Innovid CEO: “In-Banner Video Is Not Video”

ZvikaNetterInteractive video ad tech startup Innovid has changed significantly since its beginnings in 2007, when it had nothing to do with advertising.

Though it has since helped companies like Toyota, Chrysler and Sony Pictures serve up dynamic and interactive pre-, mid- and post-roll video ads, and has worked with both Roku and Sony Playstation to deliver over-the-top ads, it was founded by a group of guys with “no media/ad tech backgrounds whatsoever,” said Zvika Netter, Innovid’s cofounder and CEO.

Innovid’s initial incarnation was as a platform that dissected video content frame-by-frame and extracted metadata. “We spent three years building the platform and interacting with studios such as MTV and CBS to change the whole format of how news, sports, cooking shows (looked in digital),” Netter recalled. “It was a lot of cool, fun stuff, but absolutely no revenue.”

No revenue, however, isn’t a sustainable business model and in 2010, Innovid began its exclusive focus on advertising, working initially on pre-roll ads and now partnering with a number of video ad networks, premium publishers and tech platforms to serve interactive, cross-screen placements.

Netter spoke with AdExchanger for the next installment in a series of Q&As evaluating the video ads ecosystem. Past interviews have included: BrightRoll, Ooyala and SpotXchange. This series will also include Videology, Vindico and more.

AdExchanger: Describe the Innovid platform.

ZVIKA NETTER: Our Atom platform runs pre-roll (interactive iRoll) and other formats across web, mobile, connected devices and we’re extended that into the TV world. We’re not active on the traditional TV dollar side, but we have insight on the digital side, and we absolutely see digital video budgets increasing where we are active with things like addressable, personalized and dynamic video.

We demonstrate that through pure data and level of engagement, and time spent metrics, which we convert into ROI. We made a decision we wont touch anything outside of pure video.

What’s “pure video?”

In-banner video is not video to us. For us, it needs to be in-stream, in a player, before content, during content, after content. We’re taking the TV model and taking it to the next level in digital.

Innovid signed on recently as an exclusive partner for Facebook’s ad server Atlas. How’s this transpiring in light of the LiveRail acquisition?

We signed an exclusive deal that whenever an Atlas customer needed a video solution, everything – the engine, ad server, iRoll – that it would run through Innovid, and it’s a deep integration with Atlas. We can start a campaign on Atlas, traffic it, it will run through Innovid and the data will go back to the Atlas dashboard. It’s a deep and strategic relationship.

I can’t talk too much more about it, but Facebook is extremely serious about it and they’re rebuilding their entire platform. And I think they’re investing in the whole tech stack with the LiveRail acquisition. We have many more direct to client relationships though, independently.

Do you have a DSP?

We made the decision to stick to the technology and commit to it. That’s the fixed CPM and the fact that we’re not tapping in to the media business. We’re not a DSP. We have all the functionality of a DSP, but a DSP is a commodity – you plug into LiveRail, AppNexus, any exchange really. We tell our clients that the way that DSPs operate today, it’s a media business. It’s not a platform business. So when Chrysler, for example, runs ads on TubeMogul, a DSP, Hulu, YouTube, CBS, AOL, Yahoo and across mobile, Roku, etc., 100% of those ads are going through us, being delivered, measured for verification and fraud. There’s no way they make money from video ad serves, though. The only way to make money for everybody else is to sell media through a DSP.  Video ad serving is a terrible business in terms of margin. If all you do is traffic ads and run them and collect reporting, it’s a low CPM product and you’re providing managed services. You’re basically outsourcing the work of the agency.

How does Innovid make money?

We make money on iRoll and VOD and connected TV, where all the other players are going after the media dollars. We came to it not from an ad network perspective where there was tons of money to be made in the early ad net days. We made a decision we would be 100% technology focused, 100% media agnostic and that’s a very important distinction, because everyone sees the marketing and SaaS platform-model take off. There’s so much clutter and a lot of VC dollars are going into marketing, which hurts the whole industry.

Are you saying managed services and platforms are mutually exclusive?

Many of these guys are our partners and some are even our clients. TubeMogul, for instance, uses our iRoll technology, white labels it and that generates revenue for us, too. But at the end of the day, many of them are managed services. They’re a dashboard to buy media programmatically, and no one at the agencies besides the trading desks really has capacity or bandwidth to run this. At the end of the day, they hand TubeMogul an I/O, say “I’m looking to buy male/female,” they transact on that and they come back and say “you got this percentage viewability” and there’s so much room there to make money because it’s percentage-based.

The Innovid model is a fixed technology CPM. We don’t care how much money you spend, we care how many impressions went through our system, so it’s all volume-based. And we don’t have any media affiliation.

What’s your volume and headcount?

In 2013, we ran… about half of $1 billion. The media was not transacted through us and it does not flow through us, but if you want to compare apples to apples from a scale perspective, it’s comparative to about half a billion. We have 135 employees and 70% of our revenue comes from the US with 30% outside of the US. Global markets are growing insanely fast. We raised about $30 million to date and Sequoia Capital is our lead investor.


Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. In-banner video is in fact video, it’s just not preroll. The preroll world would love to appropriate the entire word so all those dollars and attention flow into their segment but the reality is you can get video in a standard ad placement and that actually IS video.

    In-banner is specifically the replacement to the outdated and incompatible 40k Flash banner spec. It’s not sold, and should not be sold as preroll as it’s a different use-case. VAST is fantastic but Display, in-banner video units, is about 20% the cost and can fill in all of those cracks and crevasses that VAST can’t reach. Every site and app does’t have a video player so VAST suffers from a lack premium inventory and reach.

    Is seems that every adtech category wants to sell it’s value proposition against Display. How many times have you heard that this or that will replace banners? I think we all know that every media plan will blend of all these products.

  2. Patrick Toner

    While you’re absolutely right that in-banner should not be sold as in-stream, it regularly is sold as pre-roll – and in most cases, buyers are willing to do this because they can report higher completion rates on an in-banner placement compared to a genuine pre-roll.

    When an ad runs in an in-banner position, below the fold and muted, who is going to close it? Who is going to skip it? The only reason for it not to complete is that the primary content on the page only held the viewers attention for less time than the ad unit’s length. If the advertiser’s buy is GRP controlled, and your KPIs are dominated by completion and reach, the figures are going to ‘look’ great.

    The reality is that media money has been dumped in a bin.

    Having said that, there are clearly qualitative differences between different vendors’ in-banner placements.

    Pre-roll doesn’t have to just be VAST – it’s been VPAID for 5 or 6 years, and content can be just as rich, interactive and immersive as any RM (or dedicated website) counterpart.

    It’s funny that you’d say “cracks and crevasses” as well, as there’s clear evidence to show how many deliveries are basically lost down aforementioned holes. Engagement rates on in-banner positions are dreadful compared to their pre-roll counterpart, even though the exact same content is running.

    When video runs in an in-banner position, 99% of the time, it has to be muted unless activated (by a click or a rollover). The opposite is true of pre-roll. Given that video is an audio & visual medium, it’s probably disingenuous to claim that in banner video has anything approaching the value of in-stream.

    I can’t help but feel that when in-banner video is on the buy, it’s a box ticking exercise to help hit reach and completion targets to offset the costs of placements that actually put the video front and centre. It’s the ‘best’ way to be appearing to stretch a budget, but it’s doing next to nothing for brand, and in some cases, can actually do more harm than good, due to the potential poor quality of the placement.