It’s commonplace for planning and reporting to sit in separate teams; the former in strategy & planning, the latter in data & analytics. This is understandable from an organisational perspective due to the contrasting skill sets required to be effective in each role. However, this separation can mask the intrinsic link between these disciplines, leading to a disconnect that reduces the value of the output of both, particularly planning.

Two sides of the same coin

To make this link explicit, let’s think about what those outputs are.

The key output of a media planning team is, unsurprisingly, a media plan; an articulation of what is going to run, the cost and expected results. At a minimum it should take into account commercial objectives and insights to arrive at tactical recommendations aligned to a broader strategic framework.

The key output of a reporting team is, again unsurprisingly, a performance report; a visual representation of the results being driven by activity that is live or has been previously. Generally they’ll be created via a dashboarding service, such as Looker Studio, and need a robust data collection process to ensure accuracy.

Put simply, media plans show what we’re going to do and the predicted results, and performance reports show us what we’ve done and the actual results. They’re two sides of the same coin; the more closely we align them, the more effectively we’ll identify whether or not we are performing as expected.

The need for a modernised approach

Traditionally, plans were created for offline media, where the feedback loop for measurement was longer and the ability to change activity ‘in-flight’ far more limited. This meant that so-called ‘planning cycles’ were stretched over months, with the initial media plan remaining static throughout, ahead of a new plan being created for the next cycle based on results from econometrics studies or similar. Whilst making sense within an offline context, the feedback loop for measurement in online media is much shorter, and our ability to optimise live activity has greatly increased. 

However, planning has been slow to catch up with this new reality; static plans remain the norm, even when activated digitally. The issue with this is that delivery teams working in online platforms will quickly start to optimise activity based on what they’re seeing; moving spend to where performance is strongest. This means within days or weeks, the static plan produced pre-launch will be outdated and its usefulness will decrease; at best it serves as a snapshot of what we thought would happen before  real-world data was generated through activity. 

Incorporating performance data into your media plan

Tackling this means reconceptualising media plans and their purpose. Instead of thinking about them as static – reviewed and updated every few months – we should think of them as hypotheses to be regularly refined.

In other words, a media plan is the best answer we have, given information available to us, for how to use media to achieve our goals. But, as with any hypothesis, this answer will be flawed. No matter how robust your planning, some assumptions will be proved wrong once activity goes live. The challenge and opportunity is to identify flaws as quickly as possible, learn from them, and optimise. 

This notion of generating a hypothesis, testing it and refining based on results, isn’t new. However, its application to digital media has been decidedly cumbersome, thanks to the misalignment of media plans and performance reports. Typically these will be separate documents with differing formats, meaning that refinement based on performance is manual and subjective based on an individual’s observations.

We can improve the speed and consistency of this process by connecting planning and reporting. To do this, our plans need to become data sources for our performance reports. In other words, our reporting software needs to read plans and visualise them in a way that is directly comparable to performance data, meaning the reports will highlight deviations from our hypothesis without the need for manual cross-referencing. 

Next is to incorporate performance data into media plans, via a tool like Supermetrics or similar. Once in there, the possibilities are endless but ideally you want your plan configured so that it outputs predicted performance based on emerging trends in your live data. Provided your forecasting model is robust, these predictions should become more accurate as your performance generates more data, meaning your ability to plan accurately improves over time.

Investing in an automated solution

Once you’ve taken the first step and integrated your plans and performance reports, you can improve this further by investing in a solution that can identify and apply recommended optimisations quickly and at scale. By collecting all required performance data and housing it within a platform such as Google’s BigQuery, you’ll ensure a consistency that allows you to apply sophisticated measurement methodologies more easily. You can then push these findings into the relevant media platforms via their APIs, before observing the impact those changes have on your performance and kicking off the feedback loop once again.

So, with the ability to automate this at your fingertips, how connected is your approach?

This article was originally featured on The Drum.