James Avery is CEO of Adzerk, a publisher-side ad technology company.
AdExchanger.com: Why focus on publisher ad serving as opposed to advertiser ad serving?
JA: The reason is publishing. I view publishing as crucially important to us as a society.
When I originally started running ad networks, the reason was due to the fact many of my friends - including myself - had blogs but we weren't making any money. Basically, we would run AdSense and get $.50 to $1.00 CPMs. So, as a side business, I went and started an ad network to get the $5.00 to $10.00 CPM's and make it worthwhile for them to create new content and spend more time on their site.
So, I look at ad serving as a publisher side tool that I can use to support the people to whom I really relate - the content creators and people who are putting themselves out there for free and need to make money from advertising in order to do it.
Please share a quick top line on your background before Adzerk. And, where did the name for the company come from?
Before Adzerk, I was a software developer and worked at Dell and at Avanade, which is a joint venture between Microsoft and Accenture - essentially high‑end consulting. Then I went out on my own and started a consulting company called Infozerk. The Infozerk name came from "berserk" - a term I like - and I wanted something that stuck in people's minds a little bit more than other consulting names which are a bit more tame. I'm not tame. So Zerk represented me as a consultant pretty well. After Infozerk, I started working on running ad networks and that company became ZerkMedia. The original software I used to run those ad networks became Adzerk. So, I came from an engineering background and fell into the media world.
Given your experience with ad networks, what is your take on the ad network model today and where do you see it going?
The term "ad network" is used for a wide variety of companies. You have the very first ad networks... the broad, horizontal ad networks. I think that is what people are talking about when they are saying these things are dinosaurs. There is no point. If you are Ford, and you just want to reach five million people between the ages of 18 and 34, you don't need to go to one of those networks anymore. You can go to a DSP and buy that traffic.
Then, there is the second type of network, which is the vertical ad network. We actually see this as still alive and driving forward, especially at a small [niche] level, because when you get to small publishers, they don’t have a direct sales force. They can plug into AdSense or one of these horizontal ad networks, but they don't get near the CPM that a vertical ad network can get them. So if you have a golf vertical ad network, it is going to get higher CPM's than just throwing your data or throwing your impressions out onto an exchange and letting people sell it as golfers. Vertical ad networks are where I got started with publishers who did 50,000 impressions a month.
Another thing is that the performance ad networks are the ones who are really doing the audience buying. Those are the networks that have taken off in the last year or two.
What would you say is the problem that Adzerk is solving?
If you are a publisher, you have have almost no way of knowing which resources in the Lumascape slide are going to be the best fit for what you need. We see a huge amount of friction and complexity for publishers. So our goal is to simplify that. But I think the key differentiator in how we want to simplify it is we don't want to be that publisher who clicks a button and says, "Yes, I want to sell my inventory to Google." Then, they are done. We don't think it is ever going to work that way. We think that good publishers are always going to have to learn how to get the best returns from their content.
What do you consider your target market? Is it every publisher or ad network?
It is every publisher or ad network. We want them all to eventually be our customer. As we are starting out by working with lots of ad networks because we have a lot of ad network functionality built‑in. Basically, a lot of what Adify had - we have. We are picking up Adify customers or people who are using OpenX. Then, it's from the smallest publisher, who comes in and wants to use the two ad networks to the absolute largest publisher, who wants to integrate their DMP with their yield management and also control their inventory. So it's every publisher out there.
How does the exchange model impact what you are doing?
It is complementary. For a little while, everybody thought every impression was going to go on the exchange and it was going to become this automated world of: "I'm a publisher. I switch a button and get the maximum amount of revenue for my inventory." I think that what we are seeing is that it isn't the case. Exchanges are valuable. But I think they have to be balanced with direct sales, controlling your data and understanding where your data is being used. So they are complementary. You want to send certain traffic to them and not other traffic to them.
There seems to be a trend in the industry around apps, and Adzerk seems to be taking advantage of it. Why is this app trend happening, and how is it going to work for Adzerk's product strategy?
The app trend is something you're seeing across industries where people are seeing the benefit of a pluggable model for integrations. I'm sure that's spurred by the success of the iTunes App Store on the mobile devices, as well as on the Mac OS. Everywhere you're starting to see the success of app stores. It becomes a logical jump for just about any product. Any product that has a lot of users using it, they'll see app stores on TVs pretty soon. They're going to be coming up everywhere.
Why do you think the app model is successful? And how does it relate to your app "store" solution, adOS?
The reason it’s so successful is the reason we started to look at it, As we were building adOS, we talked to publishers and they would tell us that they wanted a certain functionality. A publisher would say, "We want the ability to optimize our revenue to inventory" so we started to talk about what they needed out of that. We found out that, for some of them, something like Rubicon [Project] would be a perfect fit. Rubicon would do everything they wanted to do, and they should just use Rubicon. For some people, they needed something slightly different, where PubMatic would be a better fit. Or AdMeld would be a better fit.
What we learned was that if we did try to build the right product for all of them, it would be a super set of all of the features in Rubicon, PubMatic and AdMeld. That was something that we knew we would never achieve because those companies are continually evolving and adding new features. That's when we took a step back, and we said, "Why don't we just make it easy for our publishers to pick from these three different tools?" Then, they could pick which one works the best for them. That’s when we started to realize that everybody wants to think that they can build the perfect solution for their customers, but what the App Store realizes is we're never going to build the perfect solution, so we're going to give you access to the solutions to bring home. When you look at the iPhone, if you want a task app, there's probably 50 different task management apps in the iTunes App Store and every one of them is going to be better for a certain person.
How do you handle pricing and what might you be thinking about down the road with pricing, especially as things become more modular?
Basically, we started out with a classic, ad server pricing model, where we charged per CPM of ads served. This was anywhere from five cents all the way down to, depending on scale, one and a half cents or one cent. Basically, if you serve a million ads, it's $10 to $50, depending on your scale.
With adOS, we're doing two things. One, we're making it free for publishers under 100 million. That's because we wanted to help those small publishers get the benefit of adOS without having to pay us out of pocket. Two, we're making the pricing structure much more modular. If one of those publishers, a publisher who's under 100 million, signs up but wants to have access to our network tools - this is our tool set around running a vertical ad network - when you sign up, that costs three cents a CPM, and you can actually add it in the App Store. That way, they can pay for just what they use. Before, when charging a flat rate, it was way too much for some publishers because all they needed was a simple ad server. It probably wasn't enough for some customers that used every feature we've written.
Then, when it comes to integration, each of the integrations is going to be priced in different ways. Most of it will end up being free. If it's ad networks or tools that are free, you go add those in the App Store. Then, with ones where you're paying for it - If it's a tool where you're going to pay that vendor - we're working out deals where, similar to the App Store on the iPhone, we take a small percentage of what you're paying to that vendor. That's how we get back our costs. Then, once you're in the Enterprise level with above 100 million impressions, everything's negotiable. It all comes down to you already having existing agreements with vendors. We're not going to try to get in the middle of that. Apple's gotten a bad rap for trying to get in the middle of existing relationships. We're not going to try to do that. It's just going to be priced individually, above that level.
How do you compete with Google?
Google is becoming the Microsoft of the ad tech industry. They're building a complete, end‑to‑end stack where you can be a "Google publisher." You're going to use AdMeld. You're going to use DoubleClick. You're going to use DoubleClick Ad Exchange. You're going to use all of their tools and you won't ever have to use anything else. Certain people and publishers are going to say, "We're a Google publisher, we use the Google Stack." They've kind of drawn a line across that Lumascapes slide with acquisitions. We definitely see that, but we think that's going to become the lowest common denominator. Publishers running just on the Google platform are not going to get the benefit of all of the new, innovative things going on out there.
What's the status of funding for the company right now? Do you plan on raising any new money in the near future?
We're going to raise money in the next six months. But we've had revenue from day one and we’ve raised money back in July, so we're pretty stable as a company. We've got eight employees. We're not yet profitable, but we've got revenue and we're pretty comfortable.
Adzerk does not run its own ad network, correct?
Correct. One of the early stakes we put down was that we're not going to sell media.
Looking out over the next 12 months, what are some milestones you’d like to achieve?
The biggest short-term milestones are happening in the first quarter of next year where we're going to really launch adOS, which is currently in beta. The public launch, where anybody can come and sign up, is one of our big milestones.
The next big milestone is adding the 200th app in the App Store. Also, we will raise another round of funding, which is always a milestone.
By John Ebbert