By David Brudenell
Opening to a full and attentive audience in the main keynote hall, Sean Bruich and Ruth Krichelo from Facebook brought a supplier perspective of how Facebook used ‘Big Data’ to help its advertisers come closer to the measurement divide between social impressions and clicks with real-world sales lift.
Bruich opened the presentation by directly addressing the two main insights that Facebook is focusing on for their clients:
• Making social advertising relevant to Offline sales – to be relevant you have to fit into the attribution process
• The road to success is more important than if the campaign was successful – what did we do to make the campaign process better over time?
Ruth built on Bruich’s intermediary point that ‘impressions matter’ and that they may be more important than clicks by framing the argument around the 700MM monthly mobile Facebook users with the typical longer-term sales cycle for Telco products.
Ruth explained the closed-loop control/exposed ROI methodology that Facebook used:
1. Privacy Safe data collection – data including, but not limited to mobile ownership, location data, device data and other forms
2. Propensity Score Matching – ensuring that the control group matches the exposed group to identify the incremental sales lift that Facebook added.
Insights from this experiment showed that younger consumers are less loyal and more response than others, indicating that young consumer advertising will provide the Telco with a higher campaign lift. Furthermore, within Telco and Facebook, mobile ads were five times (5x) more effective than similar ads delivered within a Facebook accessed through a browser.