Accurate Attribution with Rauxa’s VP of Data Science Izzet Agoren

Izzet Agoren knows:  Accurate attribution models cannot be built by data scientists alone…you need a team who understands how media operates in the real world.  An electrical engineer who launched his own ISP in 2002 in Cypress,  Izzet search for the signal in the noise led him to AdTech.  As the VP of Data Science for Rauxa, hear how he’s overcome the barriers to better measurement.

 

Guest bio:

Izzet is a technically trained engineer who found his way to ad tech and data science. His vast contributions include the VAST 4.0 and VAST 4.1 standards with the IAB Tech Lab, he has served on the Mobile Marketing Associations Messaging and Programmatic Committees, IoT Council; and the Internet Advertising Bureau's Programmatic and Data councils.

 

While at Penn State, he co-authored multiple, peer-reviewed publications in the area of real-time video communications for 4G wireless communications and his work at Motorola lead to focused development of efficiencies in IEFT standards that defined VOIP on packet networks for mobile carriers.

 

He founded Extended Broadband in 2002, a regional fixed wireless internet service provider that spanned four countries — all of which had strained diplomatic and political circumstances. In 2007 Izzet joined a team that won four media awards for a Semantic targeting technology in marketing technologies. Izzet’s further engagements with Integral Ad Science and TRUSTe exposed his technical background to the enabling capabilities that verification and privacy provide the digital marketing and advertising landscape.

 

He is currently serving as the VP of Data Intelligence at Rauxa, where he leads Artificial Intelligence and machine learning product development.

 

He is an elected Senior Member of the Institute of Electrical and Electronics Engineers, holds a Master’s degree in Electrical Engineering, and is a Fulbright Scholar.

 

 

 

 

 

 

 

 

Key takeaways:

[2:00] I introduce today’s guest, Izzet Agoren, and ask him about how he came to be where he is today.

 

[5:24] Izzet remembers when he launched his own ISP in Cypress in 2002, climbing on roofs and boosting microwave signals with antennas and amplifiers.

 

[9:12] Moving into his position at Rauxa, Izzet was tasked with elevating the function of the department from being descriptive to prescriptive with data.

 

[12:17] Izzet and I unpack the concept of viewability and the issues surrounding it — a lot of ads aren’t seen but models don’t take it into account — most of the models are made by data scientists who have zero media experience and the model outputs don’t make sense.

 

[15:53] Izzet shares his steps to making a business, a department, or a team more proactive: it always begins as a philosophical concept.  He also touches on the two ways data & better measurement can be proactivity presented to a client: (i) as a media product or (ii) as a data product.

 

[18:56] In some cases, clients know that they’re missing out by not using attribution. They are unable to pursue this ROI either because they’re not ready or because there’s a perception that it’s either too complex or too expensive to deploy or that there will be political barriers.

 

[20:02] Measurement is ‘table stakes’ to buying media. Izzet has been kicked out of meetings for saying so.

 

[21:37] Izzet hopes that embracing attribution won’t take a new generation of people as the alternative is that the walled gardens will take it on. In the end, this acceptance movement should begin with more marketers bringing this function in-house.

 

[26:56] In terms of the future, Izzet believes that attribution should get progressively easier as offline behaviors and channels become digitized. Data points will become deterministic and make the analysis more accurate, near real-time.

 

[29:52] Is having complete data possible?

 

[32:04] Izzet shares one thing he knows that no one else does — he doesn’t profess to know any more than anyone, but he does share a recent discovery of his. I thank him for coming on the podcast and sharing so much of his experience.

 

Be sure to tune in for the next episode and thanks for listening!

 

 

Connect with our guest:

Izzet Agoren on LinkedIn

Izzet Agoren on Twitter

 

Mentioned in this episode:

Book: Causality: Models, Reasoning and Inference, by Judea Pearl

PegasusCRM

The Shapley Value

Markov Models

The Apache Foundation

Apache Atlas

Apache Delta (now retired)

 

About your host:

Jeff Greenfield is the Co-Founder and Chief Attribution Officer of C3 Metrics. As the chief architect of the platform, Greenfield worked directly with the former CEO and Chairmen of Nielsen to solve advertising’s Attribution problem.

 

Greenfield’s history of technology and marketing initiatives have served blue-chip clients including GlaxoSmithKline, Kimberly-Clark, Sony BMG, Black & Decker, Forest Labs, Plum Creek, and more.

 

Prior to co-founding C3 Metrics, Greenfield was a recognized thought leader in the area of Branded Content as publisher of Branded Entertainment Monthly, a joint effort with VNU Media, detailing industry statistics, gaps, and trends. He’s been a featured speaker at NAPTE, The Next Big Idea, and a news source in The New York Times, The Washington Post, The Wall Street Journal, ABC, CBS, CNET, and Investor’s Business Daily. Greenfield studied Biochemistry at the University of Maryland, holds dual degrees from Southern California University of Health Sciences and is an instrument-rated pilot.

 

Jeff Greenfield at C3 metrics

Jeff Greenfield on LinkedIn

Jeff Greenfield on Twitter

 

Tweetables (edited):

“We find signals from noise, conversions from non-conversions and clicks from those that didn’t click. We are always looking for that signal to noise ratio and always looking for that return on investment.” — Izzet Agoren

 

“The demand was so fierce that I got threats from some of my customers!” — Izzet Agoren

 

“All models are wrong, some are useful.” — Izzet Agoren

 

“You don’t have to have your entire data stack and your entire data warehouse in order to perform the correct measurement of your media investments. It’s not as complex as people think.” — Izzet Agoren

 

“You just have to have the will to do data measurement, but it pays for itself many times over.” — Izzet Agoren