Antony Taylor is VP of Display Platforms at Yahoo!'s Right Media Exchange.
AT: After a recent exchange panel, an attendee said to me, "None of it. I got none of it. Why don't you guys just start with how an exchange helps a direct marketer?" Now, we're doing this every day, but I still thought the feedback was good. We may think we're on to the next big thing, but there's so much education that still needs to happen.
How is this education going to happen around display and exchanges?
When people think ad exchange, they think biddable media, the auction, opportunities for advertisers to compete for audience. But [Right Media] learned early on that the exchange was only going to grow if we made it relevant to the business problems our customers were facing. For Publishers that meant an efficient inventory distribution platform and pricing and yield management engine. For networks, that meant seamless access to supply and the tools to manage ROI effectively at scale. For advertisers that meant solving for supply fragmentation in display, managing frequency of user interactions and leveraging data across disparate supply sources and competing ad servers.
Would you say it's about performance display these days?
Getting legacy ad servers out of the way – manual arbitration and zero predictive optimization – no doubt paved the way for performance advertising. But the bigger story here is the digital shift towards audience and our industry's run at traditional advertising mediums.
Let's talk about the features of ad exchanges. Why do you think interoperability between ad exchanges is important?
Well once you understand we are at another inflection point as an industry – one that is focused on helping advertisers reach and buy audiences at scale – interoperability becomes an imperative. Advertisers need to be able to interact with consumers at optimal frequencies across different supply sources.
Now some networks believe that they can actually service 100% of an advertiser's needs. "I'm Joe Network, I'm big enough. Just give me all of your spend, and I'll find your audience." That's obviously nonsense, unless you own the internet – especially when you consider consumer web experiences.
At the same time, ad exchanges had the vision of a single technology platform for the industry. While we made incredible progress, advertiser needs are moving faster. So instead of battling over hosted supply solutions, we are ensuring that we can plug-in to an auction environment and know that our advertisers are not disadvantaged in auctions we don't host.
In this way, Right Media is now not only an operator of the largest ad exchange but also a platform for marketers. We are partnering with large advertisers and agencies that need more efficient buying platforms. All of them require access to supply that may not be hosted in Right Media Exchange and we are going to give it to them.
Do you think that DoubleClick will give them the same access to you?
It's a great question and you should pose it to them. From our perspective, we can't not be doing this for our marketers.
Is that what Yahoo! 3PI is about?
Yes. We began exploring "Yahoo! 3PI" 2.5 years ago.
This isn't widely known. Is this a new branded term?
Yes, third-party integration. You either have a lot of demand and want to plug-in to supply marketplaces regardless of platform or you have a lot of supply and want to ensure your marketers can leverage all of their insights and capabilities to drive ROI.
But let's do some myth busting. It took 2.5 years to get to beta because there aren't a lot of companies that have world-class ad decisioning capabilities that enable them to decision on a per impression basis. There are maybe 4 or 5 companies that can scale it enough to warrant their investing in real-time integrations. However, there are a lot of advertisers with data who want to map their segments into Right Media. This is very powerful as it enables audience-based buying at scale. But, it's different from the deep 3PI we talk about in real-time bidding.
Does Yahoo! offer real-time bidding (RTB) through Right Media currently?
Yes, we are beta-testing our first few partners. The challenge is scaling. It's not easy – we're really talking about drinking from the fire hose.
What is the value of RTB through Right Media in your estimation?
For companies that can do it, we are ultimately trying to drive greater campaign effectiveness (reach and ROI) which leads to greater yield for publishers.
The other value proposition is that we have reduced technology friction that has constrained advertiser and publisher performance.
Such as ad serving? How does that evolve?
When you talk about ad decisioning, prediction, unifying guaranteed and non-guaranteed marketplaces, the next generation of ad serving innovations is extremely exciting. However, the basics of legacy ad serving systems have become commodity. Period. They are important in as much as they can service the needs of advertisers and publishers. The reality is those needs have evolved in sophistication and complexity. Legacy ad servers have not innovated to meet those needs.
Imagine in the direct mail world if as a consumer you opened your mailbox and in it was ten of the same credit card offers from the same company. Imagine the waste inherent in that kind of marketing execution – to say nothing of the consumer experience and brand impact. This is the digital world marketers face today due to these legacy systems.
RTB allows us to solve for this despite fragmentation. And it paves the way for the next wave of exciting innovations in ad serving.
Let's talk about yield optimizers. What do you think of their services such as those of PubMatic, Rubicon Project, AdMeld, YieldBuild and Yieldex?
Let me start by saying the Right Media product used to be described as an airplane cockpit of functionality – enterprise level, some up-front investment, etc. The yield optimizers have attempted to dumb that down. And, I'm obviously generalizing here…
Publishers say to us: I love getting one check at the end of the month, I love feeling that my networks have to compete, more than anything I love the workflow tools that make life easier for my operations team: billing, reconciliation, discrepancies. Yield Optimizers stepped in to address that narrow but important focus.
What I am less clear about is whether they're actually optimizing yield.
What do you mean by that?
Well, pricing time-based inventory isn't easy and it doesn't just emerge from a slick user interface. We all know that frequency is a fundamental determinant of publisher yield. Managing to optimal yield across a yield curve requires careful analysis, predictive algorithms, optimal marketplace design and a world-class salesforce or competing salesforces. It also cannot be retroactive. You need to ensure you are serving impressions to the highest paying advertiser based on time-of-day, frequency, recency, etc. The architecture of ad technology – whether exchanges, yield optimizers or 1.0 ad servers – is critical to ensuring that is happening.
So, the positive is that "yield optimizers" took a lot of friction out of publisher workflow; the negative is that I am not certain that the architecture of these systems actually enables true yield optimization. Some of these companies are thinking about it the right way but some of them are not.
Have you solved the "airplane cockpit" perception regarding a high-level of complexity when integrating and managing campaigns through Right Media?
It's bold to think that a single technology company can build the best UI, reporting interfaces, best data management warehouses, best prediction for every customer segment. We are partnering with companies who we think service customer segments that are not in our focus, helping them to design products that meet the needs of their clients. We are also partnering with technology providers that we think can augment features and functions on our platform. In that way we ensure our customers are getting the best-of-breed capabilities.
At the same time, every partner is bringing more to the table today. Two years ago it was rare to have API conversations in pre-sales. Today, everybody wants to see our APIs and build on top of them. That's cool because we're totally evolving the way we think about product development. It's almost like a shared product roadmap with our customers. Some of it we'll build, some they'll build.
Finally, for the segment of the market that is critical to us they require an enterprise level product. For them, the cockpit mixed with our services is exactly right.
So, services – as in people – are critical to the success of your enterprise product?
Yes, it is for the market we want to win. As with any enterprise product, we have built a services layer to ensure our clients are getting the most from the platform. We are then leveraging partners for segments of the market we are not addressing. This allows us to apply the right focus to our target markets and also to ensure that the entire market can benefit from the exchange through our partners. This is what it means to be an open ecosystem.
Let's pretend you work at an agency and you're a strong a proponent of demand-side platforms, but you're surrounded by traditional minds, what is the principle argument you make to your executive layer about why they need to get into a demand-side platform buying strategy?
The principal argument is margin. Digital agencies have razor-thin margins and a single media trading platform offers them the efficiency to yield more margin on the buy-side.
The second argument is that advertisers need the efficiencies that supply consolidation, frequency management and data provide. Only a platform can do that and agencies are well-positioned to be that provider, whether they build or buy.
The argument against is straight-forward. Can the agency make the level of investment required? If they can, do they have the right technology and services DNA to make it work? Finally, can they engineer the kind of change management within the agency and on the client side to make this really effective?
So there are definitely challenges to making the shift. But for large agencies and even some of the smaller ones the question is when not if.
What happens to ad networks down the road?
Everyone is surprised by the stat that in the last 3-4 years we've seen an expansion in the number of ad networks. With exchanges, we removed the artificial barriers to entry of relationship and technology – we literally opened a closed market. Who did we expect would enter the open market first? The companies who are most capable of managing risk effectively! Now as agencies, advertisers and publishers evolve their own in-house capabilities, there will be increased pressure on ad networks to prove their differentiated value. So we'll either see less of them or margins will be squeezed. Or both.
I am not providing a timeline. This will ensure I am right eventually.
Are the search marketers the first step for improving liquidity with exchanges and demand-side platforms?
No. I think they're the last, in fact, which is surprising. They have all the right ingredients; actionable data and insights from their search buying (and we now know how to leverage those into display), media buying and optimization DNA, an understanding of managing ROI across marketplaces. But the focus on display has not been there. Display is a very different animal in some respects.
No, the earliest adopters emerged from the Exchange 1.0 space. These companies quickly saw the marketers' need. They said nobody is doing this for marketers, we might as well. And they've proven hugely successful at making themselves a more meaningful part of the agency media plan. It will be interesting to see how this evolves.
Follow AdExchanger.com (@adexchanger) on Twitter.