In beginning his new mobile ad startup, Admobius, co-founder Dan Grigorivici admits he's no "spring chicken" when it comes to the ad technology landscape. He's worked in a range of product-related capacities spanning digital agencies Carat and Digitas, mobile ad network Quattro Wireless (sold to Apple) and behavioral ad network Tacoda (acquired by Aol).
Along with his two cofounders, Ray Duong and Omar Abdala, Grigorivici says the idea for 12-person AdMobius emanated from the need for what they saw as a "mobile audience management platform that is able to help both publishers and advertisers to discover and target relevant audiences of scale."
The company's products have been in "beta" since the beginning of the year, but this week marks their formal launch with $5 million in fresh, venture capital which AdMobius originally closed in January. (Read the release.)
AdExchanger spoke to Grigorivici about this company and mobile advertising.
AdExchanger: First, what problem is AdMobius solving?
The mobile display ad market is growing, but is still nascent. What we are trying to do is help advertisers with rich [mobile] audience targeting at scale. Obviously, mobile devices are the richest mediums that deliver data. There are certainly privacy considerations, too. But, in general, part of the reason the mobile display ad market is growing, is it's only one percent of total digital spend today. As a market, we are not - overall - able to deliver [mobile audiences] at scale. That is exactly what AdMobius is setting out to do.
How is AdMobius going to offer effective audience targeting, audience segmentation, given the concerns around privacy and the mobile cookie?
There are multiple questions there. Let me address the mobile cookie first and then I'll talk about privacy.
First of all, in terms of both the ingestion and delivery of platforms, it does not assume a particular format of ID.
Our philosophy is that unique user identification in mobile will continue to be fragmented. That's not necessarily a bad thing. We are ingesting multiple different types of IDs, never the original UDID, never the original device ID. We index everything in our database in terms of our own AdMobius ID. First of all, we essentially stitch together, if you will, multiple different non‑personal identifiable IDs. From a privacy standpoint, we are also progressive, we believe. What that means is that we're very close to being certified by TRUSTe.
We have an opt‑out mechanism for the consumer. Regardless of what the original source of the ID for the opt-out is, we'll drop them from the database indefinitely, from any other source that we've seen that particular ID.
Also, we have no personal identifiable information in the database. Everything is only relevant if you have specific high‑level characteristics. In other words, we have no, name, address, any personal identifiable information whatsoever obviously. This information is dynamically changing based on behavior of the IDs. We even have a minimum number of IDs in a particular segment so that there's no possibility for either ourselves or anybody else that delivers the ads using the data to reverse engineer, and try to figure out who this ID might be.
So what then is your solution around creating effective targeting? What are those datasets that they could possibly unlock mobile, if you will? Is it an in‑session mobile cookie, that people can use for retargeting, for example?
We do have solid intellectual property around the ability to identify probabilistically singular session cookies, if you will. My point earlier about not relying on knowing specific types of IDs referred exactly to this. The "teaching" technology that we have is about identifying very similar patterns between different cookies, and ascertaining with what probability this cookie is either similar or identical to this other cookie.
What's the go‑to‑market strategy?
Our customers are premium publishers, ad networks, and DSPs. In terms of the go‑to‑market, we do have effectively a global dataset - and I'll give you a couple of stats about that as well. For the go‑to‑market strategy, we're attacking the US markets this year, and extending internationally later next year. In terms of [devices], we are covering about 91 percent of data coming from iPads globally, about 67 percent of iPhone, and about 16 percent of Android data.
How does AdMobius make money? Are there buy and sell side opportunities? Feel free to use use cases if you like.
Sure. Let's say Burberry is a premium advertiser who wants to reach luxury consumers. And there's a large‑scale publisher like Angry Birds. On a purely contextual basis, it is extremely complicated for them to deliver to that audience, unless they know what portion of the audience is in fact potentially a luxury buyer. What we enable Angry Birds and similar publishers to do is identify which of the IDs are luxury buyers and enable them effectively to deliver that [audience segmentation] to their advetisers. It's a similar use case for the ad networks and DSPs as well.
In terms of the product line, we have a three‑pronged product line. First, we have "standard" segments, which is generally available audience segments. They contain demographic information. Those are offered through a CPM rate card, which varies depending on what kind of segment it is.
Then we have what we call "custom" segments. These are not generally available and enable our customers to create their own segments, store them and and A/B test them in their desired target markets.
And lastly, we have "private" segments. These are geared for helping our customers differentiate in the marketplace by providing advertisers the segments that are only to be found on their particular networks.
For these cases, we have a combination of a CPM floor rate combined with a revenue share. A rev share off of the ad media delivered through the usage of our data basically.
In general, what's been surprising to you about mobile advertising these days?
Not to be self serving, but I think what's going to kick the market into high gear is the ability to target in terms of audiences and leveraging data. The next thing I think important is the user experience of the ads themselves. We try to address this a little bit in our ads. But, the current form of the creative on mobile devices is not going to be the future of it. The fact is that today it's a very similar experience to [PC-based] display a banner on the top or the bottom. Ultimately I think that's going to fade away and be replaced by something much more engaging.
The other thing that's crucial and ties those two elements together are the data and enabling creatives to be very relevant and personalized based on the data.
How do you "define" mobile that your company addresses?
What I consider mobile is obviously smartphones and tablets. That's basically high frequency content consumption. Ultimately, consumption by the consumer will shift more and more to the mobile device.
Finally, in the next 12 to 18 months, what are the some key milestones that you want to accomplish?
International expansion and internationalization of the segments to the usual suspects in India and Asia - that's going to be later next year. Then from a high‑level product standpoint - launching a self‑service platform that enables our customers to create their own segments. Today we're a fairly managed service. We have internal platforms, but it's not yet direct to our customers.