Skip to navigation Skip to main content

Ep 189: Inside DraftKings: Mobile ad attribution

Richard Eiseman is the Marketing Tech Operations Specialist at DraftKings, one of the world’s biggest fantasy sports betting companies. Richard focuses on ad tech operations including, tracking, attribution, and driving DraftKings’ privacy attribution strategy. In this episode, Richard shares his perspective on the current landscape of ad attribution. He touches on Apple’s AdAttributionKit, learnings from early tests of SKAN 4, the possible end of fingerprinting, and a whole lot more.

Listen and subscribe:

Listen on Apple podcasts
Listen on Spotify

Questions Richard answered in this episode:

  • What is your role at DraftKings?
  • How is the current landscape of ad attribution?
  • How are you staying on top of all the changes with ad attribution?
  • Why do marketers need to account for the differences between various types of ad attribution?
  • What are the main differences between Apple’s AdAttributionKit and SKAN?
  • What high-level learnings can you share about your tests with the early versions of SKAN?
  • How do you work with your Mobile Measurement Partner?
  • What’s your take on Apple’s Privacy Manifest announcing it would stop fingerprinting?
  • Is your job easier on the Google side of attribution with their GIDs?
  • What do you think ad attribution will look like in five to ten years?

Timestamp:

  • 0:33 Richard’s role at DraftKings
  • 3:03 Current landscape of ad attribution
  • 4:35 Testing incremental attribution at DraftKings
  • 7:42 What is Apple’s AdAttributionKit?
  • 10:30 How is it different from SKAN?
  • 13:30 Learnings from early tests of SKAN 4.0
  • 16:43 How do you work with your MMP?
  • 17:23 Will Apple remove fingerprinting?
  • 19:50 Google’s marketer-friendly approach to attribution
  • 22:02 Advice for staying on top of the latest changes
  • 24:40 What to do in NYC

Quotes

(5:38-6:03) “We have our own incrementality testing method at DraftKings where we try to weigh the actual output of everything – not from taking SKAN or Android-deterministic data or web data at the base read-out, but by really trying to measure what percent of the initially reported conversions or KPIs occurred based on that advertising alone.”