MIKE YUDIN: We’re the No. 1 search partner network outside of Google and Yahoo Bing. Our clients are major brands and large marketers. We work with six out of the top 10 Google spenders.
As users become more sophisticated, they no longer rely on curated walled gardens of content fed to them by companies like AOL, my former employer. Before, everything was organized for you. You’d type a keyword into a search box and you’d get your weather, your sports or whatever. Then Google took over.
What we’re seeing now is the unbundling of search intent and we’re the only one really monetizing this at scale with data right now. There used to be a few massive walled gardens and now it’s becoming a jungle of little walled gardens.
Google is becoming more and more black box-like everyday.
Can you elaborate on that?
One day Google is shutting down tracking pixels, the next day it’s something else. That said, Google is very attractive to publishers because of its scale and because it’s relatively easy to use – but you don’t really have any power over your data. Google doesn’t want you to think that your user search data actually belongs to you.
We let our publishers access their data through an API where they can pretty much see every search query that happened on their site. Because it’s not just about making money, it’s about helping them understand their audience, trends and the quality of their traffic.
What data points are you collecting on behalf of clients?
We connect search intent with ads from our direct advertisers. More than 90% of our advertisers have our conversion tracking pixel installed, which means we get data on clicks, purchases, ROI, etc. We track all of these events – say if a user scheduled a test drive with a dealer or filled out a form on an insurance site – and use the data to make predictions about relevance and price. Each of our ads is priced differently.
Yes, our advertisers buy adMarketplace traffic for performance – we need eyeballs – but data insights are equally valuable to them. Obviously Google has scale, but they’re not sharing data, and any smart marketer understands the value of going beyond the immediate buy.
Can you share an example?
We have a client in the home improvement category which has stores all across the US, as well as an ecommerce presence. We showed them all of the data related to their campaign goals, of course, but we also shared a heat map of their sales by state and region. Turns out they had way more sales in rural areas and in particular states.
The client was really interested because the heat map correlated exactly with a map of their physical store locations, which meant that in rural areas where there are fewer stories, there was much more ecommerce going on. They found that bit of information extremely valuable.
How do you approach targeting?
Most of the companies on ad exchanges are traditional display companies and they all trade in cookies. When you think about what sort of data is being traded on, it’s very commoditized. Say I go to Amazon and search for a big-screen TV. Awesome. Now I’m in a pool of cookies and as soon as I go to another site and an impression is posted in an exchange, there’s a feeding frenzy. There’s very little that’s proprietary about this, so the only differentiator for display companies is CRM data or a smarter algorithm because they’re all working with the same data pool.
Our data pool is different. We see search intent data for 100 million users in the US. We look at longer-term trends in search history without having to rely on PII.
How does that work?
Say I’m using a local search app like YP And I’m looking for pizza in New York or I’m looking for travel information about Arizona. We’ll capture that search intent, identify when that user opens another app and serve an ad based on the search.
How is that different from basic retargeting?
We see deeper trends of user intent, which really depends on what a person is searching for. If you search for pizza, you want it now. If you’re looking for a mattress, you’re probably going to be in the market for at least a couple of days.
We’ve done analysis on this. Instead of guessing, we see the hard data on relevancy because we get the conversion data.
What’s on the road map?
We’re doing a lot of R&D and experimental work. For example, we’re starting to look and see how often certain devices appear on certain networks – work, home, people who are frequent travelers, etc. – so we can do cross-device targeting. The present is about scaling what we have and tapping into the power of search data.