Last week, DoubleVerify sued Adalytics over a report it published in March, which claimed that DV and others failed to block traffic from declared bots.
Declared bots should be relatively easy to spot, since they openly identify themselves to sites and services.
In its suit, DoubleVerify contends that Adalytics defamed the company by implying its pre-bid tech is ineffective and advertisers were charged for impressions served to bots.
AdExchanger spoke with DoubleVerify CEO Mark Zagorski a few days after the lawsuit was filed and asked what DV hopes to achieve. Zagorski would share nothing about the suit beyond the standard line: “We can’t comment on any pending or potential litigation at all.”
So we approached the question from a different angle.
Defending oneself against attacks is valid, but isn’t there a more productive way forward that’s better for the industry? It’s not like ad tech’s critics – or its problems – are going away anytime soon.
“We welcome challenges that make us and the industry better, and there are folks who want to work together to get there, like the IAB, TAG and the MRC,” Zagorski told AdExchanger. “We challenge each other and we move forward.”
But, he continued, “competitors who look to challenge us in ways that aren’t open and transparent and don’t really want to move things ahead – I don’t think that’s positive for us or the industry.”
Zagorski spoke with AdExchanger about DV expanding into performance measurement and optimization, the impact of overly zealous brand safety measures on news publishers and DV’s pre-bid invalid traffic (IVT) detection – which is at the heart of the company’s suit against Adalytics.
AdExchanger: The Media Rating Council doesn’t require pre-bid bot filtration for invalid traffic detection and sometimes even discourages it. So why do advertisers need to pay for pre-bid, especially if they pay for post-bid verification?
MARK ZAGORSKI: We look at it as a belt-and-suspenders approach. We can flag a bot on pre-bid and catch it on post-bid. The MRC focuses heavily on post-bid because it’s the most easily implemented, but we’ve shown that pre-bid and post-bid actually reduce the violation rate overall when they work together.
AdExchanger Daily
Get our editors’ roundup delivered to your inbox every weekday.
Daily Roundup
But what would you say to a client that wants to discontinue pre-bid detection because they think just a belt or just suspenders is enough?
First of all, a lot of our IVT identification is bundled and built into other products, like contextual suitability and ABS [authentic brand suitability].
But I’d also say that our detection addresses both general invalid traffic [GIVT], meaning good bots that are identifiable, and also sophisticated invalid traffic [SIVT], which is fraud. With GIVT, yes, it’s a wasted impression, but you should be able to get that back. SIVT isn’t a wasted impression; it’s true fraud, and nothing should run against it.
Last year, DoubleVerify started diversifying beyond verification and into performance measurement and optimization. What’s the strategy?
Our core value proposition has always been that if you take all of the garbage out of the system – nonviewable impressions, fraud and contextually irrelevant environments – what’s left should perform better.
Net-net, we’ve always kind of been in the performance business, but now we’re saying it more blatantly and also building and acquiring tools to do it, like Scibids for optimization and Rockerbox for attribution.
It’s like the holy grail of independent analytics, because we don’t buy or sell media.
But why isn’t it a conflict of interest to help with optimization and also provide the measurement?
We always and only act in the best interest of the advertiser. If we can compress the cost and show an ROI on average for every dollar an advertiser spends through Scibids, for example, then I don’t think there’s a conflict between cost compression and quality.
Advertisers control all of the levers. They’re the ones who determine what their suitability criteria are and they’re the ones who determine what their KPIs are when it comes to bidding. We provide independent analysis.
But DoubleVerify also has publisher clients. How do you balance supporting both the buy side and the sell side?
Hopefully, what we’re doing is providing more access to inventory on the sell side by creating greater transparency.
Because there’s the open web – and then there’s everything else, right? And “everything else” keeps eating more of the open web’s business. But if there’s a common language to understand quality, context, cost and attribution, that should create greater trust, which benefits both sides. That’s the role we’re trying to play.
Are advertisers being too conservative with their brand safety filters, especially when it comes to news?
There’s an infinite amount of inventory out there, so why would I take a risk if I don’t need to? I think that’s the way brands think and I don’t blame them for that.
What we’re trying to do is provide tools for them so they don’t have to be draconian and make overly broad decisions. You don’t have to block entire sites or categories if you have a scalpel. They can be selective but also more precise.
There’s been a lot of debate and contentiousness around brand safety recently. It’s become a hot button topic. What do you think people misunderstand about brand safety and suitability?
That safety and suitability are very different. “Safety” is the stuff that, for the most part, advertisers don’t want to be around, like violence, terrorism and pornography. “Suitability” is more nuanced, and it’s very specific to the brand. Unilever owns Dove soap and Axe body spray, but the brand suitability between those products is very different.
It’s not controversial to say that advertisers want to be around relevant content, whether you call that contextual targeting or brand suitability. But it’s not only about protecting their brand; they also want to drive performance.
There’s a narrative, though, around the idea that when safety and suitability get lumped together, publishers are hurt because certain content is considered brand unsafe and gets excluded from a buy. Yet, in many cases, the content isn’t considered brand suitable, and advertisers would never have bought it anyway.
Verification companies are never going to be loved. If you’re a seller, you’ll say, “I’m getting a bad shake,” and if you’re on the outside, you’re going to say, “They’re making a bad call.” But we’re getting paid to make calls.
No one roots for the referee.
This interview has been lightly edited and condensed.
For more articles featuring Mark Zagorski, click here.
