User Portraits and Attribution

The challenge of digital attribution solutions is that they are built on datasets that force marketers to extrapolate a conclusion based on small bits of data.

If our audiences live behind the walled gardens of Facebook, Google, Apple or some other “universal ID” solution, these services don’t play together.  As a result, audiences can’t be (accurately) pieced together after the fact.  Users are on home wifi, businesses networks, mobile devices, out shopping and increasingly on smart TVs.  An error multiplier occurs when the so-called “universal ID” is multiplied by the “people on different devices”, resulting in an attribution equation that is, at best, probabilistic and at worst, little more than wishful thinking.

It’s a connect-the-dots problem, and it hasn’t been solved online let alone between online to offline actions.  The offline-to-online onboarding problem is still treated as a black box event.  So what happens when the KPI is deterministically linking online-to-offline events?

Semcasting sees attribution differently — we believe the key is developing a deterministic location-based linkage between offline CRM records, network and mobile usage, and user’s point of engagement — all at scale.  In our approach location is framed in terms of terrestrial boundaries determined by the property assessor’s office using lot sizes and square footage at precise coordinates.  Our patented Smart Zones technology maps the ISP delivery points of home and business network providers to these lots.  We also map massive volumes of device activity and link them together in order to create a three dimensional User Portrait of people, households and businesses.

With this portrait we link together the CRM record, onboard people to multiple delivery points online and offline,  deliver counts by person matched,  link people to impressions served across multiple channels, online and store purchases, and even link physical visits to the mall.

The key to Semcasting Attribution methodology is in the compilation process where we update over 300 million device and network locations daily.  Mobile devices change IP assignment, home networks update about 5% a month, people shop in multiple locations, and they entertain and engage in all kinds of social events.  Managing updates to the data across the User Portrait results in unique levels of coverage and accuracy.

For more information on Attribution at Scale, download our SVA overview.