Episode 141: The State of Parma Advertising with Josh Walsh of BranchLab

Summary

Marketecture: Get Smart. Fast.

Episode 141: The State of Pharma Advertising with Josh Walsh of BranchLab
Date: September 26, 2025
Host: Ari Paparo
Co-host: Eric Franchi
Guest: Josh Walsh, CEO and Founder, BranchLab

---

Episode Overview

This episode explores the pivotal transformations underway in pharma advertising. Ari Paparo and Eric Franchi dive into the week's major ad tech news—Google’s antitrust trial, advances in AI, major industry M&A, and Nvidia's game-changing AI investments—before a deep-dive interview with Josh Walsh, CEO of BranchLab. The discussion with Josh unpacks the technical and regulatory challenges in pharma marketing, the evolution of privacy-safe targeting, and the implications of recent legislative moves.

---

Key Segment Breakdown & Timestamps

1. Industry News Discussion

[00:00–31:37]

a. Google Antitrust Trial—Live from Arlington

• Vibe of the Trial: Less about guilt/innocence and more a “dysfunctional negotiation” on remedies after Google already judged a monopoly.
  • Ari: “This is not a trial about guilt or innocence... It’s like a really dysfunctional negotiation.” [04:00]
  • Judge Brinkomash’s running joke: "My two favorite words in English language: 'Are we settled?'" [05:54]
• Divestiture Scenarios: Conflicting witness testimonies about what should be spun out (e.g., ADX, DFP); buyers lining up.
  • Ari: “I could run the ad exchange. There’s probably like, there’s a ridiculous number of companies who would want to bid on this thing.” [08:02]
• DV360 Revenue Insights:
  • New evidence shared showing DV360 spends $450M/month on display, ~$800M/month on YouTube, making DV360 primarily a YouTube buying platform. [09:00]
  • Comparison: Trade Desk estimated at $7–8B annual spend; DV360 (including YouTube) at ~$15B annually. [11:04]

b. Google & Browser AI Innovations

• AI in Chrome:
  • New Gemini integration and agentic features—browser acts as an assistant, booking services on your behalf. [12:49]
  • Ari: “The idea of an AI powered browser is really exciting for me... I think it could actually be good for publishers.” [12:49]
• Long-term Impact:
  • How AI agentic browsing could change web and publisher interaction models. [13:50]
  • Eric: “That’s a really neat perspective... will this be Chrome and Google owning it... or something ubiquitous?” [14:53]

c. Ben Thompson: Google’s AI Future in Video

• Stratechery’s Take:
  • Google’s AI edge may be in YouTube, not chatbots—using AI for shoppable moments and creative tools. [16:06]
  • Ari: “I don’t think we’ve seen the end of innovation in the format of video people experience.” [17:54]

d. IAS Goes Private

• IAS Acquisition by Novacap: $1.9B buyout (22% premium), following a trend of taking “legacy” measurement firms private for AI-centric retooling. [19:33]
  • Ari: “Was there something that needed to be done to the company, structurally, that was so big you couldn’t do it as a public company?” [20:55]
  • Eric: “Take the thing private. Really retool this for the AI era, likely do some acquisitions.” [22:00]
• Strategic Outlook:
  • Ari’s playbook: “Sell Publica for $500M... Or double down on measurement and buy Kantar, Nielsen, Comscore.” [23:44]

e. New TTD & Axiom Partnership

• True Intelligence—live UID2-based feedback loop, aiming to bring walled garden-level measurement to the open web. [26:23]

f. Nvidia–OpenAI’s Monumental Deal

• Nvidia Invests $100B in OpenAI:
  • Revenue commitments and equity upside—a power play in compute supply for the AI era. [27:41]
  • Ari: “It’s like Web 1.0 all over again… when the tide goes down, when the music stops... it’s not the best.” [29:02]

---

2. Featured Interview: Josh Walsh (BranchLab)

[31:37–51:56]

a. BranchLab’s Mission & Technology [32:03]

• Privacy-Safe Health Data Targeting:
  • BranchLab uses neural networks trained on health data to predict outcomes using non-health data—avoiding PII and meeting HIPAA/state regs.
  • Josh: “We’re training different types of neural networks… to predict health outcomes in non health data. It’s a privacy-forward way of activating against healthcare audiences.” [32:05]
  • Leverages demographic, psychographic, firmographic data to mimic traditional health segment performance—without exposing raw health info. [32:36]
  • Key regulatory challenges: new state laws (WA’s “My Health My Data,” NY’s S929), which heavily restrict how health audience data can be deployed. [33:11]

b. Strategic Funding from Lance Armstrong & Next Ventures [34:09]

• Origin Story: Investment came via a direct outreach from Lance Armstrong (“I had to check it wasn’t a bot”).
  • Lance’s Next Ventures sees BranchLab as part of “transforming the US Healthcare system from sick care to proactive care.”
  • Process: “We met Monday, deal was closed Friday.” [34:09]
  • Ari jokes: “Would you say it was a big step up from Aperium giving you money?” Josh: “It’s a lateral step, you know… one plus one equals three.” [36:47]
  • Lighthearted moment about joining Lance on a bike ride and potential for having Armstrong as a Marketecture keynote. [37:10–38:34]

c. Deep Intent’s $637M PE Buyout—Pharma DSP Valuation Context [39:26]

• Congrats from a “non-competitor”:
  • Josh: “We partner with Deep Intent... a lot of our customers deploy BranchLab-powered audiences on their DSP.” [39:26]
  • Industry insight: “This is a huge, specially regulated, hyper specific market that requires specialized tools, and Deep Intent at that valuation is more proof of that.” [39:55]

d. Why Pharma Advertising Needs Niche Solutions [41:07]

• Not Just Data, but Activation Rules:
  • BranchLab delivers models, not audience data, so clients can use their own CRMs without violating new privacy laws.
  • Complex state-level regulation means standard audience matching (even pseudonymized) is increasingly risky or non-compliant.
  • Josh: “The legacy way of sending health audiences is prohibited in [WA state]... probably prohibited in other states.” [42:40]
  • “We lock hyper-regulated data behind a neural net, never let it in or out, and instead just deploy lightweight probability models.” [44:03]

e. The Federal Government & the Specter of a Pharma Ad Ban [45:14]

• Bipartisan Issue, but Not Likely to Result in a Ban:
  • More likely to see increased regulation or requirements for side-effects disclosures than an outright ban.
  • Josh: “This administration is taking on what really is a bipartisan winning issue, which is, how can we better fight for the patient/consumer.” [45:14]
  • Ari: “It sounds like they’re trying to get a political win here, regardless of whether it really moves the ball forward that much.” [46:17]
  • New proposals might tweak creative/messaging and close loopholes, but legal and First Amendment precedent make true bans unlikely. [47:45]

f. Prediction: Heightened Regulation, Technology-Led Privacy

• Federal preemption possible—a move to a single, national standard from growing patchwork of state rules.
• Importance of privacy-by-design: “When done responsibly, advertising can be one of the most powerful tools for improving public health... it equals more doctor visits. Health outcome improves mathematically.” [49:52]
• “If it requires regulations to ensure… the industry does that, I applaud it. I hope that doesn’t need to happen. In my opinion, privacy by design technology can solve this problem very quickly versus having to wait for things to be settled in the courts.” [51:18]

Memorable Closing

Ari: “Stop doomscrolling on WebMD and asking ChatGPT why your hip hurts. Just ask your doctor.” [51:38]

---

Notable Quotes & Moments

• On the dysfunction of the Google trial:
  Ari: “It’s maddening for me to sit in the pews listening to witnesses who have the most vague understanding of the issues at hand and are pontificating about what they think.” [05:03]

• On the rise of AI browsers:
  Ari: “What if you could have a conversation with your browser… summarize the opinions on this website about X… So many different ways it could enhance content-rich websites.” [12:49]

• On the future of video and AI:
  Ari: “I don’t think we’ve seen the end of innovation in the format of video people experience... AI can enable things that seem futuristic right now—like a personal shopper video that knows you.” [17:54]

• On the strategic rationale for the IAS buyout:
  Eric: “Take the thing private, really retool this for the AI era, likely do some acquisitions... [IAS is] an anchor for more acquisitions and it comes out as a different type of company.” [22:00]

• On the privacy-by-design approach in pharma data:
  Josh: “BranchLab’s whole mission is to take hyper-regulated data, lock it behind a neural net, never let it in or out... just deploy lightweight probability models.” [44:03]

• On the real public health impact of pharma ads:
  Josh: “Ethically, responsibly produced healthcare ads do not necessarily equal more drug sales—they equal more doctor visits... The more time a patient spends with their physician, health outcome improves mathematically.” [49:52]

• Lighthearted moment—Investor trips with Lance Armstrong:
  Ari: “Wait, are you serious, Eric? You going on a ride with Lance Armstrong?”
  Eric: “I mean, how could I say no, right?” [37:10]

---

Takeaways for Listeners

• Pharma advertising sits at the intersection of high regulation, privacy, and innovation: The market’s “specialist” status is justified both by compliance complexities and its scale.
• The future is privacy-by-design: As U.S. state laws tighten and the federal government considers national standards, neural nets and probabilistic models will be central to compliant pharma ads.
• Major AI and M&A moves are reshuffling the ad tech stack: The week’s news—Google’s steady AI rollouts, major acquisitions like IAS, and Nvidia’s OpenAI deal—signal a decisive shift to AI-driven infrastructure and measurement across verticals.
• Marketing and measurement remain key battlegrounds: The TTD-Axiom partnership and questions over DV360 vs. Trade Desk market share show innovation and integration are crucial for winning marketers’ budgets.
• Political winds may drive regulation, but outright pharma ad bans are unlikely: Expect more disclosure requirements and constraints on data usage, not prohibitions on ads themselves.

---

For a deeper dive into the nuances of pharma ad tech and the latest industry maneuvers, the interview with Josh Walsh provides an expert’s perspective rooted in technical detail and practical experience.

---

[End of Summary]

Transcript

[A]

This podcast is brought to you by Adelaide. Media verification and measurement are undergoing major disruption. Legacy players are pivoting to performance. Advertising AI is reshaping brand safety and attention is replacing viewability. Adelaide is leading the shift with au, a new way to assess media quality that scores placements based on their potential to drive attention and outcomes. Before your ads run, think of it like a credit score for media. Finally, a clear view of quality. Before you buy, take the guesswork out of your investment strategy and try Adelaide AU on your next campaign. This episode of Architecture is brought to you by Start IO, an industry leading mobile SSP. If you build ad campaigns, Start IO is offering architecture listeners curated evergreen deals with no audience fees. Visit www.start.IO nofees to activate today. That's www.start.IO no fees. Foreign welcome to the Market Extra Podcast. I'm Ari Paparo. I am down here in Arlington, Virginia covering the Google trial all week. So I'm from a hotel room and got a little mic instead of my normal mic. I feel like I'm kind of roughing it. Eric Franchi, how are you doing? You're in beautiful New Jersey, right?

[B]

I'm doing well. I'm in beautiful New Jersey. How is beautiful Virginia?

[A]

It is like 80 degrees and 500% humidity. It is like the most humid place I've ever been in my life. And I've been to New Orleans.

[B]

How's the AC in the courtroom?

[A]

It is so blasting that everyone is. There are people wearing coats inside and the bailiff keeps coming over and apologizing for it being too cold, but he hasn't been able to do anything about it. And we're shivering, literally, while we're taking our notes longhand and trying to stay awake. I saw Catherine, from the information, if you're listening, I saw you were asleep today. You were next to me. You were out. I was gonna elbow you, but you don't snore, so I figured you'd get away with it.

[B]

I was gonna say the chilliness is probably good for keeping everybody awake and attentive, but no, not Katherine maybe. All right, let's get into all that stuff in the. In the refresh for sure.

[A]

Yeah. So we're gonna do this a little out of order for normal listeners. We're gonna do intros and the news of the week, and then we'll switch to our interview, which we recorded previously with Josh Walsh, the CEO of Branch Lab. So we'll talk about that when we intro him. And from a housekeeping perspective, I want to remind everyone that you can read my daily updates on the trial at The Monopoly Report. Monopoly-Report.com Daily Newsletter. We try to keep it entertaining. Lots of news, a lot. Also a lot of scoops that aren't really related to the trial, but that you can that I've been able to glean from the evidence. Like today we found out how much DV360 spends on YouTube versus everything else. And the answer is a lot. And then secondly, Eric and I talked that we're going to have our startup showcase at the Market live event on October 27th. I think we are announcing the winners on Monday or Tuesday and then we can discuss it on the POD next week. Eric, are you excited about that?

[B]

Yeah, I mean, we've narrowed down our field of five. We're just getting confirmations that everybody can be there in person. The five are really exciting. So I'm excited for this one.

[A]

Yeah, me too. All right, with that, let's just jump right in because it's only Wednesday when we're recording this and it's already been a pretty busy week. So what's going on?

[B]

Yeah, thank you for the M and A gods for blessing us with a deal on a Wednesday. Otherwise this would be Google and AI as usual. But let's talk about Google and AI. So this is now day three. You're probably after this, going to be writing up your newsletter for tomorrow.

[C]

Can you give us a little preview.

[B]

On kind of stuff that might not make it in or just in general if I be things that you could talk about that don't make the newsletter?

[A]

Yeah. So next week we're going to have Alan Chappelle back on. He was just on a couple weeks ago, but Alan and I will go into a lot of depth next week on the trial. We'll also have the trial will mostly be over by next week. So what I could offer you right now is just the vibe, which is this is not a trial about guilt or innocence the way the kind of the first one was. The first one was fairly dramatic. Had a lot of testimony about, you know, what happened and what was right and wrong. This is like a really dysfunctional negotiation. You know, we've all been in complex deals where the two sides have different points of view about how a deal should get done. And usually you hammer it out and maybe yell a little bit or whatever, you get the job done. But that's not how courts work in this case. The two sides have their own views for how this should be resolved. The company was already adjudicated as a Guilty monopoly. So they're not even debating that anymore. They're just saying, like, here's how we think we should resolve the problem. The two sides are pretty far apart and instead of working to resolve the differences they argue, they basically spend all time impeaching the other person's point of view. So you have witness after witness who's like, no, the other guys are wrong. They're stupid. None of that makes sense. And often the witnesses are, well, let's just say this as a. I think everyone who's listening is going to acknowledge I'm probably an expert in this area, having written the book. It's maddening for me to sit in the pews listening to witnesses who have the most vague understanding of the issues at hand and are pontificating about what they think. Like, hey, I really think it should be spun out. Oh, good for you, Congr. Congratulations. You think it should be spun out. Thank you for your input. And it's just like not. I mean, the judge makes this joke pretty often. She's awesome. Judge Brinkomash is 82 and she just makes this joke a lot, which is like my two favorite words in English language. Are we settled? It would be so much better if they settled than trying to defeat the other one rhetorically or throughout, you know, half baked testimony so that they get their way. And. And especially when the end result could be a divestiture, which will be. A divestiture will be adjudicated in the court of investment bankers. You know, someone will have to buy it and someone have to negotiate that and that'll be real because there'll be real money. So what the court says in advance of that is, you know, kind of academic. So that's my point of view on the vibe. It's a bit frustrating.

[B]

Yeah. In the newsletter Today, the Day 2 newsletter, which is really good, by the way. I love the little cartoon of whack a mole.

[A]

Whack a mole.

[B]

Yeah. If you haven't seen it, just sign up for that. Basically, it's like there's two camps. There's obviously Google's camp, which is don't spit it out. Be the worst thing ever. And then there's all these variations of spinning out. So spin out ADX separately, consolidate or shut down adex. What else? Spin addicts and DFP together. There's all these, like, different pieces. Yeah. Permutations. And a bunch of our friends are kind of at odds on this one, which I found, you know, from the cheap seats of the newsletter. Pretty funny.

[A]

Yeah. You know, the DOJ has mostly been, has only been the one calling witnesses so far. And some observers in the courtroom who are a little more familiar with the court process were sort of scratching their heads about why the DOJ witnesses were disagreeing with each other. Typically you would only bring witnesses when you know what they're going to say. And so you had like James Avery from Kevl who's, you know, I know an investment of yours, a good friend of mine and he went up there and said no, they should be spun together, which is not the DOJ's point of view. And he was precious about it and I'll buy it. Yeah, we have two, we have two parties who want to buy it who've testified so far. You know, is Index Consali, James Avery at, at Kel. And I'm sitting in the pews like give me a billion dollars and I'm in like I'm putting my chips in there. So I'm not CEO of Market Extra anymore. I could run the ad exchange. There's probably like, there's a ridiculous number of companies who, who would want to bid on this thing.

[B]

Yeah. Good marketing for, for Index and, and Kettle, by the way.

[A]

Well, yeah, sure. I mean if you, it's, it's fun to like punch a little bit above your weight. I mean Index wouldn't be that much. But Kevl is like, you know, relatively small company to be doing a multibillion dollar acquisition.

[B]

Yeah, yeah. There was, there was one quote at the end of your piece where you said that Jed from TTD was named and he, you know, there's some mention in cross examination, examination they thought SSP partners had undisclosed markups including you know, at least one of the folks in the, in the room. Was there any moment like that?

[A]

Today this episode of Architecture is brought to you by Start IO, an industry leading mobile SSP. If you build ad campaigns, Start IO is offering architecture listeners curated evergreen deals with no audience fees. Visit www.start.ionofees to activate today. That's www.start.ionofeees. well today and I put it in Twitter and I mentioned it a moment ago, they put up some slides with data I've never seen public about the percentage of all of Google's spend that goes to by format. So YouTube versus display versus app and also the same thing for DV360. They showed a chart that illustrated the spend per format per month from DV360 over the years. And that second one I made a little Twitter thread we could link in the show notes and I'll probably put in the newsletter today. And I had to write it down. So it's not entirely accurate. But the news was that display is about $450 million a month in spend. So call it 5 or 6 billion in spend per year from DV360 and YouTube at the beginning of this period in 2019 was about $100 million a month going from DV360 to YouTube and now it's 800amonth. So YouTube at this stage is $10 billion just from DV360 and is well more than half of the total spend from DV360. Like. Well, more like DV360 at this point is a YouTube buying platform with a little bit of display thrown in. Right. And a little video thrown in, which, I mean, the, the extent of that was a little shocking to me. And we know. And also it's great that we finally have a stat that all DV360, including YouTube is probably 15 billion a year, which he still puts them as the number one player. I don't think the trade desk has disclosed their total media spend, but I think it's somewhere in the 7, 8 billion a year range.

[B]

That sounds.

[A]

Listeners can correct me if I'm wrong. And, and no one knows what Amazon's number is, but it kind of gives you a little bit more of a sense of the market share and how it's moving around.

[B]

Yeah, okay, all right, cool. Well, that's a good one.

[A]

That is a good one.

[B]

And yeah, more, more to come on the Google front since they, you know, loom so large. There was two other bits of news related to AI that we should, we should touch on and we'll put all this in the newsletter. The first was. And I'll shout out Shiv Gupta from New of Digital. He was screaming about this on Friday. Again, just news on Friday dropped. We, we, we, we weren't able to cover it, but I thought it was actually really important. So they dropped a blog post on Chrome called Go behind the Browser with Chrome's new AI features. So I haven't refreshed Chrome yet, but it seems like this thing is out there and it's live and there's two things to touch on. So number one, they're rolling out Gemini and Chrome. So I think you're starting to see them bringing these tools to everyone, including just kind of like searching for the. Searching in the omnibus browser. And then the other thing, I thought this was the one. It's titled Number two Get ready for your agentic browsing assistant. So they are basically releasing an agent in Chrome. According to them, these will let Gemini and Chrome handle those tedious tasks that take up so much of your time, like booking a haircut or ordering weekly groceries. You tell Gemini and Chrome what you want to get done and it acts on webpages on your behalf while you're focused on other things. So I think this is like really big and shows the potential power of Google in terms of what they can do with AI. And it's just the beginning. Did you, did you check this out?

[A]

Yeah, I did. You know, the idea of a AI powered browser is really exciting for me. And so far there's the. Was it the D browser was the first. And then Perplexity has their sort of comet experimental thing. Right? Nothing really put this, put this into Chrome is the big move. I'm really optimistic about this. I think it's maybe as relevant a user experience as a chatbot. And I think it actually could be good for publishers. Like, not bad for publishers. This is the first time because you're going from zero. Okay, so you have zero search or zero click search, which is the chatbot where you get the answer instead of clicking out. But when you're engaged with a publisher, let's say you're on the market news website and normally what you do is like, what is that article that Ari wrote about this like six months ago? You search and you peck around and you do stuff. What if you could have a conversation with your browser about it? It's like, can you summarize the opinions on this website about blah, should I subscribe to this website? Are they known for sending spam? You know, there's so many other, so many different ways it could actually enhance content rich websites. I use the example. Maybe that's a bad example. Let's think about Yelp. Like what is. Maybe I just said this last week. I hate if I repeat myself, but like what's going to be better? Like asking an OpenAI chat box about what's the best restaurant in my area. Okay, that's option one. Option two is going to Yelp and searching. Okay, that's a good option. Option three is Yelp implementing an LLM. That's a good one, but not everyone's going to be that great at it. And option four, which I think is the answer, is an LLM that knows who you are in your browser that is on the Yelp website. Right? So Yelp content plus personalized browser experience. And maybe you have to Pay Yelp for access to that. Maybe Yelp can still monetize in that circumstance. And I think it's really, just really interesting and I think it could be a great thing for the web.

[B]

Yeah, that's a really neat perspective. I guess the question that's raised is will this be Chrome and Google owning it because it's the most popular browser, or will this be something that's ubiquitous and there's like independent chatbots, or LLMs, you know, on all these publisher sites that are either like a third party company or like the publishers or utilities themselves? That's gonna be an interesting question.

[A]

Or both.

[B]

And it's gonna be like bots everywhere.

[A]

Yeah. Also, information versus workflow is another good question. I feel like there was an announcement of a thing called operator from OpenAI where it was a headless browser that did things on your behalf and instantly everyone realized that was a terrible idea because they would have to log into every website for you and stuff like that. But having it in your actual browser suddenly becomes a good idea where you can, you can automate work related things or you know, semi work, like, you know, personal work type things. I think the examples they always give these tech companies about shopping for groceries or booking a vacation are kind of bogus. You know, let's, let's do some real work in the browser like all the listeners of this podcast generally do.

[B]

Yeah, yeah. I would say for one thing for the listener, just go to this post, it's on the Chrome blog, and just look at the animation of the Segentic browsing assistant and kind of let your B2C or B2B minds run wild on what this could be. So more to come on this? I think so, but it's pretty exciting. The other thing I thought was neat was so Ben Thompson, Stratecheri, awesome, awesome writer and thinker. He had a post this week that was a little mind bending because they took the other side of the argument basically saying Google's going to win an AI, but it's not going to be with this chatbot stuff at all. It's actually going to be through video and specifically on YouTube. And it's as these things are on Stratecherry, a super long post. We'll link to it. I'm not going to summarize it, but he points to how if the future of so much of this web stuff is really going to be video, YouTube has already won versus TikTok, but they seem to be slowly rolling out some interesting things that point to the potential power of this and the two that he pointed to again, they're all in the post is first, they're offering VO3 tools to creators to start to augment their own content creation using VO3 and like mixing videos or adding AI. And then the other side, they dropped a little hint in another post saying that they're going to be using AI to find moments in video to promote commerce. So like, literally think the entire entire. The entirety of YouTube becomes shoppable, which I think could be like, pretty powerful from a monetization perspective. So I'm not doing the post justice. He's amazing. Read the posts. It'll give you a completely different way of thinking about how Google or really other companies can win in AI and centered around video.

[A]

Yeah, My only point I'll make on that is that I don't think we've seen the end of innovation in the format of video that people experience. So, like the same way, say, you know, five or six years ago, the short form vertical video that you scroll was not a thing, or maybe it was a very niche thing, and now it's the dominant platform. I don't think we're done. And there's so many opportunities for format changes as opposed to saying, like, here's a video we already have. Let's make it shoppable. Fine. Okay, that's one idea. But there are a lot of different ideas. And AI can enable things that are, you know, would seem quite futuristic right now, like a, like a personal shopping video. Not a. Not a personal shopper and not a video, but a personal shopper video that knows you. Knows your size, knows your style. And like, we'll talk to you for half an hour about, you know, how you want to dress professionally or something like that, or how you want to, you know, dress your bridesmaids for a wedding. So there are some really interesting things that could happen if you take a kind of another leap in the technology cycle.

[B]

Yeah, yeah, well said. And we're already seeing it with ads, right? Like, you're seeing the, you know, beginnings of real personalization around ads using AI. And obviously all this stuff is super early and it'll start with ads.

[A]

Yeah, totally.

[B]

All right, that's enough Google. Let's talk ias. So, Lisa, we had you on like last week. What's going on?

[A]

She was holding out on us. She had the money in the bag underneath her desk. There was a couple billion dollars. And she was just like, hey, no big deal. We're just measuring stuff. AI is co.

[B]

Yeah, she did seem to be, you know, Fairly upbeat for. For tough interview from. From Mr. Paparo. So for those who didn't see the news again, today's Wednesday. You're gonna hear us on Friday. IIS was acquired for 1.9 billion by the PE firm NovaCap. So a couple things here and we'll talk about what this means. So first is it was a premium to Yesterday's closing price. 22% premium. This is. It implies kind of like a 4X, maybe a little bit less. Their trailing 12 months revenue was 570 million. So a 1.9 billion buyout. Right.

[A]

Like, you know, was what's it on earnings? Because I mean, obviously it's a mature company. We should do, you know, earnings multiple.

[B]

Yeah, I did it off the top line. I did it off the top line. We'll check that out, discuss on, on the X. But you know, on a top line, which is, you know, how oftentimes you look at this stuff, it's, it's certainly a discount and the acquirer. Excuse me, is interesting. So Nova Cap, they made the news by acquiring Caden, which was a pretty big deal. And then ad theoret. So they seem to be, you know, an active buyer with a thesis on ad tech and actively doing deals. So what do we think? What do we think about this? And then what happens from here?

[A]

Yeah, I'm scratching my head a little bit here because it was not an unsuccessful company, but not a leader, I guess is where I'd put it. They were certainly losing some share, maybe losing some mojo to their arch rival, Double Verify dvd. So why couldn't they fix that in public? Like it was there something that needed to be done to the company, like structurally that was so big you couldn't do it as a public company. And then they'll take it off line, do those things, either slim down or bulk up and then go public again. Because if you're just bringing it private to make marginal improvements and then try to go public again, that doesn't make a lot of sense to me. So I just, I just wonder, is this a failure of the public markets or is there some missing part of the strategy we're not seeing that needs to be done in private?

[C]

Yeah, it's a moment in time.

[B]

Right. Like we've talked about this category so much and I think the thing that we've said constantly is, you know, we want to see more innovation out of these. Verification, legacy verification, you can call them something else companies today, and you've got startups that are using AI and like really nipping at their heels. Maybe not from a customer adoption perspective, but showing, you know, what can be done using some of these tools. So I look at this as take the thing private. Really, like, retool this for the AI era, likely do some acquisitions. There's a lot of good assets for iis. They're, you know, a known trusted brand. They've got a really good customer list. They've got a really good partner list. So I think it's, you know, do some real work to kind of like build this in private for the AI era and then come out and you'll look to. Look to. To go public again. Nova Cap, they don't strike me as a rip all the costs out, you know, do some Right. You know, dump some debt on it and just keep like, kind of ripping this down. You know, kind of the bad PE style. They strike me as, look, they bought cadent and then they bought at their right. So, you know, is IAS an anchor for more acquisitions and it comes out as a different type of company. That's. That's my bet here.

[A]

Yeah, I don't think slimmed down makes sense. They're not big enough. They're not fat. They, you know, there's can't. You can't cut this thing to growth. You got to grow it. So here's what I would do. Just. I know you asked what my opinion is. Step one, you sell Publica for like $500 million to get that back off your 1.9. So public video ad server. That acquisition never made any sense. There's a hot market for video ad server. You could probably sell it to Netflix tomorrow for half a billion dollars. Yeah, well, they bought it for 250, which was probably too much. But video. There are no video ad servers. If you wanted to acquire one, who would you buy? There's like, no one to buy. So maybe Netflix wouldn't buy, maybe Walmart would, maybe Samsung would. There's a lot of options out there. And then on bulking up, you know, you kind of have a choice. You could either double down on measurement and buy. Look at Kantar. You can even look at Nielsen Comscore. There's a lot of options. Or you can kind of go after the double verify approach and try to get into more of the bigger picture. Optimization, et cetera. I don't know who you'd buy in that space, but that would be the other option. But there's a lot of stuff you could do to try to juice the growth rates and make it a more exciting company.

[B]

Yeah, I think option two is pretty exciting. And to answer the question, there's a lot of these companies that are building the custom algorithms and kind of the modern measurement solutions that I think would be super interesting to be part of is again, who has like a big customer base and a big partner base. So we shall see.

[A]

Yep, it'll be interesting. Watch the space.

[B]

Yeah, for sure. Do you think we see more of this, like companies going private to retool for the AI era, or you think this was a one of one?

[A]

I don't think you need to go private to retool for the AI era. It's not unless you're. You go private because you want to potentially have a period of some time where you're significantly less attractive as a public company. Right? Yeah.

[B]

Or just less scrutiny and like less, you know, reporting like quarterly. That's so hard. Ask anybody.

[A]

So when you say retool for the AI era, what that means to me is that I would accept like much lower margins or a reduction in revenue while I retooled. That would be the reason why I don't think that's necessary for most companies. Like you can. You can add AI and manage through. If you were a successful public company, you could get through that transition without going private. Unless there was something so fundamental to what you did that I was going to blast your. Your revenue over that period. I don't think is. Is in that position where they're going to get, like, transformed out of their business in the next, you know, 12 months. So I have to, I have to believe that they're going private because they want to do some big, you know, bulking up rather than like a retool.

[B]

Got it, Got it. Okay. All right, we'll. We'll revisit this partnership. TTD and Axiom announced a pretty interesting partnership or deepening of their partnership around measurement. So the product's called True Intelligence. It aims to create a live feedback loop from UIDs targeted by TTD with results collected by Axiom, thus giving marketers the type of measurement that they get from the walled gardens but craving in the open web. This is not a first for Axiom. They've similar product with Amazon, but I think this is a pretty good advancement by TTD in having something that's interesting and fresh for the quote unquote, outcomes era.

[A]

Yeah, I mean, I don't know how many measurement partners have adopted UID 2. And obviously the more they do, the more valuable it becomes their customers. So. Seems pretty important from that perspective.

[B]

Yeah. Yeah, it does. Like I said, I think we're going to see innovation out of ttd. Last week I had no idea this was coming, but this is the type of stuff that I'd like to see them do because this is like what marketers care about matters.

[A]

Yeah, absolutely.

[B]

What else? Nvidia. So this deal, this is not an ad tech deal, but because this is AI and it touches ad tech in so many ways, it's worth talking about and it's just like wild. So Nvidia is investing 100 million in OpenAI.

[A]

Billion.

[B]

100 billion in OpenAI. Turn the B to the M and in exchange they are getting like that amount of revenue commitment from OpenAI and then plus they get the equity upside. So this is a genius deal by Nvidia.

[A]

Yeah. Wasn't there a deal like this with Oracle just like last week?

[B]

Yep. Oracle got a long term commitment. I don't believe that they announced who the customer was. But similarly, this was what set the Oracle stock up 30% overnight.

[A]

Round tripping, man. It's like Web 1.0 all over again. It's like you give AOL all your VC money and then they advertise and then your valuation goes up.

[B]

It is, I think I picked this up on one of the write ups. But the hotline compute capacity is the new mode. They just need so much capacity to build out what they're looking to do.

[A]

Sure, that makes sense. I guess in some world. I mean.

[B]

Good.

[A]

Benedict Evans like made some snide comment on threads about it which was like, you know, these round trip deals sound great but it's one of those things that when the tide goes down, when the music stops. Yeah. It's not the best.

[B]

Yeah, yeah, for sure. I think there's reason to believe that things are different this time. But we saw it. I mean Meta, remember Meta, had that. Chuck had that meeting at the White House and he asked how much you're going to be investing in data centers and he kind of dropped the 600 billion line. These are like real numbers. So it's mind boggling.

[A]

It is. I mean there's nothing stopping the big tech companies at this point. It's just like pure Runway in front of them and like the rest of the stock market, sort of irrelevant and the Dow keeps going up and you know, doesn't seem to matter even if the economy is good or bad or shaky. There's just so much opportunity ahead.

[B]

Yeah. For these seven companies for sure. All right, I think that's a good news of the week.

[A]

All right, awesome. So as, as we said at the beginning, we're going to do the interview last. So we have Josh Walsh from Branch Lab. So you're an investor in Branch Lab, right?

[B]

Yes, I am. And I believe you're an advisor to Branch Labs.

[A]

I'm an advisor. So they, you know, ostensibly they're in pharma marketing. And you know, we've talked about, and we asked Josh about the recent Deep Intent transaction and about RFK Jr. And that's what we really wanted to talk to him about because it's a really interesting area. But also it's interesting that what they're doing really is like applying neural networks to data, basically saying like we have Data set one and we need to infer its attributes on data set 2 without using any PII or any anything like that because of HIPAA and stuff like that. So they've built some really interesting tech to do that. And Josh was, I think he was one of the founders of Adherent or was he a product manager at Adherent or something along those lines.

[B]

He was co founder at Adherent, so had a very healthy health. Health business.

[A]

Exactly. So I think it's really interesting interview for the pharma aspects, but also the tech aspects. I hope you enjoy that. And then we will be, will be back next week with sort of a remedies phase wrap up with Alan Chappelle, which should be interesting because the trial will be nearly over by the time we record that next week. So hopefully you'll tune in. All right.

[B]

And it's not just about pharma with Josh by the way. He's got big name investor that just joined the company and we get into that as well. Pretty fun story.

[A]

Yeah, exactly. All right. All right, see you next week. All right, we're here with Josh Walsh, the CEO and founder of Branch Lab, an innovative company in the pharma ad space. I'm an advisor and Eric is an investor. So I think we get all that out of the way first. But we're not here to just pump up our investments. We want to actually hear what Josh has to say. So Josh, thank you for joining us.

[C]

Sure, thanks for having me.

[A]

So what is Branch Lab?

[C]

Yeah, so Branch Lab, in a nutshell, what we're doing is we're training different types of neural networks. We're training them on health data in a way that allows them to predict health outcomes in non health data. So it's a future, it's kind of a more future privacy forward way of activating against health care audiences today by deploying only non health data that is predictive of a given health outcome.

[A]

Non health data meaning just generic data like zip codes and stuff like that?

[C]

Yeah, like think demo psycho. Sometimes firmographic information like our customers are large scale publishers, ad tech companies and for the most part agencies that sit on top of large, rich, robust CRM data sets that are otherwise not very useful in the say pharmaceutical advertising space. We can deploy neural networks on top of that non health, non sensitive data in a way that allows it to perform to the degree that say a typical health data, legacy health data segment would.

[A]

Right. And you can't use the data raw because of HIPAA and privacy and things like that. Right.

[C]

And more regulations coming down the pike. Like even the way that we've been doing it, we. I've been in this space for a long time. You know, the way that we've been deploying health care audiences over the past 10 years is under some scrutiny with different state regulations. My health, my of Washington's important New York's S929 is another one. So it's a, it's an evolving regulatory space. It's. And for me that, that means a very interesting data problem. You know, which was, which was kind of the precipice for founding this company.

[A]

Yeah. So a lot going on in pharma. So you, you raised funding a comp. A company that we've had on Deep Intent raised a giant or sold the majority of his company for a huge amount of money. And then we have the federal government talking about banning types of, of advertising. So let's go one at a time and talk about that part. Kind of the impetus for having you here. So let's talk about you first. It's easy to talk about yourself. So you raised a strategic round recently.

[C]

Yeah, strategic round. It's kind of a fun story. You know, we were not fundraising. You know, from time to time we, we get inbound from different VC is basically you know, saying hey we, we heard of you from so and so and we want to talk. I'm in heads down builder mode, like kind of not taking those meetings at the moment. But this one came in from Lance Armstrong and he was willing to. Right. And after, after figuring out that it wasn't a bot, you know, he, he offered to, to come out to our, our HQ in Boulder, Colorado and have a sit down and he wanted to talk about our business. And you know, we were very upfront. We're not raising money so you know, we're not really having these types of conversations. But if you want to come by and Talk, let's have a cup of coffee type thing. And, you know, honestly, we were kind of impressed with Next as a fund. So next Ventures, that the team at Next and their mission, you know, they're basically. They basically came in wanting more to talk about their passion to transform the US Healthcare system more from a reactive sick care system to a more proactive system. And they view Branch lab as well. They viewed Branch as a potential piece of that larger mission. The good quest on both sides aligned really well right out the gate. As you both know, it ultimately will come down to numbers as well. And we met with them on a Monday morning and we closed the deal Friday morning. It was a good deal. They're great partners. And frankly, Lance is an American icon. Everyone knows him as an athlete, as a public health advocate. Getting to know Lance Armstrong, the venture capitalist over the last few months has been in some ways equally as impressive as his athletic background. I mean, if we look at the track record of his investments, the track record of Next, next Ventures is fun. It really speaks for itself. And the partners that he has in that firm, Jordan and Mel, are just fantastic. So, you know, like I mentioned, we weren't raising, but the stars aligned and we couldn't be happier to be in. To be in business with these people.

[A]

So it's not a family office, just with his own money. This is a legit venture capital firm.

[C]

This is legit. Like Fund two had a great. Fund one onto Fund two. They know what they're doing. You know, it was, like I said it was. It was a bit of like a learning curve for me as well. Like, get to know these folks in this capacity and. And it's been impressive.

[A]

Would you say it was a big step up from Aperium giving you money? I would say.

[C]

I would say it's a lateral step, you know.

[B]

Thank you.

[C]

Yeah, Yeah. I think in my mind it's like a one plus one equals three scenario.

[A]

Sure.

[C]

And yeah, you know, we're all going for a ride, Eric included, early next month. So we'll get the whole. Get the whole cruise.

[A]

Wait, are you serious, Eric? You going on A R to Lance Armstrong?

[B]

I mean, how could I say no, right?

[A]

I could think of a lot of ways to say no. Like, my hamstring hurts. Doctor said I can't. You should find out how many hills there are before you get on that bike.

[C]

This is true.

[B]

Josh, one question that everything moves so fast with that, with that deal. I never asked you, were you, like, a fan of Lance going in and did that, like, kind of add another dimension to the conversations.

[C]

I would say to some extent, yes. Someone like Lance wants to come out to Boulder and come to the Office and cold LinkedIn message that effectively says, hey, we've been hearing a lot about your company. Can we come talk about your company?

[A]

Sure.

[C]

I do get, like, a lot of entrepreneurs do get those messages from time to time. When you're not in fundraising mode, you kick them off till later. Maybe this one wasn't kicked off because of, to some extent from that influence.

[A]

Did lance personally Cole DM you on LinkedIn or one of his associates?

[C]

It was. I think I got a cold invite. I probably have a screenshot of it on my phone.

[A]

Okay.

[C]

By the way, Lance is going to, like, listen to this. I'm going to be really embarrassed about it.

[A]

We would love to have Lance at our next Market Live, by the way, Keynote. A lot of exposure, easy softball questions.

[C]

Well, when, when, you know, Eric's just got to keep up with him on the ride to make sure he can communicate that, like, at the front of.

[A]

The peloton while he's, like, trying to breathe, trying to survive, like, Lance, I need you to. Exactly. All right, let's transition to the next topic, which was, you know, there's this Deep Intent as a company that, you know, I've known for a while. We're going to have the CEO at our Market Live, which is great. They're sort of. I've always thought of them as sort of a niche healthcare dsp. They're sort of competitive with you, but I'd like you to just give us some perspective on it. They just sold majority of the company to a PE firm called Vitruvian for $637 million. That's a really big valuation for a niche DSP. What do we learn from that? Other than they're doing well?

[C]

Yeah, I would say Deep Intent. First of all, congrats, Chris. What a comeback story. They've been through the ringer and made it through the other side. So congratulations to Chris and team. Amazing job. I would also say Deep Intent is not competitive with Branch Lab. Like, we partner with Deep Intent. Like, a lot of our customers will end up deploying branch lab powered audiences on the Deep Intent dsp. And my background, you know, founding, founding adherent, Founding Adherent Health. A bit more competitive there, of course. Of course, both DSPs. But I think what, you know, I think what is more clear to folks who haven't worked in pharma is just the upside here. Like, this is a huge, specially regulated hyper Specific market that requires specialized tools and deep intent at that valuation is just more proof of that. And there's been plenty of proof of it over the years by the way. It's really just the latest. So yeah, congrats. Congrats to that team. Stoked for Chris, stoked for that whole organization.

[A]

So you're providing data in some cases to their customers who are executing using the dsp. Why do you need a specialist in this market? I understand why you need a specialist on the data side because you can't use the raw healthcare data in any way really. But once you convert it to non healthcare data, isn't it open game and trade desk where anyone else should be able to execute? Maybe.

[C]

It depends which lawyer you want to ask. So let me back up. Branch Lab does not provide data to anyone. Branch Lab service is a neural network that can predict onto your data and make your data useful in that way. We don't sell data, we're not selling audiences like our customers are big scale agencies that have CRMs. They want to be able to target those CRMs with the same efficacy as if they were targeting say like a health audience, like a traditional health audience or say legacy health audience. That's a distinct difference with Branch Lab. We're the only company in that space.

[A]

Yeah.

[C]

The one reason why that's why we operate in that unique way is, you know, there's a, is different states regulations and how a lawyer may read that. Now the best third party credible source that we have on the most important state regulation called My Health, My Data out of Washington comes out of Harvard Law and it's a paper entitled Can Washington Change Pharmaceutical Advertising Forever? It's really the only third party credible source that we have that breaks down. How is this law, how should this lobby interpreted in the advertising space? Right. So it's like a kind of, you know, the law wasn't written for advertising, but here's how it should apply for advertising, right? A lot of brands, a lot of publishers and a lot of large scale DSPs, for example, read that paper the way we read that. And the legacy way of basically sending health audiences is prohibited in that state and probably prohibited in other states.

[A]

Okay.

[C]

And the reason for that, Ari, and this is like a long winded run on sentence or two is because of how easily re, identifiable, identifiable, identifiable an advertising ID becomes once it's activated.

[A]

So like the classic way, if I wanted to use say a credit card transaction as a targeting parameter in online, the way it works at a very, very granular level is that at some point me as a consumer go to either a credit card page or checkout page or a bank's page and they collect online telemetry about me, cookie, IP address, et cetera, et cetera, et cetera. And then there's a match and then I go into a DSP and I can pull in those matched IDs and I can say oh this person bought something at REI or wherever. That's right. And so that is, I think that's considered pseudo anonymous. But regardless of the terms in Washington state under this law that is not allowed. You couldn't do something like that.

[C]

Again, this is like a legal argument. So I'll take on the argument coming out of that paper from Harvard. That would not be allowed. The reason why that would not be allowed is even though it's pseudonymous or some might say it's anonymized on both sides because it can easily be re identified, it fails Washington's my health, my data, upcoming regulations, you know, out of New York S929 other states have things in place. Like rapidly we're getting to a place where like each state has a different regulation. It's making this whole thing hyper complicated. So Branch lab's whole mission is to take hyper regulated data, lock it behind a neural net, never let it in or out and instead just deploy lightweight probability models. Think like a transformer neural net and that way that data is never attached to a advertising id, you know, pseudonymized, anonymized or otherwise.

[A]

All right, let's talk, let's talk the government, federal government. So Eric and I were talking in the last pod about this that like at the beginning of the year we thought the signs were that RFK Jr. As the Secretary of Health and Human Services he was going to try to push through a pharma ad ban or at least he was making noise about it. And now we're here in September and we're hearing about the so called closing the loophole on pharma ads so that they have to disclose more side effects. What's your current understanding of the political wins? And we'll get to what we think is going to happen. What do you think the current state of play is?

[C]

Yeah, I mean from a ban perspective. I mean I think this administration is taking on what really is a bipartisan winning issue, which is how can we, you know, better fight for the patient slash consumer. You know, broadly speaking, this memo that you're referring to, you know, asking or mandating rather pharmaceutical advertising to like better outline the benefits and Potential negatives of a given treatment is a good thing. You know, I don't think many folks are going to argue against this. These types of things tend to be quite bipartisan. It's a way for the, you know, quote unquote Maha movement to like broadcast that they're making progress, you know, towards this larger mission that they mentioned earlier in the year that won't come to fruition. So that's kind of the way I see the current situation.

[A]

I think you're being kind of nice about it, but it sounds like they're trying to get a political win here, regardless of whether it really moves the ball forward that much.

[C]

I think it might move the ball slightly forward.

[A]

It's like a running joke though. I mean, it's on SNL and stuff that pharma ads already have too many side effects warnings and no one pays attention to them. And the side effects are like, you may die, you know, you know, if that's a side effect, then not sure how much more grave we can get. So do you know practically if, if the current proposal goes through, how much it will change ads?

[C]

It's not, it's, it's not that. It's, it's not that specific. You know, there's some, there's some talking. I think we'll have to wait and see. But better warnings, maybe not warnings that are like sped up like 5x, like the speed of a normal human speech type of thing. Like I think we'll see some movement there, but heavily creative and messaging related.

[A]

Yeah. And it'll affect digital as well, not just tv.

[C]

Yeah, it should. I mean, the ban that you're referencing for the ban that RFK mentioned he wanted to implement more than a year ago now was specific to TV only. That's not mentioned in this more recent memo.

[A]

Cool. So just stepping back, what do you see the future over the next, you know, couple of years, political and, and just technical on pharma ads.

[C]

Yeah.

[A]

Just give us your insights of what you're seeing on the ground.

[C]

Sure. I think, I think an outright ban on TV or other is, is very, very unlikely. It won't be the first time it's been challenged by the way pharmaceutical advertising has been held up under first Amendment law in the past. And it will, it will be held up again. Frankly, this administration, you know, my prediction would be I don't think they go again. I don't think they go against it because I don't see them going against the first, First Amendment rights issue. That's my personal feeling about that if they were to.

[A]

From an ideological perspective, you mean.

[C]

I believe that would be true.

[A]

Okay.

[C]

So I do think this administration will look towards ways to, you know, perhaps put further regulation on the industry, like this memoir. I think there's a real possibility that they look at the way targeting is orchestrated, which we sort of briefly talked about earlier. And there's precedents for this. Again, precedents coming out of states. There's been some legal action out of the FTC under Lina Khan, obviously, new administration over there. But is it possible that at the federal level, you know, this administration looks at multiple states, have different regulations, do we just want to make one federal thing? And. And again, I don't think folks will fight it. When my Health My data came out of Washington, when S99 passed the new York Congress, nobody on either side of the aisle fought it. Neither did the pharma companies. So I think it's a potentially easy win to say, you know, in an inflammatory, in perhaps an inflammatory way, like, let's stop taking people's most personal data and using it to advertise to them. Now I would. Now I would, I would. I'll make a statement here, Ari. And I probably wouldn't have fully believed this without seeing the data firsthand, but when done responsibly, advertising can genuinely be one of the most powerful tools for improving public health. Let me unpack that.

[A]

Yeah, I'd love to hear.

[C]

Yeah. And there are other things that influence us, but looking at the data that we have access to, look at general statement, ethically responsibly produced advertising, healthcare, pharmaceutical advertising, it doesn't necessarily equal more sales of drugs. It equals more doctor visits. In other words, like if you didn't need a doctor to go fill a prescription. I don't know how I would feel about the industry that I've worked in for the past 10 plus years. But the best that we can do is talk to your doctor. If and what we can see empirically in the data is the more times per year that a patient spends with their more times per year they spend time with their physician, health outcome improves mathematically. There's a lot of potential to do maximum good here, both for public health and for new drug creation, et cetera, et cetera. I do think we need to be hyper ethical about messaging and very intentional about data use and these other items. If it requires regulations to ensure that the industry does that, I applaud it. I hope that doesn't need to happen. You know, in my opinion, privacy by design, technology can solve this problem. Very quickly versus having to wait for things to be settled in the courts. So that so that's my hope. But I don't see a ban here. I potentially see more regulation and the output of that I believe is better health outcomes for the US Population.

[A]

All right, there you hear, you heard it here first. Ask your doctor. That's the key to success. Stop doom scrolling on WebMD and asking ChatGPT why your hip hurt. Just ask your doctor.

[C]

So I think I would agree with that.

[A]

Josh, thank you so much for joining us. This is really informative.

[C]

Sure thing. Thank you both.

[B]

Thanks, Josh.

[A]

Thank you for listening to the marketecture podcast. New episodes come out every Friday and an insightful vendor interview is published each month. You can subscribe to our library of hundreds of executive interviews at marketecture tv. You can also sign up for free for our weekly newsletter with my original strategic insights on the week's news at News Market tv. And if you're feeling social, we operate a vibrant Slack community that you can apply to [e-mail].