An advertising full stack solution integrates different components of advertising technologies, such as analytics, creative, media activation, data management and measurement platforms, under a single ecosystem. The result is easy interoperability and high data fidelity between components.
Who are the current ad tech “Full Stack” players?
Full stack is a term used differently by various tech companies and therefore varies in definition. But the main players in the current ecosystem with full stack offerings are Adobe, Google and Adform from a buy side perspective. There are others, such as Appnexus, who have a similar full stack offering on the publisher side of ad tech.
Benefits of integrating technology
The most important benefits of integrating technologies are the data synchronisation between the disparate set of platforms, and a continuous feedback loop into the media delivery systems to fine-tune the media efficiency.
Most DMPs exhibit a 40-50% data loss when matching cookies to a 3rd party system. This is because the cookie universe between the two systems does not always coincide and there are many users that are not matched between the two systems correctly. By bringing these systems within the same ecosystem, an advertiser can eliminate these losses and use their data management / analytics tools to the maximum.
For example, when an analytics tool is connected to a DSP within the ecosystem, advertisers can create multiple segments based on the user behaviour on their website and push these segments to the DSP within the stack. There, the DSP can expand the audience segments based on user characteristics and has a much bigger seed audience to play with due to negligible data loss.
This allows for more accurate audience archetypes and greater scale. Google Analytics is a good example of this connection. Similarly, Adobe’s integration of their Analytics cloud with their Ad Cloud will provide similar benefits.
In other cases, integrating creative management platforms with the DSP within the same stack allow for the DSP to feed the information into the creative tool / rich media adserver to serve richer, more meaningful ads to the end user.
Lastly, a full stack implementation across adserver to DMP to DSP reduces discrepancies in reporting and lesser overheads with setup and implementation for agency ad ops.
All in all, a full stack solution benefits the advertiser by providing a higher degree of data fidelity between the modular systems and a quicker feedback mechanism to refine campaigns based on creative or audience targeting optimisations.
Limitations of a full stack
The industry has been facing up to the challenge of rising number of walled gardens recently. Most of the inventory and data providers have started to aggregate their data and inventory within their walled gardens, which are ably supported by the full stack solution built around them.
For example, Facebook inventory is available within Adobe’s DSP with the ability to frequency cap on non-Facebook properties using Facebook Audience network. Youtube and Google properties can be frequency capped using DBM. Amazon data resides within AAP.
Therefore, in a fragmented ecosystem, agencies are forced to use more than one walled garden / full stack solution which automatically negates the benefits of the full stack described above. A full stack works very well within its boundaries, but is less interoperable with other full stack solutions.
The direct impact of this is the inability to frequency cap across different stacks. In some cases, even within a stack like Google, some properties like Youtube are bought on user ids while rest are bought on cookie ids, which makes it impossible to frequency cap within the same ecosystem.
Additionally, a full stack data management system will not allow the fusion of data from multiple sources to allow the advertiser to have a holistic view of their data. For this, they will require a 3rd party enterprise grade DMP that is able to ingest and export data across multiple technology vendors.
Recommended approach to using full stacks
As mentioned, the current rate of fragmentation within the ecosystem is predicted to continue, with major publishers like Google, Facebook, Yahoo and Amazon consolidating their data and inventory within their own ecosystem. A few points to note here:
A single stack should not be – and cannot be – regarded as a holistic solution for all of an advertiser’s needs. There are many tactical levers opportunities to leverage across all stacks which should be maximised e.g re-targeting search and analytics segments, or connecting creative execution to data management platform. But these solutions are specifically designed to support specific campaign objectives. For different objectives, multiple stacks may be used to support a single campaign.
Even if one singular vendor has a healthy reach within a market, the limited availability of inventory will always negatively impact the cost of media paid to gain incremental reach beyond a certain threshold.
Therefore, for an advertiser with multiple and complex KPIs – which is the case with almost all savvy advertisers, access to data and wider inventory across these stacks is crucial to deliver on performance and costs.
The recommended approach is to continue with an open architecture approach to selecting vendors. Mediabrands is making a strong push towards integration and consolidation of these stacks within the agency, so the advertisers can use the best mix of DSP vendors bespoke to each campaign / KPI.
To do this, we combine historical information from continuous DSP evaluation and combining it with machine learning within our tools to deliver real time optimisation across walled gardens.