Commentary

Why Multiplatform Programmatic Buying Works

Trading desk giant Xaxis announced recently that it is taking a multiplatform approach to programmatic display ad buying. This is a powerful signal that the approach is prevailing over the single-platform alternative. But what does multiplatform mean -- and if it so advantageous for clients, why aren’t more agencies doing it right now?

Multiplatform entails participating in RTB auctions on multiple platforms (DSPs and exchanges) during a single campaign. It is achieved by reconciling into a single “dialect” the many languages of the platforms that enable real-time bids to be placed.

Multiplatform is important because platforms perform significantly differently day-by-day and month-over-month, depending upon the competitors in-market, what tactics are used, etc. Only by comparing languages like-for-like and building a historical view across the broadest number of significant platforms can you appreciate this fact. Moreover, most campaigns are multi-tactic, so you need the ability to deploy a campaign across platforms to leverage tactic-related performance differences.

advertisement

advertisement

Of equal importance is the ability to locate and influence target audiences at scale, especially if the client’s target audience is niche, or if optimization during the client’s campaign necessarily results in a reduced number of prospective audience targets on one platform (for example, when layering on a behavioral element such as “auto intender”). The way SSPs work compounds the problem of access to audiences, as inventory is dynamically shuffled in and out of the bidding environment in real time, affecting the availability of the best inventory for the campaign and the price at which it can be purchased.

It is clear that being stuck with a single DSP and its fixed SSP relationships per campaign is problematic. But why do most agencies persist with a single-platform approach? 

Simply put, the time and technical and financial resources required to unite programmatic platform languages are a substantial obstacle to going multiplatform.  Agencies would require a unified data layer to enable them to place bids across a variety of platforms within the same campaign and allow for actionable insights and platform comparisons to be produced and acted on in a straightforward and timely manner. In addition they would need to manage more continually-updating technical relationships, constantly evaluate and develop relationships with new players entering the programmatic ecosystem and recruit highly-skilled traders and analysts capable of making decisions that benefit all agency clients -- while not causing individual client costs to soar!

It's worth noting here that some agencies claim to be multiplatform -- but, in reality, although they may maintain relationships with a number of partners, they still conduct client campaigns across just one platform at a time.

Some naysaying agencies argue that multiplatform could potentially result in a client bidding against itself, with an increasing CPM as the result. But for many, many technical reasons, the chances of this are actually very low. Moreover, any tiny possibility of it occurring is more than compensated for by the benefits of winning the bid and gleaning insights from the behaviour of the target audience (and platform) against the client’s marketing objectives, both in order to immediately take proven strategies back into the larger ecosystem to achieve optimised results, and to plan for future efforts.

Forward-thinking companies are creating the systems needed to aggregate increasingly disintermediated and disparate audience pools into the largest single ocean possible, for savvy marketers to fish from. Conversely, those in denial that only multiplatform allows advertisers to be nimble enough to realise programmatic’s potential, risk drowning in today’s constantly shifting technical environment.

Next story loading loading..