Danish Radio (DR) Concert Hall is a leading concert venue in Copenhagen.
DR Concert Hall is outsourcing their SoMe campaigns to an external marketing agency, they do a stellar job when it comes to promoting an event that has been sold before.
They do it through Facebook algorithms that enable them to find the right audience, through groupings on who goes to concerts, geography, genre, etc. Meta is unbeatable when it comes to finding these segments on Facebook and Instagram.
So why is this case even here? While the above is fantastic when you have history of promoting a given event that can continually be optimized, there are shortcomings…
How do you find the right audience for a NEW event, that has no previous selling history? Meta can’t take previous customer responses, purchases, tracking, etc. into account.
And by not doing so, makes it hard to find the right target audience for a new event with no history.
This gave Allyy and DR concert hall an idea.
When you take on an event that has never been sold before, there’s risk associated with it. For one, there are no guarantees that it will sell – another thing is that you don’t know who in your target audience it is relevant to, and how to push it through SoMe.
The pitfall of only using segmentation and general logic to find the right audience for a new event is that there are cosmic amounts of data available from other concerts; interactions through web and email. The Data has enormous value but isn’t being utilized to predict the perfect audience for new events.
This means that the approach ends up being handcrafted with human logic as a starting point. This doesn’t sound that bad – but we thrive by doing even better!
While DR Concert Hall was pleased with the results they achieved and easily could have stayed in the status quo, they also saw this as an opportunity.
Would it be possible to pair two unbeaten algorithms, and thereby use the data that is already being modeled and analyzed by Allyy.io, to find a red thread applicable to Meta?
DR Concert Hall was already utilizing Allyy.io’s predictions that make it possible to individualize the experience for each customer, whether it was through email or a website.
So instead of hitting a generic target group for new events without history, it would be possible to find the perfect audience based on own customer data, running through Allyy.io and then applied to and through Meta?
Together we set out to test the hypothesis.
Based on existing data from their CRM system, old event texts/pictures, and responses to previous SoMe campaigns, Allyy.io found the right audience with the highest likelihood of purchasing new events.
DR Concert Hall used Allyys predictions to find lookalikes on Facebook and Instagram for higher precision than human logic can accomplish consistently. Overall this resulted in an astonishing:
3x in conversion
+40% in click-through rate
SoMe budget was scaled down by -31%.