When it comes to advertising and media, Brian O’Kelley is something of an expert. He helped build some of the most powerful technology brands in adtech, from AppNexus to Right Media. But now O’Kelly is targeting the digital advertising supply chain, and in those giant server farms storing millions of digital ads he sees a carbon sinkhole.

O’Kelley doesn’t believe advertising should go away; it just needs to get smarter. Until Scope3, advertisers couldn’t quantify what the carbon footprint of those empty buys had left behind. 

Hear more from O’Kelley and Signal’s co-Editor-in-chief John Battelle in this lightly-edited conversation.


John Battelle
Welcome to another Signal Conversation. This month we’re welcoming Brian O’Kelley, who is the founder and CEO of Scope3, an interesting startup that I’m looking forward to him telling you all about. But Brian is well-known to those of you who might be in the advertising and marketing technology space. He co-founded AppNexus in 2007. He was the CEO and chairman through its multi billion dollar sale to AT&T a few years back. Prior to that, he was the CTO, the founding CTO, I believe, of Right Media, which was sold to Yahoo back again in 2007.

Brian has been responsible for a lot of what we understand to be the technology behind programmatic advertising, including creating the online advertising exchange, and a bunch of other related technologies like DSPs, and SSPs and data exchanges and real time bidding, and header bidding. I wish we could get into all of that this time. But really, I want to get into Scope3, which is a fascinating new company that Brian is bringing to market just now. He is a recently added member of the Board of LiveRamp where I am also on the board, the active member of Tech:NYC and lives in New York City and Brooklyn. Welcome to Signal, Brian.

Brian O’Kelley
John, thank you for having me.

I’m glad you’re here. So the mission that you have at Scope3 is, and I’m quoting here from it, I believe to decarbonize media, and advertising. What does that what does that mean?

We know that every supply chain in the world is contributing carbon emissions to the environment. We’ve got this climate change problem, and I think most people don’t realize how significant the contribution of advertising, if you think that whole supply chain from production of media and creative all the way through ad tech, ad selection, distribution, the moment the ad hits your eyeball, that whole supply chain is a significant source of carbon. Advertising funds, the internet. If you think about every server in every data center that AWS has, that Google has, Microsoft has Bing, ChatGPT, all of that, at some level is ad-funded. That’s this digital supply chain, and our passion and mission is to make that have a much smaller impact on the environment, while still providing all the wonderful things we love about the internet. That’s a fast and highly impactful change, if we can make it.

I don’t think it’s something that that people think about that much the carbon impact of all of the processing power that goes into delivering an ad. How did you come to Scope3 from your background, which is so deep in digital advertising?

After we sold at AppNexus to AT&T, after 15 years of of ad tech, I was ready for a change. I, with a friend, co-founded a company called Waybridge, which was a supply chain startup, where we were going to trace metals like copper and aluminum, from the mine and smelter and this whole supply chain, from Chile, on a boat to Florida, given a train and a truck, and finally to the factory. A totally different supply chain with the idea that if we could trace that process, we could find ways to optimize it financially, but also for sustainability purposes. What would green carbon look like if you could go source to sink to the wall in your house?

One of the realizations I had was that most of the ways to make a sustainability impact were very hard. You have to go get green hydrogen pipelines, or you had to scrape barnacles off the bottom of a ship. For all my technological expertise that’s not something that I knew how to do. But as I started to get deep into this, I took an MIT Supply Chain Management year-long course online during COVID. I realized that the theory was right, but that the industry was wrong. That the digital world, while it seems like it’s less impactful, actually has a huge footprint that no one could see, but that the changes were largely algorithmic or technological in ways that we can impact very quickly. If you think about the idea that a third of ads on websites are never seen, they’re not even viewable, but they still get auctioned to thousands of servers, they still get served. And advertisers still pay for it.

If we can just fix that one thing, you’re talking about tens of thousands of servers that just disappear from clouds. This is a pretty massive impact, and also, brands don’t have to pay for them. This is great. This is an economics problem. This is a marketplace problem. This is an environmental problem. My entrepreneurial energy started to flow. This felt a lot like the circumstances that led to the creation of the ad exchange, which was a structural problem that was costing people money, but then also had a really interesting and fast technological solution. I just pattern matched. I wanted to make an impact in the world, and I know a lot about this industry, and there was an economically viable solution. It just needed that spark of someone saying, “What if we put these pieces together and try to change the world?”

What’s your pitch to a decision maker, somebody who spends hundreds of millions of dollars in digital advertising a year? What’s the thing that you say to them that wakes them up and they say, “Oh my God, this could really be big for me?” What’s that promise that Scope3 can make that gets people excited?

Last October, I was at the World Federation of Advertisers executive committee meeting, talking to some of the biggest brands, CMOs of huge brands, and the heads of trade associations. I was really nervous, even though I’ve started and sold companies. This is the room where it happens. And I said, “Listen, we know, programmatic is deeply troubled. It is complicated, it is inefficient, it’s non transparent. There’s privacy issues, there’s national security issues. But it’s also a waste of your money to buy all of these crappy ad placements and buy non-viewable ads. You’ve known this for a decade, and haven’t fixed it. But if I told you right now that there’s an environmental problem, you’re actually hurting the environment, by doing this. All of your companies have made sustainability pledges, right? And that by fixing that problem, you’re also going to reduce your impact on the environment by a huge amount tens of thousands, hundreds of thousands of metric tons of carbon a year, could I motivate you to make some changes to how you operate, that will save you money, that will make your marketing work better, but will also have a huge environmental impact?” And they all said, “Absolutely.” Which is funny, right? Because it’s the same problem that they’ve been trying to figure out. But this was a motivation. It’s not just about marketing spend. This is about investors, putting ESG pressure on companies. It’s about employees demanding change. It’s about the consumer story for their products, where they’re saying, “Hey, we’re doing the right things for the environment.” It’s just actually a much more important mission for everybody to do their part for the environment. If it’s also good for business, it’s kind of a no brainer.

How does it work Exactly? What are you doing that changes the supply chain of digital advertising such that this kind of efficiency and impact is affected?

The first step is to measure you have to know how the money you’re spending is causing an environmental impact. One thing that is hard for companies is that supply chains are very long. There’s lots of hoops, but they’re also geographically dispersed. So even an internet, and there’s millions of different internet publishers around the world, and most suppliers don’t know their own carbon footprint. This is every supply chain.

So our unique approach was we figured out how to model the emissions of the entire supply chain because of a lot of publicly available data that we could pull in, thanks to initiatives by organizations like the IAB Tech Lab, which has been saying, every company should publish its ad tech stack, in a little file called ads.txt. We can pull that in. We can use a bunch of information that’s been contributed, or even publicly-shared by many of the thought leaders, who were, you know, companies like Axel Springer who have published a sustainability report or Criteo, as an ad tech company, or of course, Microsoft and Google and  these companies, we can actually put all of this into a big model and say, “Well, if the New York Times doesn’t have a sustainability report, which it doesn’t yet it’s working on it, we have enough other news companies that we can say, based on how many employees, the kind of properties they have, the public information we have, we can come up with a pretty accurate guess.” That’s dramatically better, actually, than if you ask them because they don’t know how to measure it. We can also do apples to apples comparisons, because we are building the model, because there’s no like regulatory requirement. So maybe News Corp doesn’t publish their travel emissions, but Meredith does, so we can figure that out and make changes.

So this is a long, complex way of saying, it’s not just about measuring your supply chain, it’s actually about modeling your supply chain. So that when you’re looking at all of these different things, how do you compare CTV or streaming video to a mobile game? Well, we had to build a model that can do that. How do you think about digital out of home? If you’re in your Uber, and there’s an ad, how do you compare that? The second thing is to make sure that brands understand, we want you to keep spending money, we want you to spend it effectively, even if something is high carbon, if it’s really effective, keep doing it. The intention is not to just get rid of carbon, willy nilly. But if you have places where your spending isn’t as effective, and my non-viewable ads example is an obvious one. We know, that’s not changing anyone’s minds to make you buy Tide detergent, well, maybe you should stop doing that. That’s the second phase.

Now we’ve measured we see where our carbon is, and we overlay that with the effectiveness of our spend, then we go down and say, to our agencies, to our publishers, to our ad tech partners, “I want this to go down.”  That reduction is something that we can inform because we now know where those hotspots are, we know the opportunities to do reduction. Then we can say, “I’m not going to buy made for advertising sites, I’m not going to buy non viewable ads, I’m not going to buy from publishers that have more than 200 ad tech partners, because it doesn’t really make any sense.” That’s the very specific reduction that in many cases is, as an example, by cutting the number of domains you buy by about 7%, you can reduce your carbon footprint from programmatic by about a third. That’s a shocking change.

If you do that, what happens is those publishers that got turned off, those 7% are like, “Wait, I don’t like being turned off, maybe I should go do something.” We’ve seen, as systematically, those companies reducing their emissions by in some cases 75%. So they just proactively say, “Well, I don’t need 200 or 2000 ad tech partners, I’m going to call the herd.” That impacts not just that one advertiser, that impacts the entire industry. What we’re seeing is the entire footprint of the industry starting to trend downward very quickly, based on these decisions by advertisers. That’s kind of exciting. I think other industries will follow the same footprint, which is once you see that it’s good for business to be green. Then you get this virtuous cycle where the greener you are, the more money you make. And especially in the B-to-B context.

You mentioned P&G, they’re one of the largest advertisers in the world. When they start tipping the scales on something, the whole market changes. That’s what’s been so exciting for me personally, in this mission is seeing the world change quickly toward a more environmentally-friendly way of operating is, I think probably the most proud I’ve been of anything in my career, is doing something that makes the world better. It’s good for business. That’s what I want to do with my life and my energy. So many people I get to work with every day are passionate about this too. It’s not just a Brian O’Kelley project. This is an industry movement that I get to be a little part of.

It strikes me that efficiency and viewability, intelligently applying your ad spent, these are not new concepts in advertising or certainly in digital advertising. Is it the ability to is to look at the problem through the lens of a new point of view or a new framework, which is the carbon impact that that’s sort of spurring this along? Because no one wants to pay for non-viewable ads, for example. Is it just that you’re interested in sort of almost de-complicating what has become an extremely big ball of yarn in programmatic?

I can only take the blame for inventing acronyms and threads of yarn in that ball. But I think you’re right. It’s having taken three years out of the industry, and walking back in. A lot of those things we all took for granted, like the status quo, it has to be this way. It’s not true. If you actually pick it that ball a bit and say, “Well, why are people paying for non viewable impressions?” There’s a procurement aspect of that. Brands have said, “I want the lowest CPMs I can get, cost per 1000.” Is it cost per 1000 views? It’s cost per 1000 impressions, right? So the agency is like, well, “I want to make my client happy. I’m gonna blend some cheap, non-viewable ads with some expensive, viewable ads, so that the procurement team says, ‘Good job, you cut the cost of advertising by 10%.’” But that’s not quality.

But that’s where you have this tension. You ask the procurement people, “Why are you doing this?” “Well, there’s all these shenanigans that we know that the supply chain does. We’re trying to squeeze the supply chain.” And I’m like, “Well, this is a balloon. You’re squeezing here. But it’s popping out here.” This is the systemic idea. The systemic reduction is pointing out to everybody at the same time, “Hey, this doesn’t work anymore. As an industry, we have to stop doing wasteful things.” I’ve been fortunate to have brands, agencies, trade associations, ad tech companies, publishers, platforms, cloud companies, everyone’s sort of on the same page of, “This year, we need to change this. And everyone’s kind of doing it together.” This is where being an ad tech company is difficult.  Because as the CEO of AppNexus, I was trying to go public, and I had to hit certain numbers, and taking a hit on this or distracting my team for this would be hard to justify. But I don’t have any status quo to protect. All I have is a mission to achieve. That means you can take a holistic approach, that systemic holistic process means no one’s really hurt, because everyone’s changing things at the same time.

Let me ask you a final question. If you’re successful, Scope3 is successful by whatever standards you judge success to be, what changes in five years? What does this industry look like in five years that it doesn’t look like now?

The biggest tragedy of programmatic in my mind is that it commoditized content. That was the sort of foundational idea that we could follow people from site to site and create this sort of open web where anybody could get access to demand without a huge sales team. That was great. Except that you saw quality publishers like these newspapers start to decline dramatically as a share of ad spend, because it was about the person, not the content. Once that started, it was really hard to reverse. A lot of that was because the business model for content production, was if you write beautiful, long form articles, that don’t get that many views, but get you a Pulitzer, that doesn’t really drive a whole lot of ad revenue in that model. Now we have AI generating articles, and we have a whole new way of aggregating audiences, and we have the world changing. So I think it’s a great moment actually to reimagine advertising.

So in five years, what I’d like to see is an advertising model that is much more aligned to quality and is less aligned to people so that we’re actually paying for quality interactions with people on quality content. Now this sounds naive to anyone watching the last decade of programmatic and social. But I also have had amazing conversations with publishers. I mentioned Axel Springer, I’ve been talking to Dotdash Meredith. We work with Fox, we work with AccuWeather. We work with these really virtuous people who are trying to do quality content. And I think the agencies have responded to this too. So five years from now, if we drain the swamp, if we get rid of all of these problematic ways that programmatic is hurting journalism, hurting democracy, causing privacy to be at issue, can we rebuild an island of traditional advertising, actually trying to put advertising on quality content. I hope we clear the way. I’m not sure it’s going to be needed, invents the next model. But I think that new model can’t emerge until we clear away some of the problems from the current generation. I think we’re going to start seeing that sustainability being the lever that causes this to happen. Then a greener, better, more sustainable future for the internet is my small aspiration for the change we cause in the world.

I think it’s a great mission. I certainly hope you succeed, something that I’ve been very driven by my whole career. So best of luck with that. You’ve got a friend here in that. For the audience, we’ll put links to all of this and Scope3 work along with this video and the article. So Brian Kelley, thank you so much for spending a little bit of time with us here and the signal conversation. Really appreciate it.

Thanks, John.