Explained: The Four Models Of Programmatic (And The Pros & Cons Of Each)

Explained: The Four Models Of Programmatic (And The Pros & Cons Of Each)
SHARE
THIS



In this guest post, MediaMath’s country manager, Yun Yip (pictured below), argues there are four models emerging for how brands manage programmatic buying and casts an expert eye over the pros and cons of each…

At AdWeek last month, there were several discussions on what the agency of the future may look like, with sessions like Is the Agency of the Future Still an Agency? The brand-agency dynamic continues to be an interesting topic as the structure of the traditional media agency evolves to accommodate marketers’ increasing need to use data, technology and insights to drive outcomes.

Yun_Yip_MediaMath_Country_Manager_ANZ (1)

Central to many of these discussions is the fact that an increasing number of brands are making moves to take greater control of their programmatic buying activities. According to a 2017 US study by the Association of National Advertisers (ANA), over 35 per cent of companies said they’ve expanded their in-house programmatic media-buying capabilities and reduced the role of external agencies, up from 14 per cent in 2016, with similar surveys revealing even bigger numbers.

From where we sit (our technology is used by brands, agencies and publishers), we’ve noticed that there are four clear models emerging for how brands can manage their programmatic buying. They all have their pros and cons, depending on the long-term aims of the brand and how it’s currently set up from an organisational and talent perspective, which we’ve attempted to capture below. And we’ve given them catchy(ish) names because well, marketing……

Contract King

This is where the brand owns its technology stack and data, but the agency continues to have full control over planning and execution. It’s often a first step to in-housing and giving the brand full, transparent access to its data and analytics, (and often the brand’s analytics team leverages that data directly to run analyses). The agency maintains a traditional role in programmatic execution and, depending on the perceived strength of the agency’s campaign management skill level, campaigns are either managed by the agency or the technology provider. The brand, agency and technology provider meet for quarterly business reviews in which they share performance results and discuss testing for the next quarter.

Drawbacks: While this model provides the benefits of data ownership to the brand, it does not help raise the brand’s self-reliance over time (if that is a goal of the brand).

Happy Commune

This is a “triumvirate” model in which all parties sit at the table together and have regular points of contact.  The agency’s role is to manage week-to-week and operational activities with the demand-side platform, including reviews of performance trends and dynamics. The technology provider manages campaigns on the brand’s behalf, meets regularly with the brand about innovating and testing opportunities and then presents formal performance reviews and insights to the brand on a quarterly basis.

Drawbacks: The transition to this model can be awkward as teams figure out how to interoperate. Intentions and desired roles need to be made clear early and often.

Special Ops

This model emerges where the brand’s overall goal is to “in-house,” but staffing constraints preclude being entirely self-sufficient so it relies on its agency for outsourced operations. The agency, therefore, maintains an operations role, providing channel planning, budgeting and processing assistance, with the brand working directly with the technology provider on campaign strategy, execution and insights. In this model, the technology provider acts in a managed service capacity for campaign management and has weekly performance meetings and larger review and working sessions with the brand; the agency has separate operations-related meetings with the brand.

Drawbacks: This is the ideal model for brands that want the benefits of more in-house control but aren’t staffed for it. However, because the agency is narrowed to a pure operations role, staffing quality may drop over time as the agency focuses its strongest resources on clients seeking a broader set of capabilities.

The Lone Ranger

In this model, the brand takes full control of its addressable media technology stack, with the agency no longer playing a role.  To date, this model has primarily been used by retailers with a heavy e-commerce presence. It means that the agency is removed from addressable media execution (although often still involved in linear execution) and the brand works directly with the technology provider on campaign operations, strategy, execution and insights. The technology provider either acts in a self-service coaching or managed service capacity.  Generally, the brand wants to manage campaigns directly, but may need technology talent as an interim solution or when staffing on the brand side changes. Weekly performance meetings, along with larger quarterly and thought leadership meetings, occur between the brand and the technology provider.

Drawbacks: This requires fully staffing a team on the brand’s side and an ongoing requirement to find the best talent and training. There is also the risk of not looking at the media mix holistically.

It will be interesting to see how these four models develop, become consigned to history or are replaced by new ones over the next decade. In the meantime, focus on what frustrations and aspirations for which you need to solve while being considerate of what your team is realistically ready to take on versus not.

Latest News

How To Drive & Measure Online To Offline Sales & Sampling
  • Marketing
  • Media

How To Drive & Measure Online To Offline Sales & Sampling

In today’s complex digital world, bringing a new product to the market is not as simple as it once was. There is no norm to the way people shop – it’s unique, personal and individual. Long gone are the days when a new FMCG product simply bought it’s space in the retail shops and waited […]