Multi-Platform

When our team initially began exploring exchange-traded media two years ago or so, I had a “Lord of the Rings” strategy in mind: we were looking for one platform to rule them all.

My working assumption was that while inventory types and sources might multiply and fracture, sitting on top of these would be “the best” demand-side platform (DSP), and we’d have one entry point into all marketplaces. I thought we’d work with different partners, figure out which one meshed best with our needs, and we’d build expertise around that platform.

That may well be what happens over time, but at this point I no longer think that’s the right approach, at least not for us. The deeper our team gets into the space, the clearer it becomes to me that there is no single magic platform.

At the most basic level, the need to operate across multiple platforms comes back to access to unique, valuable inventory and/or insight. Futures markets, online video, data, and TV are all cases where value exists that wouldn’t be accessible through a core DSP.

Futures Markets

Brand.net and Legolas Media are both doing some really interesting things in this space.

Brand.net’s Media Futures Platform lets you forecast, plan, buy, and optimize brand-safe inventory. Since it has direct publisher relationships, the same inventory isn’t available on exchanges, and importantly, it guarantees predictable, smooth delivery.

Legolas Media’s platform lets buyers apply their own proprietary data sets to specific publisher inventory. Again, the key distinction from an exchange environment being the guarantee of delivery purchased in advance, but in a marketplace environment. (By the way, purely a coincidence that I started with a “LOTR” reference and then plugged a company called Legolas in the same piece.)

Video

TubeMogul is probably best known for video seeding and analytics, but now offers a video buying platform as well. Since it’s been partnering directly with publishers for years to provide video analytics, it has access to discrete inventory augmented with a rich history of performance data to inform buying and bidding strategies.

Data

While access to third-party data is widespread, Interclick’s OSM platform has a unique approach to predictive valuation of purchasable data segments, based on actual campaign performance. Plus, the vast majority of Interclick’s inventory isn’t available on exchanges. My mantra about networks is that they need to provide value above and beyond what buyers can do for themselves with existing DSP tools, and Interclick is a great example of a company that’s doing just that.

TV

Yes, TV. People often speculate as to when TV inventory will be accessible through platforms, when in fact Google TV ads give buyers access to 37 million U.S. homes in a marketplace environment, today. TV inventory is available to buy via a platform, it’s just that it’s a standalone platform. To take advantage of the opportunity, you need to do it in an independent, not consolidated way.

Those are just a few representative examples of standalone platforms that deliver value above and beyond a core DSP, and I’m sure there are many more. That’s why my current thought about the space is that buyers need to build an overall facility with the new(ish) capability of platform buying, but understand that they’ll be using that expertise across many platforms.

It’s about the skill set as much as the technology.

Last thought: I’m intentionally not tackling the subject of whether or not it makes sense for buyers to use one or many core DSPs. I think that really is a decision buyers need to make for themselves, based on their own pilots and points of view about service and technology.

The handful of companies that have built the DSP market (Turn, DataXu, MediaMath, X+1, Triggit, and Invite) are all great companies, both in terms of their products and their teams. You’d be in great hands with any or all of those guys.

Related reading

The state of video advertising: the rise of silent and targeted ads
Report: millennials are more likely to share an ad (but also to mute it)
CZLJKT_programmatic 101_Feature Image
grayscale