Latest / Unharnessed / Unharnessed Ep. 4: Where the Agentic Money Flows in 2027 with Eric Franchi
Transcript
- Eric: Welcome to Unharnessed. Let's go boys. Moffie: let's go. Eric wants to get he's a professional, wants to get right in. He's like, listen, you guys have no clue how to run a podcast. I am a co-host of two podcasts, maybe three. I don't know. What how many podcasts now? And I will school you. I Eric: I think I just took this one over. Moffie: will school you chumps on how to manage this thing. I love that. We actually we need it. all right. Well, you you kicked it off. Welcome to another edition of Unharnessed. We have the Eric Franchi from Eperium. Welcome, Eric. How you doing? Eric: Doing great. Great to see you guys. tell me about the the origin of this of this name Unharnessed, outside of the fact that it describes both of you guys quite quite well. Moffie: Well, unharness, as we know, the harness is an integral piece of building agents. So if you're unharnessed, we're pulling back the harness to show you how agents are being built in ad tech on the buy side, on the sell side. I do think someone has some kind of hammering going on. I don't know. joseph hirsch: That's hammering on my side. Eric: Definitely do. joseph hirsch: Yeah, there's a little construction. I hope that's not a big deal for the pod. Moffie: Fantastic. This pod is off to a Eric: I got a I got a I got a good editor for ya. Don't worry about it. Moffie: This pod is off to a great start. we've got the the hammer in the back. But yeah, that's that's the origin. And as you said, I would say Joe is probably more notorious as a as a character in the space. I I aspire to be. I aspire to be a character that's that's that's known in the space as being a little little unique and different. joseph hirsch: Yeah, I mean honestly my Eric: I think you are absolutely joseph hirsch: my LinkedIn game had gotten really weak. And when Moffi came to me with this concept, I was like, this is exactly what my career needs. I need the spotlight back. I need a microphone. I need to be visible. The world needs more. my fan group was kind of I guess yearning, if you will. and I wanted I wanted to meet them where they were. and I feel like this is a perfect medium. to express myself and the views of me. and you know, sometimes I feel like people get a diluted version. And when you're on the pod, when you're unharnessed, if you will, you get a hundred percent. Eric: I love it. Moffie: And let's be honest. Let's joseph hirsch: so one one one question. I hey, I w I wanna talk, dude. Will you chill? Eric, I wanna know what tools you use every day that are AI tools that assist you in your day to day life. Eric: Absolutely. Hey, before we do this, quick disclosure. my fund is an investor in Swivel where Joe is founder and CEO. joseph hirsch: And and what do you think about that investment? Eric: I have extremely high hopes for that investment, and we are thrilled with the progress. joseph hirsch: Me too. I'm glad we're the same page with that one. the reason I ask, I feel like Eric: Yeah. I mean it's it's top investment. joseph hirsch: I feel like everyone that we've interviewed, either on accident or just in the in the nature of our conversation, discloses some net new tool. And I've found when talking to other people about the podcast, there's a lot of tool discovery. and I find the tool discovery is a useful thing in the pod because not only does it I I feel like it's bettering people's lives. and yeah, I I personally have have had some tool discovery through the pod and maybe maybe you're using something out there that that I don't yet know about since everyone is using something different. And all these tools, by the way, are somewhat analogous. I feel like a lot of tools can do the same thing as the adjacent tool. And then it just comes down to personal preference, which is also interesting to me because I think it's fascinating which tool people choose. Like, do they choose because like They saw it on LinkedIn? Did they hear it on the pod? Did they like were they marketed to anyway, you get the idea. Eric: Yeah, totally. I'm I mean I might disappoint you. but let's see. So most of the professional use cases for me, again, I'm in a different, I think, category than maybe, you know, some of your guests, and certainly you guys. I mean, it's all Claude. I use Claude in every possible way. And I almost think of it as like, you know, a team of just as many analysts as I need. To do what I do as a VC, both kind of like, you know, the analysis portion of the job and other parts of the job. and just to do it like quicky, quick and at a higher level than I possibly could do, and potentially at you know what a team of analysts can do. So, so Claude for like basically everything. I'm increasingly finding that the Gemini capabilities in Google workspace are somewhat binary. Either it just doesn't work and I get frustrated, or it's like. Absolutely amazing. So sometimes, this is not every time. Sometimes I'll get an email asking for a meeting, asking for something. And sometimes Google will draft the reply, pull from my calendar, pull from like my recent responses, and just auto-complete it absolutely perfectly. And then other times it doesn't. We're recording this on Tuesday, September twenty-second. I mean, basically, Muse has taken over. the world over the course of the past couple of days. And I've been spending a lot of time on Muse. And I can imagine some of those, gosh, I wish Google can do this to just like handle a lot of the front end stuff. I'm already experimenting with with it on on Muse. So my stack is super simple. It's just like it's basically clawed. joseph hirsch: Cloud Plus Muse. I haven't tried Muse yet. I'm over here. I'm I'm like a big town convert. one of our prior guests had had turned me on to it. and I'm I'm just enamored with, you know, the the functionality. I'll just give you like a quick use case that that happened to me earlier. someone on my sales team had sent me 10 different prospects that he wanted me to give him a contact for. I copy and pasted that list. I was on my phone. into WhatsApp to my town WhatsApp Messenger, and then it returned from my email like three to five contacts at each one of the prospects. And then I pasted that into Slack and it it happened in like the course of 90 seconds. And I felt like I had, you know, kind of lifted a large weight with very little effort. but I assume Muse does like similar things, although I know that Muse also might like get you a reservation at your favorite restaurant. Eric: Yeah, yeah. I mean the use Moffie: Yeah, let me bring this Eric: cases. Yeah. By the way, hold up. Joe, let me give you an invite to muse so we can both get a billion tokens. Please. joseph hirsch: Yeah, I listen, the more tokens that you have, the better. Eric: Yeah, and but you'll get too. Moffie: All right, I need to butt joseph hirsch: I sh Moffie: in. I keep getting talked over, and I'm the one that talks over everyone. So I want to come out as a total muse head. And basically I'll Eric: Yeah, I imagine. Moffie: give you an anecdote that is amazing. I use Muse. It does all of the actual research for this pod on harnessed, and then it's actually starting to do a little bit of outreach to net new guests because I'm doing all that because we are not. funded. and one thing, we're gonna be in New York for ad week. And we have a guest that we're gonna meet with and I wanted to find a podcast studio to record in person. And it suggested a bunch of studios near where I'll be staying. But then it also saw in my personal email, Muse is not connected to my work email, just full disclosure, but in my personal email it saw that I went to CUNY J School for grad school and CUNY has A audio podcast recording studio. And it's like, hey, do you want me to reach out to your alumni to see if we can actually get that studio for you? And I was like, holy shh Muse, absolutely. And that is where we're gonna be doing it. So that's pretty incredible. So these agents going to proactive, actual driving insights and taking action is really amazing. And it's where this needs to be going, even in our space too, of course. Eric: Agreed. And gosh, like the the marketing use cases end up being like pretty mind blowing as well. Moffie: Yeah, it's unreal. joseph hirsch: What what kind of marketing use cases are you seeing for something like Muse out there in the wild? Has has is there a startup that has crossed your desk so far where you're like, This is the next breakthrough that is combining what I'm seeing in the real world with something like Muse or Town or Instinct with marketing? Eric: No, not yet. I think what Jonah Goodhart is doing with Mobian is super interesting. They're not a portfolio company. Jonah's a great guy. and you know, we we've worked together various capacities over the years. what Mobian is doing in terms of giving media companies the ability to market to agents, I think is net new. And it could be potentially disruptive and could potentially be a new category. But I haven't seen anybody jumping on this like agentic. front end use case for marketing yet. It's only a matter of time though. Like shop you're you're able to connect your Shopify store to Muse to allow for like seamless transactions. They blocked Amazon, right? So like you can you can just like see how this whole thing plays out. it's just a matter of time, I think before you know new categories are created. But in the meantime, there's tons tons of stuff going on in Adtech. Moffie: Yeah, I Shrute Tech Gri Ben Thompson even pontificated, which I agree with. you know, Amazon blocking Muse is potentially even a good thing. Maybe it was Eric Suford. I get my really smart strategists confused sometimes. And there'll probably be a play if they can work on a deal where actually Muse will start advertising Amazon products within Muse if they can work on a deal. Just it's kind of a no brainer. Eric: Exactly. Exactly, right? Yeah, yeah. No, it's it's it's super exciting. The the closest thing, you know, to like, you know, net new formats and and these next gen channels, you know, there's companies that are trying to reimagine what conversational ads might look like, you know, in the wild on a publisher site, you know, putting a widget, putting a box out there. There are companies that are obviously providing some sort of access to Chat GPT. because they've opened it up to, you know, a bunch of partners. And then there's a handful of companies that I think are really pushing the edge, one of which is a portfolio company called First Hand. that's building agents for brands on the theory that, you know, consumers are going to be interacting with agents, agents are gonna be interacting with agents. So I would say they might be like the one that, you know, folks should keep an eye on to really push this idea forward. Moffie: I think we have a good natural pivot to new category. You created one. You decided we needed another acronym in adtech called UMPS, Unified Media Platform. What is an UMP, Eric? Eric: Yeah, no problem. I didn't create it. I would say my partner Corey should get like a lot of the credit for for this one. We collaborated on it. Six, seven months ago, we started, you know, seeing an uptick in inbound for companies that were providing a layer on top of every API or MCP available supply source for advertisers. And they all kind of had different flavors to what they did. Some, you know, optimized creative and then allowed you to, you know, build creative, deploy creative across Walled Gardens, get the learnings, bring it back, make it better and better, right? somewhat in the way that like, you know, kind of streamer one point z was geared towards. Some were using data in different ways. Some were like super all in on optimization, some were enterprise, some were SMB. Anyway. It was like a couple, then a dozen, and then like over twenty five of them. And we just said, you know, there's something happening here. this could be potentially disruptive. so we went and organized the market into just, you know, a collection of companies that are doing all this. had a bunch of debates on like what the hell to call it. we ended up with the unified media platform because, you know, theoretically it could allow unified access to all of the potential media and supply that a marketer might want. and this category and these companies are exploding. And then there's companies that are pivoting into it. It's like super, super interesting. so that's the the ump category. joseph hirsch: What do you what do you think Moffie: Very inter joseph hirsch: the value proposition for a marketer is of the ump? Like I'm I'm using the trade desk. This is kind of a rhetorical question, but I I feel like it's it's kind of directed at the ump the ump founders themselves, but why like why why can the ump do it better than the platform that I'm using independently? I'm curious if if you think that this is a durable category. If this is a category that just, you know, takes water from an existing beaker and then redistributes it to 25 new beakers, or does it does it create net new water for the ecosystem by introducing a new way to market to a new generation of marketers that have never used one of the legacy platforms of the past? Eric: Yeah, no, those are good questions. The answer to all of them is like p maybe, possibly. Like, we don't know, right? why why would a marketer use this and not just like, you know, use the trade desk? most of the umps that I think are positioned in a very interesting way, they're providing access to everything. They're providing access to wall gardens, they're providing access to search, they're providing access to open web, they're providing access to chat GPT. Right. So like all of these channels that are not necessarily available in like your traditional DSP, they're providing access to. And some of them are providing access to TTD. Right. So it's like this layer that sits on top is thing one. So if as a marketer, you want efficient access to be able to just like get to everything and be able to use your data, be able to optimize creative, be able to have like this like continuous loop. and by the way, all these companies are agentic first. that's the sort of like reason for existence, number one. number two is maybe you're not an enterprise marketer, maybe you're an SMB, maybe you're a franchisee, maybe you're like not a top 50, not top, you know, Fortune 2000 marketer that's geared towards some of these enterprise use cases that we all grew up in. This is a really interesting use case for those types of marketers. It allows them to leg in their doors. So then, you know, with your question of like, does this get new water running through the pipes of this whole like digital advertising advertising ecosystem? Potentially, right? And then there's the idea that this can be somewhat disruptive to existing categories. Like if you have this type of access and it's super efficient and it's super performant, do you need all the intermediaries in the middle? Do you need to be paying all these fees? I don't know. You know, it's kind of interesting. So it feels like something's cooking here. joseph hirsch: Yeah, I think the destruction of fees is a fundamental theme here. You know, I obviously it's really a fee exchange such that you went from fee of X to fee of Y. But then, you know, the the other kind of value exchange would be, well, I could access the platform for free. Now I'm accessing a platform that used to be free with a fee. So, you know, I I assume that the the un The unspoken value exchange is that in exchange for a new small fee for a platform that didn't have a fee in the past, I'm going to get incrementally higher, you know, return on ad spend or decreased cost per acquisition or lower cost per install or higher lifetime value per customer. and I think that's that's what could be potentially interesting is where is the breakthrough UMP that says, and now as a byproduct. not not only have we just consolidated access, not not only have we just taken the workflow and and streamlined it. you know, to your point about dynamic creative optimization, if if you're not closing the loop on the dynamic creative and the conversion attributions such that you're getting new creative with better performance, then what are you doing? Eric: Yeah, yeah. No, you're you're you're right. So that's the bull case for these companies. That's the bull case for this category. And I think that it's like, you know, something worth paying attention to. where's the risk? The risk is these companies are sitting on top of like the biggest companies in the world. They're providing access to Meta and Google and TikTok and Amazon, so on and so forth. And you know, if these companies start to do interesting things that might threaten the either relationship that these platforms have with customers or start to you know find more efficient asset you know places to spend like you know you could see see some scenarios where like with the the Facebook partner program they kind of started like unplugging like there's there's platform there's always platform risk, right? Like with all this stuff. But I think there's enough to believe that this creates like net new dollars for the space from different categories from all sorts of discovery that something interesting can happen here. the the the the best possible pull bull case is like this is the next app love it. This is the next, you know, kind of trade desk from 18 months ago in terms of like a company that is like doing something that is really big, really new and like, you know, the front end for an entire category. joseph hirsch: Alright, so we know you're invested in scope three, at least I think you are. are there are there other Eric: Yeah, you're an investor. joseph hirsch: UMPs on the list that that you guys are invested in? Have you seen one recently where you're like, wow, this this is the breakthrough UMP of the future? Eric: Yeah. so we're in scope three. we're in Newton research. Newton res Newton research, you guys should have John on if you if he's not on the on the list. I mean, you know, what an incredible founder and what an incredible company. Newton Research can do all of this. They're doing everything. They're doing like, you know, data science agents and then they're doing like, you know, this this type of buying enablement. so Newton Research, another one. They're probably one of the ones that, you know, have the most momentum out there. And then there's a couple of startups that we invested in that are doing things interesting and focused on verticals. So one is called AdLe's, super focused on performance marketers. One is called Omnike, very focused on SMBs. so we're trying to find, you know, companies that are focused on these new channels that they can kind of like open up a wedge and start to scale from there. But yeah, the enterprise use cases are covered. I think well with scope three and mutant research. Moffie: Love it. We had Adam Epstein from Gigeon last episode who built a buyer agent on top of the Amazon DSP, and he made a pretty compelling case that they purposefully went very deep and continue to go very deep on the Amazon MCP server and APIs to really build the best buyer agent for Amazon. These UMPs are trying to go broad across multiple platforms. Is anyone having success or is it just anecdotal right now? Can you tell us anything about? Managing a budget across multiple platforms. Eric: Yeah, absolutely. we're an investor in G G. Adam's a beast. we're super thrilled with with how he's doing. But yeah, his his bet is, you know, deeply, deeply focused on what is, you know, obviously one of the most important and fastest growing channels. there is there are UMPs that you have never heard of that are doing single digit and eight digit millions in ARR and growing extremely fast, faster than you know, anything that I've ever seen. most of the startups, they're not kinda like pu publishing numbers, but yeah, there's like real momentum in some of these categories and some of them are just like at the PowerPoint stage. But geez, you know, with all of the agentic tools and the ability to tap into APIs, like they're gonna be in business, you know, very, very soon. so yeah, I think there's real momentum here and there's risk and it's startups and it's asymmetric and we gotta be in it, and we're gonna track the category. Moffie: Speaking of walled gardens and the ability to scale quickly, we have Chat GPT as probably the hottest new vertical. Are you seeing any startups like crop up just to like super serve advertisers in the new Chat GPT channel? Eric: No, no. Mo most of the I mean the startups, you know, they're able to kind of like add it in many cases to to the offering. I haven't seen a GG fork Chat GPT. it seems that most of the companies that are really scaling here are are larger, more established ones. Like, my understanding is Critio you know, is like pretty deep in their partnership with ChatGPT and like seems like a very natural fit. So Seems like the enterprise guys have a head start here, but you know, it's so early. and then ultimately I think, you know, ChatGPT is going to be such an important channel that, you know, it's going to be probably another either UI or MCP into, you know, Chat GPT that, you know, every every marketer is gonna have a direct relationship with. joseph hirsch: Alright, everybody wants to know, you know, what what is what does the exit horizon look like for you and for the portfolio and for the startups in the space like Swivel that are still out there just trying to say to themselves, who is the buyer of these new products? What what do you see on the exit channel? Eric: Yeah. so there's the there is the buyer universe that we all think about and that we all know, right? Like everybody has every startup has their map of, you know, strategics who could buy the company because I am, you know, solving a problem that is going to become existential for them, right? And I think we all know those players. I think the thing that I spend a lot of time both thinking about and having conversations, and I think everyone would be surprised. Surprised at the level of interest is in non-obvious strategics. So I'll point to a deal that happened end of last year and closed the beginning of this year that took everyone by surprise. And that was Pinterest acquiring our portfolio company, TV Scientific, in an incredibly successful exit. How many people, maybe how many of you guys would have imagined number one, Pinterest being the acquirer for TV Scientific, or just number two, Pinterest is acquiring ad tech. Pinterest acquiring CTV ad tech prior to that deal happening. Would either of you have bet on that? Moffie: Absolutely not. I think I audibly gasped and tweeted immediately, like without even like reading it because I was like so shocked by it. So that was Eric: Yeah. Moffie: was that wasn't that around like Christmas, like Christmas special? Eric: It was. It was. That was a fun that was a fun note to send out to to our investors when we got the details on the deal. Joe, did did it surprise you or no? joseph hirsch: I luckily I was an investor. so, you know, I I guess it was a huge surprise insofar as like, you know, checks in the mail are always a huge surprise. as for me, to your point, the surprise as Pinterest as the is the the buyer. I mean, that's the surprise to me. I I hope you're right. Eric: Yeah, exactly. joseph hirsch: I hope I hope people realize that advertising technology has value that it's unforeseen. And I think I think when you use kind of the reverse crystal ball, you can see that data owners are in a good position to be ad tech acquirers. I mean, it it makes you think to yourself, well, if if Pinterest is a buyer, what about the top 20 or top 50 retail media networks out there who could benefit, you know, what if what if you already have a large kind of scaled advertiser base and then you want to bolt on some kind of ancillary product that then gives that scaled, you know, group of advertisers more, whether it's more supply access or more attribution, or better creative optimization. I I'm hoping that, you know, that group of players starts to realize that they could be acquisitive. Eric: Yeah, yeah. Which is yeah, that's that's a punchline, right? Like I think that enough companies, whether they are RMNs or VMs, right? Like vertical media networks, think of a PayPal, so and so forth, not naming them specifically, or those who have, you know, interesting data or relationships with a lot of companies, they realize they can get into the advertising game. They see the margin profile of the advertising business that's being built in front of them. With some of their competitors or complementary categories. And then there's a whole set of companies that can bring them to your point either net new capabilities or net new customers. I spend a lot of time with those types of companies. And you'd be surprised at the amount of lean-in, but then also not surprised at, you know, just some of the fundamental, gotta wrap my head around this, you know, advertising business. Where do I start? How do I think about this category? How do I put together like you know my strategy? that happens in those conversations. But I think we will over the course of the next twelve months see more head scratching at the time acquisitions, but then, you know, months later, especially in the case of something like Pinterest where T V Scientific is already starting to drive real results for them, like, you real ROI being created. That's my bet. Moffie: Well, we've seen we've seen that thesis accelerate. We saw Walmart acquire Eric: Yes. Yes? Moffie: vibe at Cannes, another shocker to the industry. So we see this retail media network plus performance CTV play continue to be very hot. Is that the category or those two categories combining going to continue to be the hot categories going into Q4 and into 2027? Is there other categories emerging? What are your thoughts? Eric: Yeah, I think for sure, right? Like what what are the growth areas that as an acquirer you can, you know, lean into. It's super obvious. It is C T V generally performance C T V, you know, kinda you know, new new new types of customers, specifically is is thing one. Thing two is commerce and retail media s you know, absolutely. And then there's just like the capabilities that companies like both of yours are building. That I think are gonna be part of, you know, what'll be a big wave of MA over the course of the coming months, maybe not to not not Q four. but you know, this whole agentic wave is changing everything. And if you're not building it, you're going to be flat footed. And I think we see a cycle of MA that just happens from companies just like buying great buying great startups that have these capabilities, have these technologies, that, you know, before they get too expensive. Or maybe they'll get too expensive and they need to do it anyway. So like per like like like the sort of upgrade of ad tech infrastructure in this AI era is the other thing that I think every strategic that I talk to is like thinking deeply about and trying to map out the ecosystem of of potential acquisitions. joseph hirsch: All right, we need a hot take to grab listeners' attention on LinkedIn, some byline that just it it can be it can be uncontextualized. but as long as I think you say it in the pod, then we can use it to grab people's attention. Is there something that you really want to share with the world in this moment that's kind of salacious but exciting, tangentially related to ad tech, also involving agents and relevant to unharness the podcast? Moffie: I'll I'll I'll rant in a little bit. So we know AI slash agentic will be the most investable sector that everyone talks about in 2027. You asked earlier why we launched Unharnessed. Although all the podcasts, all they do is talk about agentic, there actually wasn't a podcast that just focused on agentic until now and also has real operators. Sorry, Eric, Ari, Joe. guys you guys are on the sidelines these are these are two operators we're in the freaking cockpit driving agentic so 2027 Eric: Hell yeah. I'm I'm I'm I'm here to support you in every way. Moffie: what most investable ai categories in twenty twenty seven in agentic Eric: Most investable categories in 2027. Think of it as three layers, right? So it's at the foundational layer replacing existing programmatic infrastructure. that would be number one. Number two, how do I frame this? Like agentic on top of existing platforms, and maybe GG. Is an interesting example of that. because that is not only, you know, kind of a play on agentic, but it's a play on how agentic is changing how customers buy and how customers work. I'm sure, you know, I didn't hear the pod, but I'm sure Adam talks to you about how he just spends a lot of time helping them think about how they can become an age an agent forward organization. and then I think, gosh, there's like new category creation to come, right? And whether that's UMP is with a fresh take on buying, whether that's marketing to agents, whether that's new formats, like that top layer is where I think you know, the most asymmetric opportunities are, the most exciting, you know, companies are going to be built, and where I'm spending a lot of time just trying to think about and understand, let alone evaluate opportunities. Moffie: Love it, love it. Joe, anything else you wanna ask? joseph hirsch: No, I think I feel pretty good. I think Eric is a fountainhead of wisdom and I'm I'm thrilled that he was able to join us today on the pod. Eric: You're you're too kind, Joe.