Subscribe: Google Podcasts | RSS
Subscribe: Google Podcasts | RSS
As we put the awfulness of 2020 in the rearview mirror, we thought it might be fun to look back to another bleak period: the 2007-2008 financial crisis! Why? Because Tim hasn’t stopped talking about Subprime Attention Crisis—the Tim Hwang book that draws a parallel between the digital advertising ecosystem and the subprime lending crisis from a decade ago—we decided to all give it a read and then sit down for a discussion with the author. From the opacity brought on by the many moving parts to misaligned incentives to the fact that, well, even more than just the internet is built on digital advertising dollars, it was a fascinating discussion!
0:00:04.4 Announcer: Welcome to the Digital Analytics Power Hour. Michael, Moe, Tim and the occasional guest discussing analytics issues of the day. And periodically using explicit language while doing so. Find them on the web at analyticshour.io and on Twitter @AnalyticsHour. And now the Digital Analytics Power Hour.
0:00:27.9 Michael Helbling: Hi, everyone, welcome to the Digital Analytics Power Hour. This is episode 159. In real estate, the old phrase is location, location, location, and maybe when you’re writing books about current events, it’s like comedy where timing is what matters most. But whatever the case, we’re here today to go one level deeper on a topic that actually most of us touch almost every day in our jobs as analysts. It’s the digital advertising space. Hey, Tim Wilson, you’ve made the book “The Subprime Attention Crisis” a last call on the show and subsequently mentioned it multiple episodes since it came out this past fall. Are you super stoked to be so plugged into the zeitgeist here at the dawn of what is hopefully a post-COVID-19 world?
0:01:13.0 Tim Wilson: I feel like I’ve been living with this topic for years and had not seen somebody put it together with so much evidence, so I am. It is like the last part of Christmas for me, it’s the last gift.
0:01:28.8 MH: There you go. The season of giving can now officially end with this episode. And Moe, how are you doing?
0:01:34.8 Moe Kiss: I’m good, I’m just trying to plot how I’m gonna get a word in with Tim today.
0:01:39.1 MH: Literally my next question is like, “Are you gonna ready to join me and try to get a word in edge-wise?” on episode, I’m sure will be awesome. And I’m Michael Helbling. But if we’re gonna discuss the book, “The Subprime Attention Crisis,” we needed a guest, someone who could lend us their insight into it, so we got the book’s author. Tim Hwang is a writer and researcher, he’s the former director of the Harvard MIT Ethics and Governance of AI Initiative and previously served as the global public policy lead for artificial intelligence and machine learning at Google. His work has appeared in the New York Times, the Washington Post, Wired, The Atlantic, and The Wall Street Journal, amongst other publications. He lives in New York City. And today he is our guest. Welcome to the show, Tim.
0:02:26.1 Tim Hwang: Hi, everybody. Yeah, thanks for having me on the show.
0:02:28.7 MH: Your timing is excellent, we’re delighted to get the chance to chat with you and by delighted to chat, I mean Tim has… Tim Wilson has a series of questions here…
0:02:41.4 TH: And I am ready to answer them, so how can I help?
0:02:46.3 MH: No. It’s really exciting, and we’ve all had a chance to read your book, and I think all of us found it pretty thought-provoking and fascinating. I think as a great starting point… You spend a couple of the first two chapters of the book laying up the landscape of the digital advertising world, and I think that’s probably a good… At a very high level, let’s have a conversation starting maybe there. Most of our listeners will have some familiarity, but good for them to understand like, yeah, what are you really kind of like… What is the case you’re making with this book?
0:03:16.1 MK: The thesis, if you will.
0:03:18.0 TH: Sure, yeah, yeah. There you go.
0:03:22.6 MH: Very literary, Moe.
0:03:25.8 TH: The best place to start with that, I think is just like how the book came to be, and I’m really glad that that is in fact the first question, because I think that’s the point of the book, in many ways, whether or not you buy the argument is just to explain in simple terms, “What is this money machine that’s at the centre of the internet?” And the inspiration for the book really actually came from the time that I was working at Google, and what’s surprising is for a company where 80% plus of its revenue comes from ads, basically the business model of the company is a rumor. You talk to people on a daily basis, they’re excited about AI, they’re excited about self-driving cars. But even engineers at Google, you say, “How do we make money? And could you explain the system?” And a lot of them don’t… Frankly don’t know.
0:04:09.0 TH: Right, and so the point of the book, at the very first part of it, is really to explain how online advertising functions on a day-to-day basis. And there’s a lot of complexity, but the basics are pretty simple, and some of it is that ads are same as it ever was, you have people who have attention that they wanna sell, these are people who are known usually as publishers. So this could be everything from your local newspaper that has a website to Facebook or Google, and on the other hand you have people who wanna buy that attention, they wanna spend money to get their message in front of people, and those are the ad buyers. And so far, that’s basically the structure, the supply and demand of the ad economy that really has defined how ads have always worked.
0:04:52.7 TH: What’s different with the advertising that you have online though, is that’s a form of advertising that’s known as programmatic advertising. And essentially what it is, is that you have basically the buying and selling of attention coordinated through algorithms, and specifically these algorithms that compete in these auctions on a split second basis. So typically the way it works is that you load a website and immediately at the point in which you load the website, there’s a signal that goes out to this marketplace and says, “I have a male 25 to 35 living in New York. Does anyone wanna buy an ad to deliver to him?”
0:05:24.6 TH: And at that split second there’s basically an auction that occurs, and depending on who wins, they basically earn the right to load that ad. And all of that happens basically between the time in which I click on the link to when the website loads, and this happens basically billions and billions of times a day. And this really is sort of the rocket fuel, this is the engine that is really creating Willy Wonka amounts of money to companies like Google and Facebook.
0:05:48.2 MK: And that’s the bit that I… Like the way that you explained it actually in the book, I had several very conflicting emotional thoughts reading this. But the one in particular, when you describe that, is like, “Oh, you’re scrolling through things and you’ve seen an ad and a transaction has happened… ” I mean, obviously, we work in the data space, so I guess we know, kind of know that a transaction’s happened, but I’ve never actually thought about it as like, “Oh, I just took part in a transaction that’s happened online that I had no input or oversight of. It left me feeling very strange. I’m not gonna lie.
0:06:27.3 TH: Yeah. And I think it is like… What’s amazing about it is, as from a pure engineering feat standpoint, it’s crazy, what we’re talking about. We’ve created this massive global infrastructure that does this billions and billions of times a day, and it’s so funny because I think on a day-to-day basis, we’re sort of not aware of these transactions that are always happening. And I do think that there’s almost a feeling of peering behind the curtain a little bit, which I think is a little bit uncanny. It’s a little bit unsettling.
0:06:56.3 TW: But I think one of the things you quickly get into, and this is where is… Coming from a digital analytics background, and for the first, I don’t know, five to eight years of that, never really working closely with media agencies, and this was pre-programmatic, when there was still a lot more transparency, but it was felt to be a blunt instrument. I may be buying ads on a specific property or a set of properties, or even if I’m buying audiences, it wasn’t full-on programmatic, but it does seem like there was this idea, this race to get more efficient is a really good story.
0:07:38.8 TW: We can build the machines, we can get the response time to be quick enough, we can now get everything in the magic really fast black box that’s using machine learning and AI, but where… Can you maybe talk about how… That felt… That was a really good story, but then ironically, it actually introduces way less visibility if you say, “I bought a million impressions, but I have no idea where they are shown. All I know is the type of person that I was expecting to have them put in front of. And I have to take this big a leap of faith that they were put in front of the audience I wanted to put them in front of.”
0:08:22.2 TH: Yeah, that’s right, I think the core origin of programmatic advertising, like this crazy algorithmic Rube Goldberg machine that we’ve created, it’s really driven by scale, and it’s in some ways, no surprise that Google was one of the first companies to really implement these systems. Because suddenly when you’re serving up search results and you wanna serve ads against them, you really can’t have a salesperson that takes a call and says, “Oh, you’d like to deliver an ad on this time and this date. Sure, we’ll put that up on the website.”
0:08:49.2 TH: And I think in some ways that is really one of the key dimensions that has also simultaneously produced a lot of opacity in the market, which is when you deliver ads online, you will sometimes deliver hundreds of ads to huge amounts of people, and suddenly it becomes very difficult to track where your ad goes, and in fact why your ad ends up in a certain place. And I think that is one of the features of the marketplace is ironically, even though we have a lot more data about when the ad is delivered and how much money you spent on it and that sort of thing, it’s also introduced a lot of new opacity in the sense that this plumbing has ads end up in all sorts of places where we don’t always understand why it happened the way that it did.
0:09:31.3 MK: I’m just thinking a lot about budgets at the moment, and suffice to say there has been lots of shift from some of our more traditional advertising platforms like TV and radio, where you do still call up a person and you book a time and all that sort of stuff. And without getting into whether either marketing channel is effective or works, ’cause that’s a whole another discussion, companies largely have shifted a huge portion of their budget to that programmatic advertising. Do you think that’s fundamentally because all of the ease of it, or do you think it’s also very deeply coupled with, we think it works, so we have access to more data, like what really drove that shift?
0:10:14.2 TH: Yeah, so there’s a fascinating anecdote this reminds me of in the early history of Google, where basically Google for a period of time has a team of sales people who do it the old school way based in New York. And they’re constantly paranoid that basically like Mountain View is attempting to automate them with all this programmatic advertising stuff. And I do think that one way of thinking about the relationship of these companies, the new digital giants with the media channels that have existed in the past, really is a competition between sort of this data-driven automated way of delivering ads and then sort of manual seat-of-your-pants way of doing so.
0:10:50.1 TH: And I think in some ways, Moe, to your point, I do think that the perceived credibility, the perceived power of delivering messages in this way is one of the reasons that the dollars have flowed from one place to the next. And I think the question my book asks is, “Is this kind of intuitive fairy tale we tell ourselves,” which is like, “I have more data, so I should be able to target my messaging better?” Is that really the case, and is the programmatic advertising system really delivering on its promise, and what does it mean that it if it doesn’t deliver on its promise?
0:11:20.7 TW: And is there a lineage? It’s interesting that if you go to paid search, when you’re getting a super powerful signal and that somebody is typing… They type thong into their search bar and…
0:11:34.4 MK: Wait, are we talking about Australian thongs?
0:11:38.1 TW: This example, actually, Moe, was leading up to a localised… Exactly where I was heading. So you saw where I was going. But if I type that in and if I’m a company that sells thongs generically of some sort, then I’m buying on that, because I know right now this person has expressed that they have an interest in it, and then it does become a, “Well, wait a minute, are you looking… Where are you? Are you looking for a footwear, or are you looking for underwear?” Okay, well, Google also generally has a pretty deterministic sense of where you are, so now I can geo-target it, and so that seemed like a very, very good…
0:12:15.6 TW: And even just what’s happened with AdWords, with all the bidding and all the thought that went into figuring out relevance and the way that machine has grown, it does seem like there’s this idea that when you’re saying, “Oh, I wanna target 18 to 24-year old females who have children and are making decisions about healthcare.” That there’s this idea that, “Oh, we have the same level of specificity about targeting that person in many different places on the internet,” that we have of somebody who said, “Thong, I’m interested in thong.”
0:12:50.4 TW: Is that, again, sort of part of that myth that we treat all signals… We want to think that any time there’s a field available that we have very specific knowledge, which we did with it, what are they looking for on the web now, you don’t know if they wanna buy a thong or clean a thong or whatever, but still you had… That was much more a stronger piece of information than these kind of attitudinal things.
0:13:20.3 TH: Yeah, I think there’s two subtleties here. I think one of the subtleties, which is important, and then I think the book sometimes gets misinterpreted in some of the interviews that I’ve done, which is people say, “Oh well, you think that ads just don’t work then.” And I think that isn’t the claim, we definitely have cases in which it seems pretty certain that the ads work. The famous case in search ads is like Mesothelioma, if you’re searching for an obscure medical condition, it’s because you probably have it, and that ends up being a really powerful signal to deliver an ad. The question the book is asking in some ways is like, is that case that we can think of, is that the norm, or is that the outlier?
0:13:56.4 TH: And I tend to think that it is the outlier just because even in search, which has this reputation of being like the strongest signals, there’s so many things you can search for that provide totally ambiguous things you might deliver ads against. And so I do think that we have been lulled into a series of received wisdom about what works and what doesn’t online, that may actually not really be the case once you start looking into it. Again, I think, Tim, one of the subtleties there too, that I’m sure we’re gonna get to in this interview is gonna be, is the data that you even have on that person accurate? We can get to the question of, does an ad work if you’re actually targeting it to the person that you intend, and the whole system is working as intended? But there’s so many lines of intermediation and places where it can fail, and one of them is just like, “How much do we know that we’re actually targeting the right person? And we think that we are.”
0:14:43.5 MH: So yeah, something that really stood out to me as I read the book was that historically, the publisher was the person who provided the information about the person who was being advertised to, and in this ecosystem, we’ve really changed the ownership of that data. Now the platforms and networks own all that information, and they’re the ones controlling who gets what targeting, so Google knows everything about you, Facebook knows everything about you, and so whatever advertisement you get on time.com, is it really being driven by Time as a publisher, as much as the network that they’re choosing to participate in, and that’s doing that? And I just thought that was very interesting, ’cause when you talk about data and personal data, which you sort of… EU concept under GDPR, that has implications potentially.
0:15:39.1 TH: Yeah, I think that’s right. Again, I think this is one of the reasons it’s really useful to think about ads as almost like a lens for thinking about everything that happens on the internet. You can almost look at anything that happens on the internet and be like, “How did advertising inform the way this turned out?” And I do think that… Michael, your point it’s a really good point, which is basically that even though the surface of it doesn’t seem to have changed all that much, the underlying question about who’s holding on to the data and who owns the data has radically shifted. And so in some ways, like what’s immediately obvious may not actually reflect the deeper reality of the situation.
0:16:17.7 MK: And that’s the funny thing, is that now these marketing platforms are also asking us for more data. When you send them a particular list that you wanna target, they’re like, “Oh, you can enrich it with this, this and this, and that will help us with better targeting, which will make your ROI more efficient.” And they try and incentivise you to share even more data about your user base, which ultimately there then the… The word guardian comes to mind, but it feels really inappropriate in this case. And that just makes their engine better and bigger, which…
0:16:54.7 MK: There was no question there, disappointment.
0:16:56.9 TH: Yeah, just general existential angst. Yeah, no, I think that’s right. I do think that in some ways… ‘Cause I come out slightly differently on the debate of, is this data all worth anything in the end. But I do think that regardless of whether or not it’s actually effective at targeting, it’s actually effective at making ads more impactful. In some ways, the collection of data, the aggregation of data almost has gotten its own logic now, where it’s just like, “We just want more data,” because if you have more data, then you’re better, and then if you’re better, you get paid more. And I do think that that does create these perverse incentives in the space, which is like even if the core of it is no longer working, we would still see a market where people are trying to aggregate all this data, which has these very real, like say, privacy consequences.
0:17:43.0 TW: And the book, sort of… The connection you make is from this space of digital advertising to the 2008 subprime lending crisis, the financial crisis, and maybe if you could talk through that a little bit because it was… Or maybe I’ll give my very quick version and you can correct it. If the subprime lending crisis was all of these obscure… These financial instruments that we’re actually propping up an entire financial sector, nobody understood it, there were some shenanigans going on, and that was like powering economies, and then it fell like a house of cards. And since digital advertising is in many ways powering the internet, and that’s not an uncommon argument for the advertisers or for the publishers to make say, “Look, advertising is not bad. We’re funding the internet.”
0:18:40.0 TW: And my understanding of your case is, well it’s the same thing it’s built on a house of cards or a bubble that’s due to burst, and all we’ve really touched on is the opacity of, “Does it really work?” But you laid out other ways that it’s analogous to the financial crisis.
0:19:00.0 TH: Yeah, I think there’s basically a lot of lessons to learn. It’s actually funny, I was joking with a friend the other week, which is like, you usually try to pick an analogy because it’s simpler than the topic that you’re talking about.
0:19:11.6 TH: The book landed on the 2008 financial crisis, which is famously wonky and complicated and difficult to fully get in your head.
0:19:21.1 TW: Read this Michael Lewis’ book as the pre-requisite for this book…
0:19:24.7 MH: That’s right.
0:19:24.9 TH: Yeah, exactly. My mom was like, “This is the most boring book I could ever… ” So…
0:19:29.4 TH: Yeah, I think that’s right. So I think there’s a lot of lessons that can be learned from the 2007, 2008 financial crisis. But I think maybe let’s… Before we dive into that, there’s just the preliminary question, which is, is it a good metaphor? Is it a good parallel to what we’re seeing in the ad market? And the argument that I make in the book is really like you look at the history of financial bubbles, market bubbles of all kinds, 2008, but also… Take your pick. And there’s really these three elements. One of them is what we talked about, which is opacity. It’s very difficult to know what’s going on in the marketplace. And that’s really important because the second element, which is what I call to subprime asset value. The thing that we thought was really valuable is actually decreasing in value over time.
0:20:13.1 TH: And opacity basically, allows that to happen without anyone really realising that it’s going on. Alright. And then the final dimension that you need is this perverse ecosystem of people that have incentives to keep flogging the value of it even though that might not actually match the reality. And that’s certainly something that you had in the subprime mortgage crisis. You had these very opaque assets. You had these mortgages that were going under. And then you had these banks and rating agencies that had very strong incentives to push the market even though there were these structural problems with it. And the argument is the same exists in ads.
0:20:49.1 TH: The market’s really opaque. There’s only a few people who really understand it. The quality of ads and the attention they capture is declining over time. And then you have this whole ecosystem of people who have made fortunes on pushing the dream of programmatic ads. And at least, what that says to me is like, “Yeah, okay, we have all the ingredients that kind of set us up for a market bubble.” And so that’s where I’m at.
0:21:10.0 TW: Well on that group that are all sort of… I mean, it really is… It’s like the ad tech vendors who are the technology supporting and the infrastructure supporting programmatic, they’re incentivised. The media agencies are saying, “Give us this money. We have access to this magical ecosystem. And we get compensated based on the more you spend the more we get paid.” And then you’ve got even the advertisers it seems… Well, I guess… Even the advertisers are generally incentivised… I think you got into this. It’s gonna be hard for them to say… Or maybe this was in the freakonomics episodes… That it’s gonna be hard to say, “Well, I bought into this for the last five or eight or 10 years. That’s what I’ve been telling the board I’m spending the money on.”
0:21:55.1 TW: And by the way, the more I spend the more my stature grows and the next bigger CMO job I can get. So it’s literally no one is incentivised to… Even though there’s lip service to efficiency it’s… Everyone is incentivised to continue to let this grow, which it does seem like is analogous to the financial crisis.
0:22:19.7 TH: Yeah, I think that’s right. And I think one of the things that that example teaches us is that people in the game can know the problems in the game for a really long time and have very limited incentives to solve it. So in the 2007 crisis banks were like, “Yeah, we just gotta keep this going, even though we know it’s not really gonna stay around for long.” And I think the same is here… The case here as well in the ad markets, which is like… I’ve heard from some people in the ad industry and they’re like, “Yeah, we know about ad fraud. We know about the rise of ad blockers. We know that there’s opacity in the market.” So a lot of these problems are well known in the industry. The question is like, “Well, do you think that they’re gonna take the painful steps necessary to stop it?” And my answer would be no.
0:23:03.1 MK: But that’s the bit that left me… Like I said, I had all these mixed emotions reading this book. And the glimpse into what would the internet look like if this did fall over was something that… I found myself like sitting in the garden, potting plants, thinking about, “Would I be prepared to pay for Google Maps because Google doesn’t have a revenue stream from advertising?” And I was like, “Oh yeah. You know, I’d be fine to pay for every time I needed to use Google Maps.” And then you make this really eloquent point about the fact that the whole internet is based on this, and that there are people… Groups of people that can’t afford to pay for things like that. And then it just… Yeah, do you wanna paint us a little bit of a picture of what it would look like if this whole advertising world blew up?
0:23:52.2 TH: Sure, yeah. With pleasure. I’d be happy to do that. I think that…
0:23:58.1 TH: Again, this is like lessons we can learn from 2007, 2008. Part of the trouble of what we have here is that the bubble has grown so quickly. It generates so much money that it’s very difficult. We may have all of these disagreements with the privacy consequences of this ecosystem and what not, but we’re chained to it in some ways. And I do think that a bubble bursting would have huge impacts. And there’s three ways that I usually think about it. So one of them, which is on a very personal level, is to think about all the services you rely on everyday, which are free, and subsidised through advertising that you might suddenly lose access to. So it’s a little bit like a Google Maps question.
0:24:37.4 TH: And I think the huge inequality that might emerge from a world in which suddenly like, “Well, if you can pay you have access to these amazing services. If you can’t you don’t.” I think the second thing which is interesting is to think about the whole ecosystem of media and journalism that rests on programmatic advertising. If you think that those things are key to… I don’t know, things like democracy, a downturn like this could really impact how many people are employed, what kind of media is being produced, that entire universe of things. And then I think the final one which is very top of mind for me is… As was mentioned earlier, I worked at Google and I worked with a lot of people who are part of their AI lab.
0:25:16.7 TH: And that’s like the… That’s cutting edge research, but it’s also a huge loss leader for the company. That it is essentially something that’s also subsidised by ads. And so I do think like a downturn would also even impact things like… That we don’t expect like scientific progress in the field of AI. Which is something that seems like very unrelated, but it has become subsidised by this ecosystem. And so… Yeah, I think it touches a lot of parts of the economy that we don’t normally think about.
0:25:44.8 TW: So… Yes… This is the… Yeah, let’s get 2021 rolling with real pick me ups here.
0:25:53.6 TW: But it does going back, the one thing we didn’t mention with the financial crisis was that you even had consumers, the borrowers were able to get access… People were able to buy houses because they could even get access to credit that they didn’t understand. I wonder if that’s the advertiser analog, but…
0:26:13.3 TH: Yeah, I really… One of the things I worry about is also the psychological price of things. So the example that I always give is like, because Amazon created this super controlled ecosystem called Kindle, we don’t think twice about spending $15 on what’s effectively like a text document. But for music, it has this history where there was massive file sharing, so our psychological price of music is zero dollars and zero cents. And I do think that one of the impacts of a shock like this is, suddenly all these consumers that were trained that certain services were zero dollars, suddenly in face increased prices to go after those services. And so there’s questions about, what becomes sustainable in a market like that as everybody needs to re-adjust to a world where, “Oh, suddenly this thing that I’ve lived for a decade, that has been free, is something something that I have to pay for on a monthly basis.” Which I think would be a really interesting outcome of the bubble bursting.
0:27:11.3 TW: So you lay out a bit of a plan of how to gently burst the bubble, being at an agency consultancy, to me, it seems like the near term for me to talk to my clients is, measure this stuff, do it in our… We just had an episode about randomised controlled trials and actually basically value your advertising a little bit better. You take a much larger view of how could we have a controlled bursting that doesn’t drive increased income inequality, doesn’t put the already anemic journalism industry completely under in a never recoverable way. Are you able to do summarise how you envision that potentially working?
0:28:04.1 TH: Yeah, sure. So the main thing that I’m really interested in is, in some ways, I think it’s like the macro version of, Tim, what you’re talking about on the agency level. ‘Cause on the agency level, you’re basically saying to clients, “Look, maybe you should measure this better. You’re gonna learn somethings that are maybe a little bit uncomfortable, but hey, we should probably be having those conversations sooner rather than later.” And in some ways, my proposals like to do that, but at a market wide scale, because right now, basically people don’t face any real consequences for distributing garbage information essentially into the marketplace. And so the dream is, again, this is like a super wonky thing, but the idea is like the government…
0:28:43.6 TH: You could imagine a world with the government says, “Look, there’s certain data standards for the kinds of ad inventory you’re gonna put out into this market, and if it turns out that those metrics are false, you will actually face liability for them.” And what’s interesting is that the idea here is not to put the government in a world where it says this ad is good and this ad is bad, which is I think it’s very difficult to do. But more in a world where it’s trying to at least increase the credibility of the information that you get through these markets. And I do think that that is a powerful way forward, just because I think part of the reason we have a bubble is because the information is so cloudy, and also because obviously the duopoly has been very reticent about giving up information.
0:29:22.5 TH: So all of these things I think quite combine to make a market that’s deeply unhealthy and so I think one thing that can be done certainly is regulation that helps shine a light on what’s going on.
0:29:31.1 MK: But I feel like this is that space where regulation is always so behind. Even if you look at data privacy. Yes, Europe was very forward-leaning, but I feel like generally regulation is so behind because they don’t understand the problem. If that’s the option, I don’t know, I still feel really bleak. [chuckle]
0:29:53.6 TH: Yeah, I think it’s a totally fair point. Maybe you could try… I’m curious about, Moe, what you think about this, is like one of the things I’ve been thinking a lot about is whether or not privacy laws end up being like the detonator for the bubble. And the reason being is basically like all of the advertisers are like, “Don’t take my data away! If you take my data away ads will just not work anymore.” And I almost imagine a world where there’s this almost perverse outcome where like GDPR really comes into force, CCPA really comes into force and actually turns out that ads are okay, like that it turns out that all this data we’ve collected is pretty much the same as really dumb contextual advertising from decades ago.
0:30:31.9 TH: And I think that sort of thing would really shine a light on maybe the fact that all this data actually is not what we thought it was. And that really could re-adjust the market in some ways, but what’s interesting is the regulators wouldn’t be intentionally doing that, they would sort of be like elephant in a glass shop sort of causing this to occur.
0:30:50.5 TW: Well, and it stood out to me that you didn’t delve into privacy regulations as a mechanism in the book necessarily for doing that, so I’m really glad to hear you bring it up ’cause that sort of stood out to me is sort of like, “Well, that might have a part to play.” Because again, like who is owning that data? That’s really what GDPR and CCPA are really getting it, which is, “Okay, well, Google has all this data on you, and Facebook has all this data on you,” and it sort of like, “Well, to what extent are they allowed to?” And oh, by the way, is even as Apple does what they’re doing with sort of browser level privacy to combat that, they’re already… Like Facebook is already working behind the scenes to basically create a server-to-server API call, where they’re actively encouraging brands who advertise with them to pass email addresses and phone numbers as part of the data that they send through.
0:31:45.8 TW: And it’s like, “Whoa, whoa, whoa, you want what?” And then in addition to that, you now have potentially evidence… In December, there was the lawsuit that was filed by the State of Texas against Google, where there seems to be even potential collusion between Facebook and Google, if this is actually true, I guess alleged collusion on price fixing with what’s called header bidding, which is basically… Not to get into too many weeds, but you allude to this, which is publishers who basically say, “Who can give me the best price for this advertising inventory that I can sell?” And in the header of the page, as it loads, there’s an auction going on and Google and Facebook basically agreed to basically split the difference and not drive the price up on each other.
0:32:38.1 TH: Right, yeah, and I think it’s really interesting seeing regulators wrangle with some of these problems because… The ad ecosystem is very wonky, and I think it’s exactly the thing that Moe is talking about, which is it’s a case where regulators are trying to get up with everything that’s been happening in this ecosystem, and I think they’re doing pretty good so far. I think the complaints that were launched are pretty informed, and certainly the collusion case, if proven is like… It’s a really big deal. I think it’s also really interesting, seeing… I mentioned this before the recording started, but the FTC basically put out an informational request to a bunch of social media companies, and I thought it was one of the most interesting things is that the FTC is ostensibly a privacy regulator, it’s really interested in what data these companies have on consumers. But hidden in that questionnaire was actually a long set of questions about all the inaccurate information they have.
0:33:32.6 TH: And so I do think that this relationship is sort of between privacy and the effectiveness of ads are tied in these pretty complex ways and I think it’s gonna lead to some really interesting conversations because traditionally it’s been… They have all the data, and that’s why we should be worried. I guess in some ways, my book argues like they have all the data, but maybe we should be worried not necessarily because it gives them a unique persuasive power, but because of all the other problems that it creates in the marketplace.
0:33:58.8 TW: Which is, you take the social dilemma and how so many of the social media companies stumbled in or followed the lead of Facebook that had no idea they were gonna become all of their revenue basically being advertising. And so clearly no plan on the route they went, but then you have so many of these shadow industries… I try to explain to people, “Every time you click through on the 120 celebrity couples who’ve been together forever,” you’d never believe in just that user experience that is so painful, but it’s because that it is that ecosystem, it is all those transactions, just impressions being bought by intermediaries to resell impressions, I’m like, that to me seems like the clearest evidence that this is clearly broken.
0:34:53.5 TW: You have junk content that is actually paying for advertising so that it can get people where it can have more impressions that it can resell. And there’s so much money slashing around in that area, that when it does come to regulation, policy changes, it does feel like a pretty tough row to hoe when there are gonna be a lot of dollars pushing back saying, “Yeah, yeah, this is really a problem, but go fix another part of the problem, don’t come after… The DSP technology is totally fine, go chase the bots or the… ” That just feels like the struggle is that it’s gonna be whack-a-mole with point solutions that will have people seeing dollar signs to work around whatever point solutions get regulated or put forth.
0:35:51.9 TH: Yeah, and I think ultimately, this may end becoming a conversation around sort of activist strategy almost, is like a way to think about it, which is like, what are the evidence which will cause the biggest advertisers to back out of this ecosystem? And I think that’s ultimately gonna be the thing that will really reshape the market, just because… I agree, we have very blunt instruments when it comes to regulation, we have very blunt instruments when it comes to public opinion, but the minute Procter & Gamble is pulling millions of dollars out of the ecosystem, I think it does start to create incentives that are moving in a better direction. I don’t think it’s a silver bullet by any measure.
0:36:27.9 TW: Well, but that… You mentioned the P & G, and that was a multi… “We’re pulling our dollars out” and then everybody was like, “This is big, and it’s bold, and it’s right,” and a couple of other large companies have done it, but it seems… Why has that not really gotten traction? Is it because there’s so many forces saying, “We don’t wanna look too closely and find out that’s right.” We can just say, “Oh, that’s P & G, they’re massive. We’re only a 100 million company, so this doesn’t apply,” ’cause that was like two or three years ago…
0:37:00.7 TH: True. Yeah. Good point.
0:37:02.6 TW: That P & G started saying that, right?
0:37:04.7 TH: Yeah, and I think that that’s right, yeah, exactly. And it may just be, that’s one of the most interesting things about financial bubbles, if you study the history of them is the thing that triggers the stampede out of a market is not always the biggest advertiser moving out. And yeah, I think, I don’t know, Tim, I think it suggests to me at least that the incentives are pretty strong here to keep the ball rolling, because at the point at which you have these really big advertisers moving out, it doesn’t create that cascade. I guess we are left wondering like what does.
0:37:39.9 MK: So I have a very odd question, and I’m hoping no one has asked you this in any other form. But it’s something I couldn’t help thinking about as I was reading your book where you really lay the foundation when you start talking about… Like when Google started, they thought the percentage of revenue coming from advertising, I think you said it was 17%, but don’t quote me.
0:38:00.6 TH: 17%, exactly.
0:38:02.4 MK: Now obviously it’s like their main revenue stream and no one anticipated that. You’ve obviously spent so much time and energy, and I think you write about this problem so incredibly well, have you seen those early indicators that we’re somehow setting ourselves up for failure in another new space that perhaps we’re only just starting to see those signs of? And I think, yeah, I don’t know if that’s something that you’ve given thought to, but I couldn’t help trying to think of like, “Oh, I wonder if there are other situations where if we look closely enough, we could see that they might become problematic, like advertising have?”
0:38:43.4 TH: Sure. I mean, I think that… I think the answer to that question is, in some ways, advertising has become the business model hammer for every nail, is the way I’d put it. Essentially, programmatic advertising has been just this pure capitalist rocket fuel, that it’s just if you can turn businesses into a programmatic advertising business, you’re doing great. And I do think that’s the case, as you see people being like, “Okay, search is over, display is dominated.” There’s just this increased press to basically say, “Can we get programmatic advertising into other channels, expand it into as many corners as possible?” And I guess part of my worry is that it sort of spreads the infection, if you will, that it actually creates these pathologies in entirely new media channels that may have not had these problems before. I mean, one example, which I don’t think is large in terms of market size, but it’s near and dear to my heart is the world of podcasting, which is that podcasts have run on a very old school ad model, is there a push right now to make them programmatic, for sure. And the part of the problem is that the idea is like, “Okay, but do you also want all of the market warping incentives that are created by introducing that as your life stream for your economy?” And so I do think that those are the new frontiers I’m looking at, but it’s largely extensions of the empire of programmatic advertising.
0:40:10.0 MH: I guess for us, the question is about how much money do you think we could make if we kind of sold our souls… no I’m just kidding [laughter]
0:40:14.1 TH: Squarespace, build it beautiful. Yeah, right.
0:40:22.4 TW: ZipRecruiter, this is brought to you by, snack in a box, subscription service of the month, whatever it is.
0:40:29.6 TH: It’s not in the book, but it’s a thesis I’ve been playing around with a little bit, which is, is there an optimal amount of data you want in a marketplace before it becomes terrible, which is like, okay, imagine a world where basically you have no information about your readers. Well, you’re just flying by the seat of your pants and maybe you produce lots of stuff they don’t actually like. And then you have a little bit more data and you’re like, “Okay, well, yeah, I can calibrate stuff that they’ll like.” And that’s actually a good place to be, and maybe we’re there for podcasts right now, but then eventually you have the dashboard up on your office and you’re like, “Does this get as many click views as… Or view count as this other post?” And that introduces all sorts of terrible incentives to be like, yeah, what maximises that metric? And so there’s an interesting progression, which is like, there’s a point at which the data becomes so granular that it actually produces all of these negative effects on the quality of the work, and so… I don’t know, it’s just like a vague armchair theory at the moment.
0:41:25.3 TW: I like it. Wow, I wish my vague armchair theories were… I wish my well fully thought out, well-formed theories were 20% of that.
0:41:33.7 MH: But you can actually see large corporate interest jockeying into position for the future you’re describing, Tim, on podcasting, like Spotify with all the moves they’ve made in buying a number of companies, and then they made that massive investment to… I don’t know if it’s acquire, but basically acquire the Joe Rogan podcast which is, I guess, a really popular podcast. And so it’s coming, ’cause everybody is sort of like, “We presume the algorithm, we’ll put our weight on the stone and once it starts rolling, we’ll reap the benefits.” I ran across an article, I don’t… I think it might have been on Bloomberg or somewhere, but I can’t remember, but it basically was something around Black Rock as a private equity company, was exploring as they acquire companies in private equity packaging and selling their data as another asset to basically recoup the investment they’re making in the company. And I was just like, okay, so that’s another horizon we haven’t really gotten into as a society as sort of like, yeah, we’re just gonna have corporate raiders now coming in and being like, “Ah, here’s some data. Who wants it? Who wants it?”
0:42:51.4 TH: Yeah, for sure. And I think it’s a disturbing thing. I think this is a story that happened when RadioShack went out of business in the US, which is basically that, it turned out that one of the big assets they sold in the bankruptcy was their enormous data set on American consumers. Yeah, and I was just talking to a friend the other day about, there’s also a very vibrant data ecosystem between the ad economy, political advertisers and churches in providing data. So I do think that the ad ecosystem creates all these interesting links between organisations that otherwise wouldn’t have reasons to talk, just because it suddenly becomes profitable to mine certain types of data that otherwise in the past, wouldn’t be considered an asset at all.
0:43:33.7 MH: Yeah, it’s definitely a time in which advertising is one piece of this, but it’s spreading everywhere, which is probably why we need a law about it.
0:43:48.7 TW: I will say this, I’ve often maintained it, I think your book was useful to me in this, which was I’ve often said of the analytics space, a lot of shenanigans happen because of the lack of relative maturity of the space, we haven’t been around… Digital hasn’t been around for all that long. And so it’s sort of like, yeah, we’re all still learning, but the juxtaposition of the mortgage crisis to this, I was like, “Well, the mortgage space wasn’t immature”, it’s been going on for a long, long time, and yet they still manage to work themselves into a pretty intractable situation, and so it’s really given me some pause on the way that I’ve been mentally framing it up, and so for that reason, I found it really useful and helpful to read your book.
0:44:36.5 TH: Cool, yeah, I really appreciate it. Yeah, I think that’s right. Again, I think there’s an interesting question about what programmatic advertising should be considered as a subset of, ’cause if you consider it as the latest generation in advertising, then it almost is similar to the mortgage situation where mortgages existed for a really long time, and then we were like, “Hey, actually, we could collateralise this debt in spite and sell it as an asset.” And in some ways, I think that’s what’s happened with advertising, which is, we’ve bought and sold attention for many, many years, decades, but only now have we been like, “Oh, you know what we could do with attention, we could slice it up and sell it at rapid speed in this particular way.” And I think that’s producing all these effects, it hasn’t been just evolution, it’s actually been a revolution and sort of the impact of the technology.
0:45:22.9 MK: I can’t help but wonder, this is what I… I just keep going down all these crazy paths, [chuckle] I don’t know why. What made marketers accept, I guess, the black box of programmatic advertising in the first place? And it does seem like the shift in the industry and the way things are going with privacy we’re moving back to that state again. Yeah, I laid here being like, “Why would they just accept that and why wouldn’t they challenge it, and how can we get them to challenge it more?” It’s the thought that’s running through my head.
0:46:02.0 TH: Yeah, I think it’s a good question. And I think in some ways, this is an interesting critique that I’ve been thinking about how to wrestle with, which is the book came out and some ad people said, “Well, look people spend a lot of money on it. They wouldn’t spend money on it if it didn’t work. How do you explain that? How do you explain that?” And again, I know there’s a little bit of a fallacy there, but I do think that we have to address the question of like, why do people adopt this stuff if it is so faulty? And I do think that one of them is… One of them is, I think the myths that we tell about data, which is that it’s very hard if you’re sitting in a room and being like, “We’re either gonna buy this old school media channel called a newspaper and it doesn’t have a lot of data, or we can buy this big fancy one that everybody’s talking about that has all the data.” And almost regardless of whether or not the data is faulty or not, the fact that they have the data is pretty powerful. And so I do think that that is one thing for sure. There’s a thesis again that I really wanna write about that didn’t make its way into the book, which is sort of a funny…
0:47:01.7 TH: Again, a totally speculative theory that I’d love to research some more, which is that in the early days of banner advertising, the click-through rate was 44% and now, it’s more like 0.01%. So you basically had in 30 years, 100-fold decrease in the effectiveness of this ad channel. One theory about it though, it’s like maybe the reason we pay attention to ads is just that they’re novel, they’re new, that’s why we look at them. So you imagine the first billboard that ever was put up where people like driving by and they’re like, “Wow, that guy on that billboard told me to smoke cigarettes. I should go buy some cigarettes.” And then over time, we just got used to it as a society and the effectiveness declined, and so there’s always that theory which is that early on, online advertising looked really, really effective, but it had nothing to do with the data. It had everything to do with the fact that it was new, and so we just happened to pay more attention to it. And so now historically, we’re now like banners and search are now becoming the next billboards. And if you buy that, there’s almost like no way to recover, but it certainly explains why early on, advertisers would be willing to go to this new channel because in the early days, it is actually in fact that much better. But again, it has nothing to do with data or technology.
0:48:08.0 TW: But it was a dual… It was not only was it was novelty, so therefore had more attention, but the people who were doing it were the ones that were like direct response digital advertising. ‘Cause I had commented on… I’d made a… I tweeted or posted somewhere about your book and somebody did chime in and they’re like, “Well, every time we run advertising, we get results.” And I’m like, “Yeah, if you’re a retailer and you’re selling shoes and you’re buying paid search for those shoes, sure” and you could quantify from a last click. But to me, part of what happened is that… Well, that became part of the mythology is that everybody can do it. Oh, you’re selling pet food? You’re selling boring, boring old pet food? You gotta be in digital, too. So to me, it’s this marketers are somewhat kind of lemmings, they had been beaten up for years about not being able to quantify the impact of their investments. Somebody came along, the pied piper of ad tech came along with a story and they were like, “We will happily follow that out, and now we’ve got numbers that we can point to. And when you combine that… “
0:49:15.4 TW: I loved how you made the point that it’s a steadily decreasing effectiveness, but nobody’s really watching that sort of trend of effectiveness. They’re saying, “Well, we didn’t get quite the click-through rate that we got last year, but you know what, X or Y, it’s noise in the data.” And they don’t necessarily look and they don’t have a constant way of measuring, the things are changing so much, new channels to advertise on, they can’t even see a truly long-term trend and recognise that they’re spending more to chase lest, and they’re trying to do it for everything, not just a direct response, direct sell where it would actually work.
0:49:52.3 TH: Yeah. And I think this is connected to the opacity point. One of the ironies is that we have a lot more data about like a given ad, but we have a lot… It’s basically no man’s land in terms of trying to figure out how the market as a whole is evolving, and so we’re like… We have really, really deep transparency, but in some ways, it’s extremely narrow. And that I think that is a big problem to try and get to the bottom of what’s actually going on.
0:50:16.9 MH: So much more to data, but we know less.
0:50:18.9 TH: Yeah, ironically.
0:50:22.8 MH: And as marketers, we’ll end up chasing channels. So this week it’s TikTok, and next week it’s whatever is gonna be up next.
0:50:31.7 TH: Yeah. And I think we do also… This is very natural, but I do think we suffer from a bias of like… We hear about the success stories, but we don’t have a really good measure of all of the failures.
0:50:43.5 MH: Yep, 100%.
0:50:44.3 TH: And so everybody can point to an anecdote where ads worked, and I agree, there are situations where ads work. The question is, is that the norm or is that the outlier?
0:50:52.8 MH: Right, yeah. This is such a great topic and we do have to start to wrap up, but I think all three of us, Tim, really got a lot reading the book. And the book is “The Subprime Attention Crisis,” and I think you’d get in the official podcast, recommend here from the team.
0:51:11.6 TH: Thank you, thank you.
0:51:11.6 MH: So really nice job. Thank you so much for writing it. Thanks for coming on the show.
0:51:17.1 TW: I wanna throw in on that just the book is like… It’s like 140 pages of content. It’s got a bunch of… It is both super well-researched, it is not a lengthy read like it sounds like, “Oh my God, is this gonna be some super dense… ” And I’m still blown away with how you managed to pack the pieces together and have… I don’t know how… You seem like one of the people who maybe just doesn’t ever sleep, but for… I really do think people should… Like this is not a book that takes a long time to read, and it is so informative.
0:51:49.8 MK: It’s really informative.
0:51:51.7 TH: Thank you.
0:51:52.9 MH: You mentioned Flash Boys early in the book by Michael Lewis, and as I was reading it, I was like, “Yeah, we need a character through this thing.” [chuckle] So I was like, you should… Maybe Version 2 is like the Michael Lewis treatment.
0:52:05.4 TH: A Hollywood adaptation of this book.
0:52:11.5 MH: Yeah, that’s right.
0:52:13.7 MH: I think Seth Rogan’ll be ready for a serious role and like… Anyways. Okay, we do have to start to wrap up. One thing we love to do on the show is called the Last Call. We just go around the horn, we share something of interest that we think might be interesting to our listeners. Tim Hwang, you’re our guest, do you have a last call you’d like to share?
0:52:33.0 TH: Sure, I’ll plug a weird side project that I’ve been working on for the last year or so. I don’t know how many of the listeners are based in the US, but I run a small project called the Trade Journal Cooperative. One of the things I’m really a big fan of is obscure trade journals, and this is a quarterly subscription service that sends you an obscure trade journal once a quarter. So previous issues have included Elevator World, which is the magazine of the elevator and escalator industry. They actually call themselves the vertical transportation industry, which is really great.
0:53:04.5 TH: And last year we basically… Yeah, and last year we’ve covered what, magicians, the most recent one was private investigators. So if reading those magazines gets you interested, I’ll just plug that project.
0:53:16.9 TW: It’s like you’re the Gary Vaynerchuk of trade journals, so instead of getting your wine subscription, you’re getting obscure trade journals.
0:53:25.6 TH: Yeah, it’s something like that.
0:53:26.6 TW: Do you make a video each time one goes out where you talk through the…
0:53:30.9 TH: Actually no.
0:53:31.5 TW: The fascinating…
0:53:32.2 TH: Yeah, I really should though, I really should.
0:53:36.8 MH: That’s awesome, thank you for sharing that. And that’s… Definitely gonna take a look at that. So, Moe, what about you? You have a last call you wanna share?
0:53:45.4 MK: I have a request, and then I have an odd last call, so the request is one of the things that I did reading this book is I tried to remember the last five ads I saw, and I didn’t have a freaking clue, and then even as I became aware of trying to… So first I just did it of like, “What do I think the last five ads were?” I had no idea, and then even when I tried to flip my awareness to try and recall the next five ads you see, I still really struggled. So I just feel like it’s a really interesting exercise, particularly for data analysts to try and have a play around with and be like, “Oh, here’s this thing that we spend all of our lives trying to measure, and yet we can’t recall them ourselves.”
0:54:28.2 MK: The other thing, so I came across something really bizarre, but I really enjoyed reading about it, so I don’t even remember how I stumbled on it, but it’s called Auction Theory, and there’s two Stanford economists, Robert Wilson and Paul Milgram. I think they actually have won the Nobel Prize now for their work… It was a BBC podcast, more or less behind the stats Auction Theory that I somehow ended up listening to randomly, and I just found it really fascinating, and the reason I found it so fascinating is because in Australia, all properties pretty much go to auction. You don’t normally make someone an offer, it’s like they auction everything to try and get the highest prices. And I just found that this podcast was really fascinating how it talked about the whole idea of like you have a mass auction and how everyone actually ends up better off. Anyway, if auctions are something you’re interested in…
0:55:20.2 TW: Wasn’t part of theirs… Part of the winner’s curse… I don’t know if they were the winner’s curse, but isn’t that part of auction theory is like when governments that have adopted auctions for bidding for contracts that they’re like, “We award it to the second lowest price, not the lowest, because if you go to the extreme, the person who actually won gets screwed”?
0:55:39.6 MK: They talk about the example of the telecommunications bandwidth, and how you auction that off, and everyone actually ends up getting what they need. But I don’t remember the winners curse so maybe I’ll have to do more reading, but anyway…
0:55:51.2 TW: I don’t think the winner’s curse was auction… Well, it is part of auction… I don’t know…
0:55:55.0 MK: Anyway, I need to clearly do more reading, but it’s a fascinating topic.
0:55:58.9 MH: Tim, the quintessential analyst Tim is who we’re talking.
0:56:02.6 TW: Jeez…
0:56:02.9 TW: Well, so, I almost thought we were gonna wind up hitting my last call while we were talking, so, one, I completely missed the Morning Consult, did a whole special report last year on podcasts, and it is like a series of different… And a lot of it went into different… The Morning Consult, they’re always collecting data and doing surveys, so it has a whole bunch of different articles where they dug into some of it being the payment model for podcasts, some of it being advertising for podcasts, so it’s just a very interesting… I haven’t worked through all of the articles in it, but it’s a very well-curated set of articles about different aspects of podcasting. But I did read one talking about the effectiveness of ads, and it was talking about how much higher…
0:56:54.7 TW: The example used was from Crooked Media which, three podcasts that I listen to every week are out of Crooked Media, and those are hosts that actually have a whole banter unscripted advertising, like the multi-touch moment that we used to have, and it leads off with the examples of Jon Lovett, Jon Favreau and Tommy Vietor on that. So my second part of my last call is to actually recommend Just Egg, which was one of the mungbean-based egg substitute that they plugged, and I did actually go out and buy some, and I was delighted. I do eat… We eat a lot of eggs, and I was like, “Okay, they’re just bantering about this like these really are a good egg substitute,” and they really were. So I thought I’d help out with Pod Save America’s podcast advertising impact. I don’t have a website with a special code to go to, so their measurement may be screwed, but…
0:57:53.4 MH: I was gonna say, “Good luck on the attribution with that one.”
0:57:55.8 TW: Yeah. [chuckle] So what’s yours Michael?
0:58:00.9 MH: Well okay, so Moe to chime in on your observation, literally the day before we recorded this and we record shows in advance, Twitter went crazy with this speculation about a specific advertisement that was a bit cheeky, let’s say that a lot of people ran across and Vice even did an article on it like today, which is basically digging into why is everyone seeing this ad right now and where is it coming from? And so it was just like, how timely? So anyways, so there you go. Okay, so as I was researching for the show, I got interested in who published this book, and so I was looking at the publisher and I ran across it was a joint publishing effort between I guess an imprint of Macmillan and a group called Logic, and then I was Googling that to figure out what the heck that was. Turns out it’s a little magazine called Logic Mag where they’ve taken a stand of we’re gonna write about technology, but we’re gonna write about it really thoughtfully and really slowly and on paper. And so I subscribed and I’m really looking forward to getting some of their stuff and it looks pretty neat. So just following threads. So Logic Magazine, kind of nifty. I’m looking forward to it. They only publish three times a year.
0:59:31.7 MH: Okay, I’m sure you’ve been listening and you’ve been thinking, “Wow, this is a crazy topic. How do I find out more?” Well, here’s a cool thing. You could actually follow Tim Hwang on Twitter. He’s at Tim H-W-A-N-G, Tim Hwang on Twitter. So you can follow him there, and you can also reach out to us on the Measure Slack or on our LinkedIn group, and we’d love to hear your thoughts and questions about the show’s topic. As always, our show would not be possible without our inimitable producer, Josh Crowhurst, doing what he does behind the scenes to make these shows a reality. So thank you very much Josh. And lastly, Tim, thank you so much for taking the time to write the book, come on the show today. Been a really pleasure having you. Thank you so much.
1:00:20.8 TH: Yeah, this was great. Yeah, thanks for having me on the show.
1:00:23.1 MH: Awesome, and I know that no matter where you are in this opaque ecosystem that sometimes gets our goat, you know that both of my co-hosts and I all agree. You should keep analyzing.
1:00:39.5 Announcer: Thanks for listening, and don’t forget to join the conversation on Twitter or in the Measure Slack. We welcome your comments and questions. Visit us on the web at analyticshour.io or on Twitter @AnalyticsHour.
1:00:53.9 Charles Barkley: So smart guys want to fit in, so they made up a term called analytic. Analytics don’t work.
1:01:00.5 Thom Hammerschmidt: Analytics, oh my God. What the fuck does that even mean?
1:01:09.8 MK: I feel like this year, I haven’t seen anyone. It’s been very strange.
1:01:13.3 MH: Wow, yeah.
1:01:13.4 TW: Why’s that?
1:01:15.6 MK: Don’t even get me started. It’s been the week from hell.
1:01:20.1 TW: It’s Monday Moe.
1:01:23.5 MK: You have heard Christmas is cancelled in Sydney.
1:01:26.9 MH: What do you mean Christmas is cancelled?
1:01:29.4 MK: I didn’t think that was a legislated…
1:01:32.2 MH: Yeah, we’ve got Dr. Fauci here in the United States telling kids that Santa’s been vaccinated. It’s totally cool. I’m like, thanks, Dr. Fauci.
1:01:42.2 MK: And Tim, author Tim if I suddenly sound really exacerbated and I’m like, “God Tim” it’s definitely the other Tim I’m talking about. [laughter]
1:01:54.8 TH: I mean it’s fine if it’s to me too. It’s alright.
1:02:00.3 TW: Yeah.
1:02:00.6 MH: Tim Wilson is gonna ask you… You will get asked a four-part question at some point today.
1:02:07.9 TH: Great. Bring it on.
1:02:08.8 MH: So I’m sure you can handle it.
1:02:12.1 MK: Well bring it on if you can distinguish parts one from two, three, and four ’cause sometimes they mould together.
1:02:18.6 MH: Tim Wilson, you’re hilarious.
1:02:21.5 TW: That’s actually really…
1:02:28.8 MK: I think that’s the thing that’s gonna be really…
1:02:29.2 TW: Okay, I’d like to call out that this was actually, I was not the one talking either one of you. You talked over each other.
1:02:38.7 MH: I started talking first so I don’t know how that happened.
1:02:41.1 TW: Rock flag and collateralised digital advertising. Sorry, we do a It’s Always Sunny in Philadelphia little take off at the end that goes in the out takes and I botched that one.
1:02:58.2 MH: That’s okay, ’cause that’ll go in the out takes too.
Photo by Andriyko Podilnyk on Unsplash
Subscribe: Google Podcasts | RSS
This site uses Akismet to reduce spam. Learn how your comment data is processed.
https://media.blubrry.com/the_digital_analytics_power/traffic.libsyn.com/analyticshour/APH_-_Episode_220_-_Product_Management_for_Data_Products_and_Data_Platforms_with_Austin_Byrne.mp3Podcast: Download | EmbedSubscribe: Google Podcasts | RSSTweetShareShareEmail0 Shares
Subscribe: Google Podcasts | RSS