Difference between revisions of "Pelican"

From Bitnami MediaWiki
Jump to navigation Jump to search
Line 20: Line 20:
<references />
<references />
[[Category:Privacy Sandbox|Turtledove]]
[[Category:Privacy Sandbox|Pelican]]

Latest revision as of 23:36, 23 January 2021

Neustar's Private Learning and Inteference for Causal Attribution (Pelican) provides key market requirements for causal attribution of sequences of exposure across decentralized publishers.[1]

The focus of Pelican is to ensure marketers can continue to provide multi-touch credit for publishers, given other proposals would only support last-click attribution. See Private Click Attribution.

Causal measurement makes it possible to say how effective advertising is at changing user behavior.[2] Operationalizing causal measurement in practice requires not only measuring marketing's impact on a consumer's decision to purchase, but also quantifying the consumer's inherent propensity to convert. [3]


By improving marketers' ability to attribute credit to more publishers, they will allocate more spend with those publishers.

Pelican lists impacts to marketers if attribution is impaired:

  • Impairing accuracy in attribution will likely lead to lower spend in digital channels overall.
  • Impairing accuracy in attribution will disproportionally impact smaller publishers who do not have the ability to leverage large amounts of first party data about their users.
  • Impairing accuracy in attribution will shift spend to digital channels that are not impacted by interference with cross-publisher identifiers (such as Search)
  • Impairing accuracy in attribution will cause marketers to allocate their spend less efficiently, which will increase waste and hence likely result in higher customer acquisition costs, and may in turn be passed on to consumers in the form of higher prices.


  1. https://github.com/neustar/pelican
  2. Singh, K., Vaver, J., Little, R., Fan, R. 2018. "Attribution Model Evaluation." https://research.google/pubs/pub46901/.
  3. Kelly, J., Vaver, J., Koehler, J. 2018. "A Causal Framework for Digital Attribution." https://research.google/pubs/pub46905/.