
Loading summary
A
Foreign.
B
Welcome back to the Commerce Collective podcast, brought to you by Flywheel. You're listening to this Month above the Fold, a monthly series on the Commerce Collective feed where Patrick Miller, co founder of Flywheel and I cover the most important commerce topics each month. I'm Emma Irwin and today we are covering the balance between performance and transparency, the Trade Desk Q4 earnings and what they signal for the industry and valuation versus differentiation across LLMs and how branch we think thinking about investment of time, money and resources. Let's get into it. Patrick, how are you? How have you been?
A
I'm well. I'm well. I woke up with a cold, so I apologize for any sounding horsey, but, yeah, otherwise.
B
Well, I wonder what the ratio of episodes that we've done where you have some kind of cold, like, versus not hypnotized.
A
Last month it was like the only month I had off and now I'm back to being sick again. It's great.
B
Beautiful. Okay. I am curious though. This morning I woke up and I was looking for dishwasher pods on Amazon. And I can now see Amazon. Amazon now. And it was like, delivery in 12 minutes. I was like, holy cow. Do you have. You had access to Amazon? I know it was in Philly and testing out in Philly, but did that expand to Baltimore?
A
Yeah. So Baltimore had like, the BWIFC was one of the. I wanna say it's like the fourth or fifth fresh one. So we've always had access to sort of early rollouts around CPG stuff everybody makes a big deal about, like, oh, I can get it really quickly and like, yes, interesting. But I think actually more interesting is, is building the baskets and delaying. And so Amazon makes ways. Like, one, if you want it right now, you have to pay a premium. And they stagger it based on, like, how quickly you want it. So either way they make more money. But then the inverse is also true, where you can actually say, like, with precision, oh, I want it this day. And then like, bundle up everything so they solve both problems and they make money both ways. So if you delay, you're like, you get 6% on like your prime Visa and it all shows up in one box out of convenience. But if you need cough syrup right now, you can get it and pay an extra three bucks. And so I actually find that the, the network and the unit economics fascinating on this one.
B
Yeah, I just, I saw the little widget up at the top and I was like, what do you mean I can get it in 12 minutes? How? Yeah, where is it? Like, great.
A
Well, and the other thing to look for is like when you, when they deliver it, it's when are they using a network, like when are they using a prime truck versus when are they using a flex driver? And so it's the same thing with like looking at grocery and so like you can see it sort of like they're still building, you know, the plane and so it's not all on the truck but like longer term they will bridge to that and then that then sort of brings the sort of the, the outbound freight cost down, which will then bridge to more.
B
Prof. That makes sense. Okay, I will carry us into story number one, which is inspired discussion on recent piece from the substack Madison and Wall titled the Trade Desk, Google and the Fight over Performance versus Control. And it essentially, it's like taking a look at how brands think about performance versus transparency in regards to how they spend their budgets. And for more context, for anyone that hasn't read this piece, which you absolutely should and you subscribe, the substack is very good and it actually makes things make sense. And I can understand it, which means everyone can understand it. But it makes the point that is growth stuff slows and industry spending is a bit more constrained, competition intensifies, incremental gains come from share shifts, hence the investment into channels like ctv, bundling of services and offerings, et cetera. What are your thoughts on this environment that we're in? And like, can you give maybe even some examples of what we really mean by like performance versus transparency in context to this article?
A
So it's this natural tension that like brands and agencies will say, hey, I want transparency. And they want to like, they'll ask for things like log level data or they're asked for like, oh, what's my impact of, you know, of shares? But in reality what they really want is performance. And so it's like, it's a little bit, you know, hand wavy, you know, of, of the transparency because like given a choice between something that is performant and opaque versus transparent and doesn't work as well, 90% of the time they're going to go with where the performance is. But at the same time it's like, okay, hey, industry is slowing down, advertising slowing down, tariffs, macro, blah, blah, blah. And then it's like okay, now they can make their dollars go further, they can use as more leverage and they can say, okay, well I want to Y or Z but you know, you have to give me whatever transparency. And whether that's log files or what have you, there's then Sort of the tension with like if you think about like how bidding works, it's like it's not, you can't explicitly, you can't bring a scalpel to the game. Right? Like, and because like, like we would always have like these debates with the platforms where it's like I don't want to bid on that inventory because it's like it doesn't work as well. And they would be like no, you kind of like you have to. And so it's sort of like when you think through like bid modifiers and like the sort of like how they think through like the economic of like where the ads are showing up, you know, that becomes a fascinating question. And I think lastly there's the question of like who is taking the risk? And so what I mean by that is like if you're buying on a CPC basis, the platform is taking the risk because you don't pay unless the money clicks. That said, it's like a 90% gross margin business. So like you know, that much risk versus like if you're buying on the open web or like stv, you're buying on a CPM basis. So you, the brand are taking all the risk. And so that to me is sort of the real tension. It's like who is taking the risk? And then do you justify that risk? You know, be a performance and then what transparency do you need to justify the reporting to say that hey it is that performance.
B
I'm curious like the reasoning for wanting more and more transparency if at the end of the day they really want performance, like was that just kind of like fluff and trying to be better, does it come down to like quality of the inventory that they were buying that they wanted more transparency?
A
It's a power dynamic. And it's like I remember earlier on like when like brands were like I really want like on sponsored ads and sponsored products, like what's on mobile, what's on desktop? And you're like okay, but like they don't give you bidding controls versus between the two. So like what are you going to do about it? It's like a, you know, it becomes like a navel gazing exercise. And so like yeah, you know the breakdown but like, well why? What does that actually get you? And so you know, I think sort of like the end of the day like what are the brands trying to do? Well, they want the transparency to then sort of better look across the various platforms and what is then the efficacy? And so that's like whereas like the platforms are going to be like, hey, give me your money. And then I'm going to decide like how I sort of, you know, spread it across the various units and then, you know, and then how do I sort of make money, you know, in sort of across all of it? And so as the brands get more transparency, they can then give, you know, pushback to the platforms and be like, hey, I don't want to, I don't want my ad going here, this type of bidding or whatever, this user. So it's just the natural tension between, you know, the pubs and the brands
B
at the platform level, like, do you have like one that ranks the highest for performance and one that ranks the highest for transparency? Just to kind of like level set,
A
it will be like the open web ones will be the most transparent, so they're also the least performant. And so it's like Google doesn't have to be very transparent on like their SERP ads because guess what, they work really well and they got a line of people around the block. Okay, if I'm doing a 300 by 250 on the Open web, it's really transparent, but it doesn't work that well. And so that's the tension and sort of like, and you know, look, if, if Google opens up more and they start talking about, you know, you know, serp and you know, it's the same thing with like the LLMs now. And I know by the way it's probabilistic and they don't even know, but like, some of this is just like the machines doing the machine things and other things where it's like, so you don't even, you couldn't even have the transparency if you wanted. But it's also not necessarily in the economic best interest, you know, of the platforms. And that's where you see things like the ANA and other industry groups that will sort of try to push for this, see where it goes.
B
For your mention of the open Internet, this helps transition us into the next story nicely because I can actually say now, all these years later, I genuinely understand what is the open Internet versus what of what is a walled garden.
A
Well, at your age, you don't, you never, you've never been to the open Internet because you're like, you're like, why would I go there? There's nothing good.
B
Oh yeah. And I, and it's like my point, I read all this commentary of like, oh, basically. And I mean, the commentary's been around for like 10 plus years of like, the Internet will die and fade away. And I don't know how I feel about that. But anyways, on this element of like, open Internet, which trade desk is a champion of, and that'll be the next story kind of talking about their earnings and what that signals for the industry. But is it possible, like, can the open Internet, like realistically win on performance ever? Is that just like an entirely dying art? Is there something they can do to get better in terms of performance versus transparency?
A
Yeah, I mean, like, the LLMs are gonna gut a lot of the garbage sites and so like, we're gonna go through like the valley of death where like all like the awful, like recipe sites will go out of business. And that will be wonderful because I will be able to cook chicken and figure out what the temperature is without having to scroll through like endless ads that get stupid view attribution two weeks later. It doesn't make any sense. And so like when they go out of business and then you're left with like actual quality sites that do work, like, yeah, it's gonna be fine. It's not gonna be huge, you know, especially relative to like STV or the wal, but like, it's going to be a nice business and it's going to work and you're going to get rid of all the garbage and all the charlatans. Good thing, you know, maybe a lot of fake reviews, you know, we'll see. So, yeah, I mean, that's, I'm, I'm sort of hopeful of that on the other side of this, you know, at the same time, a lot of the plumbing is then going to, you know, sort of help drive how people plug into STV and, you know, and what that looks like. So, like, in many ways I got to get like, the open web is like, it's where you can experiment with like bidding and audiences and you know, frequency caps and day parting and geos. But it's inexpensive. Like you're paying a couple of bucks, you know, CPM as opposed to, you know, STV where 20 bucks. So great place to experiment, figure things out. I mean, the problem is like, you know, depending on what signal you're using, clicker versus View, you can have a lot of noise there just because of sort of, you know, false positives.
B
Gotcha. Okay, digging into this is Cleveland Research Company did a whole wonderful insights and takeaways from the Trade Desk Q4 earnings. And then kind of comparing that also to Magnite and other publishers. I think the Trade Desk reported slowest growth that they've seen in five years. And so I think you, you've Kind of touched on this. But like why is that? Why are they specifically seeing that slowed growth versus the magnite on the other side, which it couldn't be real. I had to Google what that was and figure that out. But now I know what sell side really means.
A
Well, I mean with the trade desk, their CEO in the last few earnings calls has been asked about, you know, sort of competition with Amazon and he has explicitly said that they are not a compet competitor. And, and like, and a lot of people in the industry are like, huh. And now this, you know, this quarter, you know, he's acknowledging that and I'm just like, man, like as an investor it's like CEO gotta be straight and just acknowledge like, you know, what's going on. And the fact that they did not acknowledge the competition with ADSP for many quarters in a row, it's just like what, like what are you doing? And so now all of a sudden like they're seeing a slowdown and so they, you know, then sort of turn the knob up, you know, on their take rate and like, well, how sustainable is that really? And so I think it was a huge miss to not acknowledge that Amazon was coming for them and in the research community have long called that out. But now that they actually are acknowledging on the earnings calls and you've seen a sell off after that and after that. And to me the big question for the investment community is how much of a trust buster was that? And then what does this look like going forward? And what is the role for the trade desk relative to the other DSPs that are out there? And what does the connectivity look like especially to STV as they pivot away from the open web.
B
In your opinion, what is the role of the trade desk going forward compared to other DSPs?
A
It will be interesting to see what their partnerships with like Walmart and some of the other retailers look like is sort of what I'm most interested in. And so you know, if they are, you know, providing, you know, sort of the bidding. But if like some, if those retailers, you know, decide to build, you know, their vehicles to buy video, you know, or display advertising themselves, that's going to be a huge headwind for them. And so it's like if I'm trade desk, it's like, okay, well I'm going to lean into transparency. I'm going to give you, you know, you know, whatever, you know, information you need. To me though, it's like just like what do the retailers do, what do the partners do? They're leveraging Their tech. And then, you know, what does that look like going forward? What's their moat? You know, quite frankly, like, and that's, you know, it's look, it's a tricky one.
B
I think about how even I like was I would see the commentary of not treating Amazon as a legitimate competitor. I was like, I don't know a lot but that what, how I do know that that seems like a bad
A
take and it's coming from the CEO. Like, like on earnings calls. Like, that's what I don't get. Like I would like listen to them and be like, oh, all right, you know something? I don't.
B
But okay, last question on this one. And this is, this is where you said keep it simple. So I'm really dumbing things down. And since this, so this email talks about Magnite and kind of they saw this great, not great, but like promising Q4 results and that is talking about the sell side of things. So I'm trying to, to like comprehend in my brain, okay, we have this really large cells, if not the largest, like sell side, platform growing, seeing good growth. But then we have a dsp, the tripe test kind of this slowed growth and it's concerning. How should I think about that? Is it that like the other DSPs such as Amazon continue to grow and that's what's fueling. Okay, I was like, yeah.
A
Yes, yeah, exactly. It's sort of like if you're seeing if the, if you know, if the sell side's growing and some of the demand side is not, and then you're just like, okay, well, it's coincid somewhere. I don't know we're going to be going.
B
Well, yeah, well, the email, I was like, okay, like the email is making me think and tie these dots together. But okay, cool, we got confidence.
A
But it's also, it's important because it's like Google will break out their DSP business and that is, you know, sequential decline for years. You know, Trade Desk is nothing but a dsp. And then you know, something like an Amazon, like they don't break it out, so get full transparency. And so it's sort of like a game of triangulating. Like what do people say publicly and how does that then, you know, align with these, you know, these various businesses?
B
Last up story number three, friend of the show, Jason Del Rey wrote this really awesome piece about the state of AEO looking at platform valuations versus like what brands are actually seeing when it comes to differentiation and figuring out how to invest there. And we can't give everything away because I know he's working hard to build the aisle and everybody should subscribe and follow along with that. But for the gist of it, like VCs, they're valuing profound, for example, at like a billion dollars. As I said, brands are saying, you know, it's hard to tell the difference between a lot of these various platforms and what they offer and measurements. Very hairy. There's a significant price compression going on, which I know interests you. Can you give your, your overview of what's going on there?
A
Sure. Jason has launched a new, you know, substack the aisle. It's a, it's a very good read and it's much closer to his writing during Recode, so more industry specific as opposed to sort of general audience, encourage folks to, to check it out. So what he did was he, he went and he, and he interviewed sort of a number of folks, you know, folks on the brand side, you know, some agen folks, you know, and then just people in the ecosystem. Well, how are brands thinking about tracking, you know, success, you know, within these, you know, within these LLMs. And because, you know, none of the LLMs are giving out information, you know, as far as performance thus far. And so you've got like a cottage industry, you know, of, of companies that have popped up, you know, that are, you know, sort of pinging the LLMs that are trying to figure out, like, how are they returning and, you know, doing some really creative, interesting things. What I find fascinating that he pointed out is the price compression, you know, over the last year. And so it's, it's like if I go back to like when Amazon started, you know, taking off, like, it's like people were aware of it, but it wasn't like on the front page of like the Wall Street Journal every single day. And versus, like the like versus the LLMs, it's like it is every single day like people are talking about them. And so you have then just like a rush of people and companies and money that is being thrown at, like, hey, how do we figure out LLMs, you know, both from, you know, an oper, you know, you know, perspective, but also sort of like a reporting perspective. And so, you know, you know, you know, Jason had a quotation from our brand of just talking about like, you know, people's, you know, you know, price, you know, cutting in, you know, 50, 75%, you know, in the last, you know, in the last year. And so that tells me it's like, okay, there are a whole lot of People that are trying to sort of, you know, figure this out and are coming up, you know, with novel ideas, but there's lack of differentiation. And so that means like these companies do not have pricing power. So I, so if I'm a brand I would be, you know, I would not want to lock into a long term contract here just because it looks like that there's going to be more and more price compression because it's just there's so many people trying to figure this out and then like what is actually differentiated around them. And he further goes into, you know, a, you know, some of some founder he was talking to a larger brand and not being crisp and precise with, you know, how the tools are working. And to me that's always a red flag. And like it's like, I get it, like everybody's got like their IP and like you need to protect that but at the same time, like you need to be able to explain to your customer how things work in a way that builds trust. If folks are not doing that, like, you know, for me, like, I'd be real hesitant of engaging.
B
Would you say? Does it seem like in terms of like all of the kind of like technical technological innovations and things that have happened specific to kind of the advertising industry and even tech in general, like this one is moving faster in terms of adoption. And it's like a lot of what I've read seems like brands are kind of just like throwing money at it and seeing what works. And this is really the first, first like kind of new thing that I've really been able to grasp as to what's happening per my little five year career. But is it moving faster? Is it, Are like people throwing money at it faster than they did for let's say like sponsored product? Okay.
A
Oh yeah, well, because just look at the Capex, like look how much like an Amazon, you know, or Microsoft or Google are spending on this stuff. And that's the, that's the like, that's the tell. I mean hell, just look at your electricity bill and how you're paying more because the data centers, you know, also a tell. But then like sort of, then the second way to look at it is like, well, how is the growth of the data centers? How are they funding it? And so if I'm a Google or an Amazon, I was like, well, my business spits out enough cash that I can then just spend whatever money I want on data centers. And even if everything crashes, it's like lame broadband 25 years ago. It's like, okay, yeah, we had the dot com bust, but eventually all the broadband cables were used. It's just the thesis was correct, the timing was wrong. The data centers is likely the same thing, but the risk is, well, how are they financing it? So if you have business that is like spitting out like a bunch of cash and then it's like, okay, so investors will get nervous because like, you know, they're, they're sucking up all the cash to put in data centers, but then it's like others are like taking out a bunch of debt and so like a meta, you know, or Oracle, which is like terrifying, like, you know, and just the amount of debt that are taken out and they're buying like Warner Brothers. That's a whole nother thing. You know, it's like, okay, this is riskier because it's all debt. And then lastly, you then get into people that are doing sort of circular financing like, with like OpenAI, where it's like, God knows, like, you know, they got some really creative finance people and it's like, and, and everybody's taken, you know, is, is putting money into, into OpenAI. But it's like, it's like how much of that is just a hedge? And like, maybe they become Google, you know, maybe it flames out, you know. Either way, like, everybody wants to put a little bit of money in just as a hedge, just in case they actually do pull off, you know, just the, the greatest landing, you know, of all time.
B
There you go.
A
Boom.
B
Thank you as always for your insights. And that's it for this month's commerce news. We'll be back next week. Month covering everything you need to stay on top of it. In commerce, I'm Emma Irwin and we'll see you next time.
The Commerce Collective Podcast – Episode Summary
Episode: This Month Above the Fold – Performance vs. Transparency, Industry Signals from The Trade Desk, & LLM Measurement
Date: March 5, 2026
Host: Emma Irwin
Guest: Patrick Miller (Co-founder, Flywheel Digital)
This episode dives deep into three headline topics: the ongoing industry tension between performance and transparency in ad buying, analysis of The Trade Desk’s Q4 earnings and their broader implications, and the current state and challenges of measuring success in the rapidly evolving world of large language models (LLMs). With characteristic wit and candor, hosts Emma Irwin and Patrick Miller dissect these industry flashpoints, forecasting trends and unpacking challenges that commerce players face today.
[00:56 – 02:39]
[02:39 – 07:34]
Inspired by the Madison and Wall Substack piece, Emma introduces the core dilemma: as growth slows and budgets tighten, brands focus more acutely on choosing between ad channel performance and transparency.
Patrick defines the “tension”:
Power Dynamics & Platform Nuance:
Transparency requests are bargaining chips for brands seeking better deals, but platforms balance what they show to maintain their economic advantage.
Patrick’s memorable summary:
[07:34 – 09:44]
[09:44 – 13:37]
[14:00 – 19:21]
On Amazon’s delivery experience:
“I saw the little widget up at the top and I was like, what do you mean I can get it in 12 minutes? How?” – Emma (02:06)
On the nature of performance vs. transparency:
“Given a choice between something that is performant and opaque versus transparent and doesn't work as well, 90% of the time they're going to go with where the performance is.” – Patrick (03:51)
On the dying “open internet”:
“The LLMs are gonna gut a lot of the garbage sites… you’re going to get rid of all the charlatans. Good thing.” – Patrick (08:32)
On The Trade Desk’s competitive missteps:
“I think it was a huge miss to not acknowledge that Amazon was coming for them…” – Patrick (10:56)
On the fevered state of LLM measurement vendors:
“Brands are… throwing money at it and seeing what works… But there’s lack of differentiation… these companies do not have pricing power.” – Patrick (14:41)
On the hype cycle being faster this time:
“Just look at your electricity bill and how you're paying more because the data centers… [are] a tell.” – Patrick (17:40)
Episode wrap-up:
“That’s it for this month’s commerce news. We’ll be back next [month] covering everything you need to stay on top of it…” – Emma (19:21)
This episode is a must-listen for anyone seeking real talk on where commerce media, ad tech, and retail innovation are heading—whether you’re leading a legacy brand or launching your first DTC campaign.