BRIAN BROWNIE: We have been pressing hard to make sure the buy side understands that eBay has an immense treasure trove of first-party insights that can be combined into the programmatic space so that they can buy more effectively.
We're trying to evangelize this dataset to ensure the buy side can link the very exclusive data to the eBay inventory to drive value for the advertiser. We've gone headfirst into the market and pushed this approach eBay is taking around private marketplace. How can we push up the stack, get to a first-look opportunity so we can begin discussions around guarantees? That's been a major change over the last six months. The buy side is waking up and saying, "Yes, I would love try out the data set and get into a realm where I can have this first look capability, elevate above the open-market noise and get into a guaranteed space."
We've been out talking to all the agency trading desks and top demand-side platform, evangelizing the supply source. Making sure they understand how to find eBay inventory and what it means to their bottom line. More importantly, it's [showing them] how the first-party dataset plays into this.
A perfect example is what we did around Black Friday. We lined up with a buy-side machine, a major automotive OEM to support guaranteed [buys]. We put them at the "first look" above the stack, they guaranteed they would spend, we agreed on a price and away it went.
We're all about connected commerce, we're about putting relevancy in the hands of the consumer. So we want to make sure the brand – the ad itself that's going to be delivered through this amazing set of machinery – is the right brand for our consumers at the right time. The hand-picked [inventory] availability allows buyers to leverage the data side of their equation and then couple it with what we have surfaced from a private marketplace, and benefit the consumer.
How many guaranteed deals can you do for one time block, approximately?
I would love to get to a point where I was pressed to make that decision. The good thing is we're able to partition inventory in the private marketplace scenario in an infinite amount of chunks. It all depends on what is the driver from the other side. What do they need, who do they want to speak to and then assessing the value of the opportunity. When they get hypergranular in the targeting, it's obviously tougher to fulfill.
Can you talk about using the eBay dataset to extend advertiser campaigns to sites off of eBay? What are you doing there?
We do have a buy-side piece of our business. It's the eBay Audience Platform, a team here that buys on behalf of advertisers through programmatic channels based on the eBay data. It's relatively new to the market but we're finding major traction.
The genesis of it is that historically direct advertisers have come in and consumed eBay's inventory. We recognized we needed to extent this, and the efficiencies based on the dataset and performance we're able to drive has driven healthy results for marketers and advertisers. More importantly, we leverage this tool to buy on behalf of eBay's marketing, so we have a lot of experience in how to do it.
What tools support the eBay Audience Platform? Is it home-grown tech?
The data side, which is extremely important to us, is based on our home-grown DMP. We have a wealth of information and from a consumer privacy standpoint we need to be in full control of that. Therefore we built our own toolset, which is essentially a DMP and the capabilities to plug into whatever tools we need to execute buys.
What technology solutions do you use to manage yield?
We leverage PubMatic as our supply-side platform for on-eBay inventory. Working with PubMatic very closely, we've built out a series of reports. They've been a great partner.
When it comes to selling data-enhanced eBay inventory and off-eBay inventory, who are the most important sources of demand?
We do see the majority of control of dollars coming out of the trading desks: the Cadreons (Interpublic Group), Accuens (Omnicom Group) and other top guys in the market. But we have seen a recent uptick in the tier-two remarketers. In that core group you have Criteo and companies like Rocket Fuel taking a much bigger chunk of inventory out of the marketplace.
The underlying technologies that have made the majority of this possible are the DSPs: DoubleClick Bid Manager/Invite Media, Turn, MediaMath. Those guys are plugging in and making all of this happen. It all depends on who's leveraging it and how they're leveraging it.
What about advertisers buying direct? Are you seeing that?
We're having conversations, yes. It's an area that's going to grow very quickly in 2014.
At the end of the day we're trying to make sure that whoever it is we're talking to from the buy side, that it's appropriate for the consumer. More and more growth will happen in the programmatic space. It's truly about being agnostic from a supply-source standpoint to how that connection happens.
Could you talk about ad ops for a minute?
Our ad ops is quite different from traditional IO-based campaign creation from inside an ad server. We're building out private exchange pieces in conjunction with fully controlled datasets from the eBay side. These builds from the private marketplace can live forever. Our hope is that we have figured out the right dataset for the buy side, align them with this private marketplace and allow them to decide how to utilize it.
It has lessened the trafficking and has now become more closely tied to an analytics-based role, making sure everything is working smoothly and the client is happy.
We're just at the beginning stages of how the insights can be leveraged to accelerate the value for the buy side. The beauty is, the automation is there. The piping is all there. It's a matter now of how to embrace it so it works on both ends. And to do it at scale. That's the key.