I recently sat down with MadHive Founder and CEO Adam Helfgott to discuss his firm's recent deals with Inscape and TEGNA, why OTT is so important to the future of local advertising, and how OTT fraud actually works.

You recently signed deals with Inscape and TEGNA around local advertising. Why is local so hot right now?

Local television buyers are starting to understand that buying OTT, on top of buying linear TV, is a great way to extend reach. And that’s what they’re trying to do: get more reach and more frequency in order to drive conversions, whether that means store visits to a car dealership or purchases from a furniture store. Measurement is critical so that they can buy media that's working harder for them.

So where do Inscape and their ACR data come in on this?

What we're initially doing with Inscape is to look at what networks in the local market are driving action. So it’s not individual viewers, it’s looking at say, does WXYZ drive more traffic to the local Ford dealer than WABC? And we can even look to see whether WXYZ drives more traffic when it’s raining-- any sort of variable.

And then we look at the IP addresses of people who watch WXYZ and we can see which OTT programming they’re also watching, so that the brand can target them there. If you know that your optimal audience are those people who watch WXYZ and you can reach them on OTT, then you’re extending your reach with your best customers.

Is that something that can be extended out to national campaigns

Definitely. At some point, all national campaigns are going to be purchased the way local is, just at a national scale. The top 10 national advertisers are already doing it. And what I mean by that is that they'll focus their marketing nationally, but they’ll vary the amount of spend they put against different markets. They’ll see where their best customers are or where they need to make inroads against their competitors and they’ll adjust the spend accordingly in those DMAs. And so even though it’s a national spot buy, they can double-down in different local markets based on data.

We can also help them figure out where to allocate budget and how to extend their reach to OTT on a national level. If they need to amplify their message with a certain demographic in the Chicago area, we can use the Inscape data to help them figure out how to reach those same viewers on OTT, so that they’re getting more reach and more efficiency out of their ad spend.

We’re seeing a big shift to ad-supported OTT viewing, one that’s only going to get bigger once the Flixcopalypse happens and we have all these new high-quality ad-supported services from the likes of Warner and NBCU. So why has the industry been so slow to shift dollars to OTT advertising?

Over the past several years, we’ve seen rapid adoption of OTT by consumers, but measurement has been, to put it mildly, wonky. So it’s funny--20 years ago, you had a pretty good idea who all your viewers were just from the Nielsen data. It wasn’t perfect, but it was close enough. But now, the OTT viewership data can be so all over the place that brands are feeling they have no idea who is watching what.

In a way, that’s great for us because brands are starting to demand tools to let them know that they’re reaching the right audiences. They want to know how to find the consumers that matter to them, the ones they’d previously been reaching on linear TV. And we can provide them with that information using Inscape data and we can even track the ads using cryptography to make sure that there’s no fraud involved, that the ads are being seen by the people the brand is paying to reach.

How big a problem is fraud for OTT right now?

We’ve found that as much as 18% of OTT inventory is fraudulent in the open marketplace. And with Beachfront Media estimating that the open auction marketplace contains about 5 million OTT impressions on any given day, this could quickly become a lucrative opportunity for fraudsters especially with the premium CPMs on the platform. This was one of the big driving forces behind our research -- and of course AdLedger’s Fraud In OTT Report, which brought together Hearst Television, The Hershey Company, Reprise/IPG Mediabrands and other members of the consortium to address the issue in a collaborative way. One of the big takeaways was that the predominant methods of fraud on OTT - misrepresentation, app based and device based - mirror that of traditional digital, and the only way to ensure we don’t follow in the footsteps of our digital fore-bearers is to bring together key industry stakeholders to tackle these problems head on.

Can you explain how criminals actually defraud OTT ad networks?

OTT is fraud-free if I do direct deals with all of the providers. But fraud can happen when you start dealing with aggregators. So a fraudster goes and gets connected to a tier five SSP, who has a deal with someone who aggregates a ton of traffic for these tier three SSPs. They then spoof the traffic to look like something else, and then they send that spoofed traffic to the tier three SSP who then sends it into a tier one SSP.

And then people buy that and they think they're paying for a U.S. viewer watching Fox, but they’re not-- it could be a viewer in a smaller market overseas, it could be a bot or it could be a fake impression.

We use cryptography to track all of these views and make sure that the traffic is coming from where the aggregator says it’s coming from. So a company like Beachfront Media, who works with MadHive, has eliminated fraudulent traffic. And that’s super important because if we can eradicate fraud now, in the early days, then it won’t become a thing, you won’t have more people trying to make a quick buck off a fake OTT ad.

One final question: how do you reconcile digital style metrics, which rely on direct response style KPIs, with television, which is more about branding and emotion?

There’s a tendency to measure ads by their actual response, especially for local advertisers, where the advertiser is often a car dealer or a health club and they’re measuring the amount of foot traffic generated by the ad.

But if a national advertiser wanted to measure something squishier, it’s easy to change what you’re measuring from “did the customer walk into the store” to “does the customer have a slightly different impression of the brand?” It's possible to measure image and branding, you just need to ask the right questions and then you can determine the impact the commercials had and which audiences changed their perception of the brand and which didn’t. From there, you can either change your creative or your media plan or both based on those metrics.

Keep up with Adam Helfgott on Twitter and LinkedIn.

Header image by photo by Constellate on Unsplash.