If you’re not getting John Dick’s weekly newsletter, “What we’re seeing,” you may not know Dunkin’ Donuts customers have higher credit scores than Starbucks’ drinkers, or that people who eat ice cream from cups are happier than those who prefer cones. Starting with fewer than 50 readers, Dick’s weekly letter is the voice of CivicScience, a data intelligence company that runs upwards of four to five million polls a day. Readers willingly answer questions on subjects from consumer spending to streaming platforms, or whether they’re willing to eat a peanut butter and mayo sandwich. (That gets a nope from us.)

Next up for CivicScience? In this latest conversation with John Battelle, Dick, CEO and founder of CivicScience, talks about his new commercial advertising platform, already in a soft launch, which uses context to understand readers and customers in a privacy-friendly manner without cookies or tracking.

You can hear more from this lightly-edited conversation, or read the transcript below:


John Battelle 

Joining me today for a Signal360 conversation is John Dick, the CEO and founder of CivicScience, a consumer data intelligence company, really remarkable and unique. John, welcome to the Signal conversation.

John Dick 

Thank you, John. Great to be back with everybody.

All right, let’s get right into it. Before we get to the news, I was wondering if you could refresh our audience’s memory. What is CivicScience? And what makes it unique?

Well, we started as a polling company. I wish I had a better word for it than that. But we engage people in questions and answers across the web in a very proprietary way. We had a clever idea, about a decade or so ago at Carnegie Mellon to engage people inside of content they’re consuming on the web, reading an article on a website about the Beatles. We asked them a question at the end of that article about the Beatles, turns out they’re willing to answer three or four more questions about other topics in order to see the results of the first question that they answered.

We figured out how to do that at a tremendous scale, we’re doing upwards of four to 5 million of these polls a day, billions of answers now historically, trended over time. We’re now evolving that as world of digital advertising is evolving and or running toward some significant privacy evolution. It turns out the data we gather actually has a lot of really strong applications in a privacy-compliant way for media activation. So we’re growing that part of the business on top of our consumer insights business.

You announced some news that I found very interesting. I think our audience will as well since the majority of our audience is very involved in the marketing world and in the digital marketing world, which is one of the mainstays of the Signal conference at Signal360. Tell us about that news. What did you announce? Let’s get a little deeper into what makes it unique, particularly in the context of what you mentioned earlier, as it relates to the shift in the ad tech world.

The most notable news of that is this ad platform that we’ve been working to build now for almost two years on top of our data set is now commercially ready. We knew that because, and made that decision based on the last wave of campaigns that we’ve run, using our platform have just been increasingly performing. So I’ll explain a bit of how we do that.

We’re asking surveys, massive amounts of them every day, fully permission, CCPA, GDPR compliant. The respondent explicitly gives us the ability to capture and store that information and use it in this way. The thesis behind this business was that future-proof privacy compliant self-reported audience data, analyzed at scale and an aggregate, would allow us to define audiences on the web in a very ethical way. And by using that data and really sophisticated models that we’ve built, we can facilitate highly performant campaigns. The outcomes of the last sort of five to 10, POCs that we’ve run, really just prove to us that, yeah, we’re onto something, we’re seeing, even with fairly early stages of the technology, campaigns with engagement lift of engagement of over 30%, on average, compared to some control campaigns we ran with maybe not as ethically sourced datasets for comparison. And we’re seeing upwards of 60+ percent conversion rates. People are buying the thing our advertisers are trying to sell. Having done all that work, and all that due diligence, made the decision that the business is ready to launch. And we announced that just this past week.

Now, from the point of view of a marketer, what is this compared to in terms of channel? Is this sort of open web only? Does it go across social? And how are you acquiring the inventory that you’re then offering to your marketing partners?

Our early POCs are open web only. We have absolutely have designs on connected TV. We have done a couple of case studies and social within social environments, but it’s a bit of a different story there. Right now, we’re running it as a managed service, just as we’re kind of proving the capability, longer term, because our polls are derived from first party integrations that we have across a network of hundreds of premium publishers, Microsoft News to Penske media, and the like. Our longer term plan is to essentially create a private marketplace across that universe of publishers. But right now, we’re running more open web campaigns as a managed service. And as we’re building the data capabilities, and the modeling capabilities will map that to other to other channels down the road.

Make sense. Can we pull back a bit? Maybe I’ll ask you to do a bit of a, you know, admiration of the problem. You hinted at it. But what is the shift that’s occurring in the digital advertising landscape that set up this opportunity for your business to get into?

John, it’s a mix of the shift that’s happened, to some extent, so far. The shift we’re betting on is going to happen around privacy and identity. I understand there’s a broad swath of points of view about what will happen and what should happen. We tend to be in the extreme end of the privacy mindset around this, that we need an Internet that does not have cross platform identity, because it’s not good for humanity. It is not good for publishers, we believe for particularly premium publishers. And frankly, we don’t believe it’s great for for large brand advertisers. And I’ll explain that a bit.

First, let’s define a term just as a proxy for the audience. When you say, cross platform identity, are you speaking of large companies like Google, Amazon, Apple, Facebook, know who you are everywhere, as well as the open web kind of ad tech ecosystem has trackers of one kind or another and, and creates that kind of an identity network?

Yes, all of that. And the third-party cookie is the thing people probably talk the most about in the open web scenario. And Google’s sort of been head faking about that now for a few years. But if you take them at their word they fully intend at some point to fully eradicate it. Obviously, the ad tech world has been frantically trying to develop alternatives, hashed emails, and the like which frankly, if you step back and you think about it, are just cookies have a different flavor. They don’t really do anything to address the privacy implications of cross platform tracking. So, yes, obviously, some of those bigger players do it and Google and Facebook and the like. But there’s a full there’s a full sort of cottage industry now of these new identifiers being spawned to track people across the web. The most nefarious terminology for it is surveillance marketing, which is quite frankly, an appropriate terminology. We deeply believe that it’s not good, not good for humanity, not good for publishers, and really not good for big brand advertisers. It’s nobody’s business, what I do as I traverse the web. While I may have a first-party relationship with Target, and I’m comfortable with Target, knowing everything about what I buy at their store, it’s not Target’s business, what TV shows I watch if I don’t volunteer that information to them. Right now, we have some plumbing and infrastructure that allows those things to happen to the detriment of pretty much everybody involved.

From the point of view of a marketer, what is on offer through the insights you’ve derived through billions of surveys and paying attention to what people are interested in what they tell you. Are you creating kind of a contextual audience network, as opposed to a first-party identified audience network?

It’s a bit of both. Within a given publishers, let’s take a Nexstar, for example, where there is a first-party identifier within the constraints of the portfolio of Nexstar properties, we are able to identify that user, at a user level within the constructs of that platform. So, enabling advertisers to reach that person on a first party basis inside of say, an Nexstar is feasible and privacy, compliance and an ethical and all of those things. When we roll back to the broader open web, we are able to do on a modeled basis by understanding the given attributes of an archetype of segment and an understanding of content they’re consuming on the web, can pair the identity or the look alike model of that user against a piece of content they’re consuming across our broader network of publishers.

What’s the scale? If I could be the proxy for Procter and Gamble, they’re always interested in the question of scale. They’re a mass reach organization, as are many others. So what is the scale you’re talking about here as it relates to this new product? Or is it still scaling up?

At the identity level, across sort the, quote, private marketplace model I discussed, that identity sort of footprint is absolutely in its early stages and scaling up. But the campaigns that we’ve run have been internet scale. It’s all based on building models. If you think about it, we use a survey question to ask someone, ‘Are you currently in the market to buy an automobile?’ And we get tens of thousands of people to say, ‘Yes, I’m currently in the market to buy an automobile.’ Turns out, that’s a really good signal of somebody who’s in the market to buy an automobile. Often that is inferred, maybe based on a search someone does, which of course, is very reliable, but also maybe can be misleading, or maybe a piece of content that they consume. So we take that audience of people who say ‘Yes, I’m in the market to buy an automobile,’ we’ve got now billions of data points to connect and analyze among across that universe of people who say they’re likely to buy a car, and we’ve got demographics, and psychographics and all the other content attributes we have about people like that. We take that seed audience of people and we can map that to a much larger audience of the web based on some of those contextual signals that you mentioned, or other local-like attributes that we can build against.

Another one of the things that Signal pays attention to is innovation and entrepreneurship. You’ve clearly over the past 10 plus years now with your company been an innovative entrepreneur. This is a big step. We’re talking about moving from being a data-driven insights, business and an intelligence business, into potentially an app advertising business that could possibly be much bigger than the business you’re currently in. How do you think about that? What kind of conversations and deliberations did you have, as you decided whether or not to get into this new line of business?

Well, I did it kicking and screaming, John, to be honest with you. Quite frankly, I’ve always found ad tech to be somewhat repugnant. In the earlier years of our business we were tempted and heard frequently into that world of, ‘Hey, CivicScience, you have all this fabulous data, you should put it into these, you know, swamps of other data, and people can buy ads against it.’  First of all, we feel like we have a covenant with our respondents that we weren’t going to share their individual level data with the broader universe. It just always struck me as a market that was very murky, and maybe, again, sort of nefarious in some ways. But it was really our publisher and partners initially that were coming to us. This was now probably about three years ago when they spawned, saying, ‘Hey, you know, John, a lot of this infrastructure we use to power ads on our platform is going away and or will go away. It occurs to us that the way you gather data about our audience is sort of future-proof, vis a vis, a lot of these changes that are going to happen.’ We hummed and hawed about it and said, ‘No, we’re not really in that business.’ Then some of our brand partners, whom we inform with our insights that they use to drive kind of media planning and strategy, started saying the same thing. ‘Hey, John, all this data you provide to us that helps inform our strategy, it occurs to us that it’s being gathered inside of these premium brand environments where we would like our creative to live.’ So there was just a groundswell of that.

The tipping point was the CES conference in 2020, right before COVID, where I met with a lot of these different people and felt like this is a really big potential business opportunity. To give a little bit of behind the scenes, we engaged with one of the very large consulting firms whose name everybody knows, to assess the market opportunity for us. That assessment came back very promising. We worked with them to develop a business plan, which then became an investment document. We brought in a large swath of new investors to the company to spawn the business we now call Rouleau. I will say it’s not in any way replacing our intelligence business or even supplanting it. It’s turning out more and more that the two businesses kind of create a flywheel between one another and that’s the end game. We think more about what the aggregate of those two businesses can be, that can it can be a very large business if we can execute on the advertising side.

That leads me to my final question in the newsletter that I read a week ago or so actually, just a few days ago, you said, ‘We haven’t really even scratched the surface of what’s possible.’ Can you sketch out that if everything goes largely to plan, what does this look like in two to five years?

Right now, we’re focused on lots of proof to both the advertisers as well as our publisher partners, that this new way of driving audience creation and driving activation, can perform on par, if not better than a lot of the ways it’s been done for the last 10 to 20 years. To get greater scale, we need more publishers in our universe. And that’s a bit of, I don’t know if chicken egg is the right metaphor, but as we demonstrate greater efficacy of the campaigns we run, we find more publishers who are interested in bringing our platform onto their site. That a bit more of the flywheel as well. It’s starting to create a snowball effect. Because previously we would sell our capabilities to publishers who implement our widgets inside of their content. We don’t pay for that. So we’re getting six to seven million page views of our JavaScript across a lot of premium publishers. And we don’t pay a penny CPM for that. Why? Because we provide some insights back to those publishers to help support their ad sales or editorial content, and so forth, a very sort of analog barter with the publisher.

Now we’re saying to them, ‘Hey, in addition to all those analog services we provide, we’re able to activate this first party data for you and drive new ad revenue to your platform.’ As we tell that story that we’re creating a greater groundswell of publisher interests, which helps to grow or scale and so on, and so on. So there’s that sort of virtuous loop that that creates. So really driving and accelerating that virtuous loop is our huge focus of ours this year. And a bit of the chicken and egg there is we need enough scale to attract the right advertisers, to bring the right dollars to the publishers, to get more publishers and so on. So that snow that snowball is the thing that will drive the company. Now in terms of scratching the surface, there’s also a tremendous amount of technology investment we’re making in machine learning and AI to get smarter and smarter about the models that we’re building, testing those models, learning from those models making the next wave of models smarter.

And then last but not least, the thing that we certainly advocate for and try to be as outwardly outspoken about as possible, is privacy policy actually evolving. There’s still a big part of the market that’s either in denial about coming change, or actively resistant to it. That tends to emanate a lot from the existing ad tech infrastructure players as well as the agencies who benefit from the current state of play, either because they’ve invested immensely in the infrastructure or they benefit from the complexity and the opaqueness of it. Advertisers need to pay a lot of people to help them advertise because its positioned to them as an incredibly-complicated thing to do. So we’re swimming upstream against a lot of industry energy among players who don’t want to see things change. So our position in say, two to five years is going to really depend a lot on how that evolves.

I can see it, and congratulations on this big step for your business, and I very much appreciate you spending a few minutes with us to explain what’s going on at Civic Science. Hopefully, we’ll see you this coming July in Cincinnati once again.

I’ll commit right now if you’re when you invite me John and I had a blast last time. It was a great group. incredibly well put together event so would love to join. And let me just say, we did announce that kind of news last weekend but we’re not working on bottles of champagne yet. We still have a long way to go. It’s a big market to try to move. Advertising is deeply entrenched in the way that it’s done. There’s a lot of sunk costs and infrastructure and other things that we’re not strong enough to move on our own. It’s going to require a lot of visionary leadership from marketers to drive change in advertising. That’s where the money comes from when we talk about government, maybe regulating us toward better privacy or the platform self-regulating, which is never going to happen, really, to its full extent. It’s really going to depend on the spenders, they’re the ones who drive market. So anyway, thanks for having me, John, and I look forward to seeing you again.

Thanks very much to John Dick, CEO of CivicScience for this month’s Signal Conversation.