Programmatic Blog

Measurement challenges will continue in 2019, why is this and what can you do about it?

Written by Chris Cox | Feb 5, 2019 9:57:16 AM

Digital industry changes are ripping up all that we know about how to measure. Our existing approaches are becoming less and less reliable but having the right principles in place can allow for reliable and insightful measurement.

The challenges of accurate measurement were brought into stark relief in 2018, with a multitude of systemic shocks, each making it harder to create a holistic picture. Key drivers of this included:

  • Large scale changes to user privacy in the online space, driven by GDPR and ITP 2.0 made it harder to track users over time, and across devices.
  • Google and Facebook have largely closed their doors to the independent measurement of their activity or limited it to only a few preferred partners. In the case of Google, they have further restricted access to user level data, and in the case of Facebook they have never offered this level of access.

2019 is likely to bring more of the same, with the potential for several more disruptive events, most notably the possibility of ITP 2.0 being applied to Chrome. However, there are also some potential bright spots, such as the full launch of Ads Data Hub, a product that will allow user level analytics again on Google properties.

For brands to take advantage of this however will mean bringing together a wide array of different solutions and be able to manage the sometimes-contradictory data signals this might produce. Specifically, brands must have a structure in place which encompasses:

  1. A top down perspective: Budget setting across channels, many of which will be offline, must be subject to a central source of truth. Due to the nature of the channels involved this will by necessity be a modelled approach, typically a Marketing Mix Modelling approach. Usually, this is used to make key decisions on a quarterly or twice-yearly basis.
  2. A cross-channel understanding: Again, due to the nature of the available data, any understanding of the interplay between channels will come from a modelled approach, mixing deterministic and probabilistic data. This will allow for more nuanced budget shifts and some campaign optimisation at a more-brisk cadence, for example monthly.
  3. Channel led optimisation: Day to day decision making should be done at the channel level, relying on the best measurement approach available within that channel. Due to the nature of the walled gardens this will typically need to use their source of truth. These types of optimisations can be approached almost daily.

Bringing these together, linked through a business orientated KPI is the gold standard going into 2019.

Chris Cox, Client Analytics Lead, Infectious Media


This article is part of the IM programmatic trends 2019 series, read more here