How Orchestra measured true effect without third-party cookies
Orchestra is a Danish full-service media agency, which - since 2013 - has helped a large number of small, medium and large Danish brands. Orchestra aims to stay transparent about their choice of methodology and is not afraid to challenge traditional media industry conventions.
Orchestra wanted to measure the effect of a large campaign for an entertainment company by documenting how many conversions could be attributed to the campaign and thus what the Return On Advertising Spend (ROAS) of the campaign was.
However, measuring objectives like conversions in today’s fragmented media landscape is impossible if relying solely on third-party cookie technology. There’s an abundance of media channels and devices providing their own set of identifiers (or none at all). This along with the rise of walled gardens, ad blockers, ITP and regulations such as GDPR and CCPA have made it challenging to tie cause and effect across channels.
By utilising AudienceProject’s cookie-less attribution model, Orchestra was able to cope with the challenges and successfully measure the effect of the entertainment company’s campaign without the use of third-party cookies.
AudienceProject has introduced a new and better way of measuring effect and attribution at scale – an intelligent model utilising geography to determine cause and effect by comparing groups of people rather than tracking individuals through cookies.
Please note: You can learn more about our cookie-less and privacy-safe attribution model in this article
Not only was Orchestra able to document a 2.5x ROAS of the entertainment company’s campaign. By utilising AudienceProject’s new geo-based model, they were also able to properly document the effect attributed in a uniform manner across channels.
Words from our client
This new attribution model is the tool we as an agency have been searching for, as it helps us document effect across different media platforms at scale and eliminates the problems we have been experiencing with typical first / last-click attribution models in a lot of common platforms.