Omar Tawakol is CEO of BlueKai, an ad technology company.
As part of its “State of…” series of articles with industry executives, AdExchanger.com spoke with Tawakol late last month to discuss his company, his views on the space, and the state of BlueKai today.
Click below or scroll for more:
- BlueKai In The Past Year
- Data Exchange Strategy
- First-Party Vs. Third-Party Data Trends
- Cross-Platform And Clients
- Apps, Marketing Stacks
AdExchanger: What are some of the differences between the BlueKai of today and a year ago?
OT: So, early on in our evolution as a business, we were helping the media ecosystem get access to good data so that they could start doing audience targeting. It was all about access to data assets that they didn’t have before. Over the last year, what we’ve noticed is that people have become more mature. They started getting access to different data assets and the focus shifted towards, “Give me a platform that would allow me to create my own proprietary advantage. Let me get my own first‑party data, figure it out, layer on external assets I don’t have, and then make it incredibly useful ‑ drive ROI, make it actionable, integrate it in every platform I need to execute on.”
So that’s the major shift. The data buyers expanded their scope – from testing and experimenting to, “How do I now create a proprietary advantage?” And the answer has been, “I need my own DMP.”
And where has this left the exchange strategy which I think BlueKai was initially known for?
It’s been very interesting. What we’ve noticed is that people who have DMPs end up buying more data. So it’s very helpful to the exchange business. However, we run the two businesses entirely separately ‑‑ meaning, if you buy our DMP, you don’t have to buy a single bit of data from us. And anything that goes in your DMP stays in the DMP, meaning it’s only one‑way access. Data from the exchange makes it into the DMP. Data from the DMP never touches the exchange. They’re run, sold and priced separately. The digital data rights and contracts are separate.
I would almost think of this as similar to Android. Android has an app store, and many more people want [to work with] the Android platform than Blackberry. It’s an open platform. Platform marketplace combo business models are very powerful.
The pendulum is swinging. When we first started the business and talked about the data exchange and access to intent data, it was sexy and interesting for people. Over time, people became more used to third‑party data and, frankly, even confused by the myriad of options that we presented to them in the exchange. And so the pendulum swung to the other side to, “Well, let me understand my first‑party data first.” The reason is that when they understand their first‑party data, they can overlay third‑party attributes on it and better understand what they have and what they are missing.
So, interest has been so strong over the last year, that if you just said the phrase “first‑party data,” you could meet with almost anybody.
And so I think the pendulum will be swinging back, as people realize that there’s a harmony between these two things, and the whole point of a data management platform is to sort through it all.
What’s the next disruption in the data world?
Well, the first disruption that has occurred is data has moved to the edges – meaning it started out in the realm of the ad networks, and then it ended up becoming usable for the trading desks, agencies and DSPs. And as of late, it’s moved to the land of the marketer and the publisher, where they’re really getting savvy on its use.
But, there is one disruption that has not yet occurred in the data world and that nobody is prepared for. That is, “What happens when it moves to the land of the individual consumer?” Where the individual consumer says, “It’s mine ‑ whether it’s on this device or on 72 different platforms, it’s mine.” That’s going to be interesting.
Looking at the “cross-platform Holy Grail,” is cross‑platform usage of data possible? – as in audience buying across channels?
So, one of the biggest trends we’re seeing in this second year of the data management platform business is that people are starting to move from a myopic view to an enterprise view. The myopic view says, “I’m going to retarget my data from my website via RTB, and through a DSP, and that’s it.” And that’s a good starting point. However, the way we see it evolving is into several layers of expansion. Expansion, number one, is ‑‑ well, not just an RTB DSP. It could be, “I buy a lot of ads on premium publishers who aren’t even putting their inventory in an RTB [environment]. Let me also apply my data segments there. Let me apply it to ad networks that I am used to buying from. Let me apply it to my video ad networks, which are different partners with different platforms. Let me now take it outside of the realm of advertising, and let me put it into my own site optimization, so that I can change the content of my site, and drive better activity and personalization on my site” and so on.
That’s what we are calling the enterprise DMP – where a marketer moves from a single application of the data, which looks like fancy retargeting, to multiple uses, sometimes across different departments in your company, and across different channels and touchpoints. You’re starting to see the marketer have this concept of the enterprise data management platform. But, it’s going to take several years to unfold.
What are the challenges on the client side right now with a DMP strategy?
The biggest challenge is getting all these corporate groups on the same page and consolidate marketing’s use across many, many different groups within the company, and eventually across [geographies].
When you think about the enterprise DMP, nobody can step immediately to the enterprise DMP. What they need to do is paint the vision, “Hey, there are fantastic advantages to segment once across all our available data, and then apply it everywhere, across all our execution platforms.” That’s the vision. But in order to execute that vision, you really need to start with a solid use case. So some people start with a use case of retargeting. Some people actually start with tag management. Some people start with third‑party buying. Some people start with site optimization, and other people start with audience analytics, which is just applying data to all my running media, even if it’s not data‑targeted, to figure out how my media is performing.
And so, what we find is by allowing people to choose the use case they want, and nail that initial use case, and then roll it out, we succeed.
It’s too much for marketers to think they’re going to integrate 50 different point solutions in the market.
Regarding “apps” strategies, do you anticipate BlueKai coming out with its own apps?
Yes. That is something we’re working on very actively. We currently have two companies we’re working with today who already do look‑alike modeling on top of our platform. We’re going to be announcing one of those soon (Announced 4/10 – DataLogix and BlueKai). And the vision we had there was, instead of BlueKai being the be‑all and end‑all, where we do every application, we opened up our platform for other people to run their applications on top of us. We’re starting to do that with analytic providers, where several analytic providers can run pieces of their business on top of our platform. We’re going to be taking the same route with attribution where we have some rudimentary capabilities in BlueKai’s applications, but instead of us saying you have to stay with us, we’ll open up our platform for other companies with great solutions to run on top of us. This is an important strategy.
We believe that the data doubling rates of our industry are as aggressive as what you saw in Moore’s Law. Rather than convince the world that we have the only look‑alike models, we have opened up our platform for other good modelers to produce look‑alikes, and then put those look‑alikes back into the platform as if it was just another data stream.
So it allows us to add up new algorithm providers quickly and scalably, so that the market can almost turn algorithms into data. This leads to the question, “Are more data or more algorithms better?”
We’re saying that you don’t have to choose. Have more data, and then plug in other people’s algorithms.
Do you consider yourselves, at some point, providing an end‑to‑end solution, a marketing stack?
We definitely see ourselves as an end‑to‑end data stack for the CMO, and believe that you want to keep your data stack completely separate from your execution platforms for a very simple reason. No big enterprise, across all of its departments and countries is going to have one company do every bit of mobile, video, social, search, CRM, site optimization and so on ‑ no company uses one vendor tor run all their execution platforms.
They choose different best‑of‑breed partners in these different channels, different countries, different use cases. And so instead of fragmenting your data across all these disparate execution platforms , we’re saying that we centralize all the marketer’s data into one place and then drive segmentation across all of the execution platforms.
Follow BlueKai (@BlueKai) and AdExchanger.com (@adexchanger) on Twitter.
