expedia team are particularly good. They've got a very sophisticated content marketing function. They have hundreds of thousands of webpages or millions of webpages. It's a giant machine that those guys have built.

Nakul Mandan:

Best startup book that inspires you and is one that you really have taken your operating principles from.

James Cadwallader:

Frank Slootman, Amp It Up.

Nakul Mandan:

Worst startup advice that is getting repeated as gospel these days around AI and all the other things.

James Cadwallader:

AiSDRs are stupid.

Nakul Mandan:

AiSDRs are stupid, great. The AI app that you love the most, which is not ChatGPT, Claude or Profound.

James Cadwallader:

Dust is very good for us.

Nakul Mandan:

Okay.

James Cadwallader:

Another Sequoia company. It's like our intelligence layer. It's a great product.

Nakul Mandan:

So last question. Let's say you're talking to a Fortune 500 CMO today. They know they need to change everything about how they run their marketing function. Do you sit down with them for an hour? What would be your recommendation on their 90-day plan to revamp around Profound and the future of marketing?

James Cadwallader:

It's the human transformation piece. The failure mode that we see is exec gets really excited. Team below doesn't get much of a say so it gets kicked down to the team to go and execute and they've not been part of that journey. So I think that the CMO's role ... Because I think the CMO is correct in that, sure, this is a big moment in time and you do not want to miss it. But they need to bring the people on their team along on the journey with them. They need to bring the marketers who are going to be clicking the buttons and doing the work and give them the context as to why this is important, why we're building this muscle, why we want to use agents to make us more efficient. Why we care about what ChatGPT says about our brand and products and services. So I think ironically in this world of generative AI, the biggest challenge for execs today is not a technology challenge, it's that human transformation challenge.

Nakul Mandan:

Have you found something across the experience with Profound that a tactic or a strategy that is working for CMOs to bring their entire team along on this transformation?

James Cadwallader:

We physically go and do offsites. So we did one yesterday with the biggest bank in Canada and we went up to Toronto and spent the day with the team showing them what to do. These tools are extremely powerful, but they are a new category of technology. It's not intuitive. People don't get it straight away. Even when it's like we have all kinds of ergonomic functionality now where you can just talk to our platform and it will give you the answers. You'd think it would be intuitive, but it's not always the most intuitive and I'm sure we don't do the best job all the time in our UX as well. Yeah, it's really understanding that we are not just a vendor to these big brands, we are a technology partner. We will help you navigate the transformation. You have to transform people who have been doing the work very manually for decades into people that will now be conductors of AI, conductors of agents. And that's the hardest challenge I'd say today.

Nakul Mandan:

Awesome. Well, James, this was amazing. Thanks for coming on.

James Cadwallader:

Thanks for having me. Yeah, appreciate it.