A reputed client in the marketing space had substantial limitations in their market intelligence. Users consume content across multiple devices, meaning getting an accurate picture of who is viewing what can be difficult – with no attribution system, our client was unable to accurately model performance or map audience characteristics. Due to limited and inaccurate intelligence, investment into new marketing strategies was not translating into performance – wasting funds and damaging our client’s relationships with their customers. An accurate device attribution system was sorely needed to get performance back on track, with the added challenge of guaranteeing user privacy
Our initial priority was to explore our client’s existing data streams and map the cloud system supporting their ETL. Our experts worked with our client’s specialist R&D group, designing novel methods to build multi-touch attribution and guarantee GDPR/CCPA compliance. Two strong options were presented to our client’s leadership team, who chose to work with server-side pixel tracking (SSP) and a bespoke algorithm set – all running on AWS. First, a compliance layer leveraging a range of AWS tools was implemented to maintain GDPR/CCPA compliance, and all architecture was governed by these strict cloud protocols. The next step was to establish effective web beaconing via AWS Lambda, then feed anonymised data to MySQL database storage using AWS Kinesis. The final component was a series of custom Python algorithms able to attribute device-IDs to individual user profiles and connect these to anonymised audience profiles – delivering accurate metrics in near real-time
