As I just joined The House of Communication a few weeks ago, let me briefly introduce myself. I started my career in a media agency nearly 17 years ago as a digital planner…
And fell in love with this job because it matches with what I am: a MathMan, a strategic thinker (for both my clients and my company) and passionate in mentoring young and less young people.
About a decade ago, lots of advertisers were doing massive campaigns on broad audiences with one single message.
Then we started to use targetable media
(In Belgium with Microsoft 15 years ago before programmatic and social) and some advertisers wanted to run campaigns on hyper-targeted audiences.
As demonstrated by the academic world and maybe also by your sales figures, most of the products like FMCG, retail banking, insurances or health services still need to be advertised on broad audiences because the highest growth potential is on new customers recruitment (and not trying to sell more to your biggest customers).
So, we went from broad to hyper targeted, and now we are in the era of hyper-targeted broad reach: smart advertisers use specific messages on hyper targeted audiences but the global campaign is a sum of all these clusters (1 audience – 1 message). One of the main purposes is to learn which audience(s), message(s) or platform(s) provide the best output to maximize your campaigns and to nurture your marketing know-how (on content, call-to-action, creatives, audiences)
To evaluate the output we need first to set KPIs. There are two main considerations: first, optimize your campaign on ONE MAIN KPI and add max two other soft KPIs (eg: % reach is main KPI and used to optimize media plan whereas cost per full view and engagement rate main be soft KPIs, just “FYI”). Second consideration: for a single goal, there are marketing KPIs and media KPI. Both of them can lead you to very different conclusions.
Let’s take an example: your campaign goal is full views of your video:
In order to get clear learnings from your campaign, I have two other recommendations: test maximum 2 variables at a time. For example: test different audiences and creations but use only one platform and one format to avoid bias from them. And finally make sure you have enough and comparable volume of impressions/reach for each cluster (eg: 1 audience + 1 message = 1 cluster).
The final benefit of the approach is to get better performances for same or even smaller budget. So we all need speed and flexibility when analyzing KPIs, creating new messages to test, optimizing targeting and switching platforms. We shaped the House of Communication with that spirit: one unified circle of experts in data, media, creation and production around each client.