Sabio’s Aziz Rahimtoola on Mobile and App Data

“Sabio means wise or experienced in Spanish,” he noted. His company strives to address the issues with current panel measurement systems through mobile and app data.

Charlene Weisler: What is the state of CTV/OTT in general at this time and where is it headed?

AzizRahimtoola: The CTV/OTT media landscape is quite fragmented due to the various operating system platforms and increased competition from every angle. This is good for consumers as it allows choice, but it will continue to be confusing for marketers.

Weisler: In terms of methodology, what approach is best for marketers?

Rahimtoola: I believe that app data is more effective than panels in understanding consumer behaviors. This has been validated by Nielsen and other panel-based companies that admit to not effectively understanding streamers (app users) or diverse audiences via panels. In the 1970s, when society was 11% and monolithic, panels were more effective than now when at 42%, our population is more diverse.

Weisler: What is App Science?

Rahimtoola: App Science combines data signals from 280 million active mobile and 110 million CTV devices and closes the quality gap between multicultural and general market data. Traditional survey panels have historically underrepresented and undercounted multicultural segments of the population. By analyzing CTV and mobile device data, App Science draws a complete picture of multicultural audiences, their preferences, and interests. They also provide trends and insights for their clients to help them better understand consumer actions and interests so they can make better strategic business decisions. This can entail programming preferences, competitive analyses, industry vertical insights, or multicultural and diverse audience trends that traditional measurement companies typically underrepresent.

Weisler: What data do you at Sabio collect and apply?

Rahimtoola: We have GDPR and CCPA-compliant ways of obtaining mobile and CTV/OTT IDs. We also use machine learning extensively to understand consumers via predictive modeling better. In addition, we provide advertisers with a pixel, or a snippet of code used for tracking, that can be applied across media buys to measure campaigns. The information can help to compare viewership among different platforms and avoid targeting the same audiences repeatedly.

Overall, we capture over one terabyte of mobile and CTV/OTT data daily and ensure that everything is ethically sourced, privacy compliant, and cleansed. This data contains a wide range of metadata that allows us to apply machine learning algorithms and other predictive modeling to considerably understand consumers’ demographics, points of interest, behaviors, and life stages. From there, we can create granular audience insights and segments that can be applied to media buys and measure campaigns’ effectiveness.

Weisler: How do you craft segments from your data?

Rahimtoola:Segments are created from a collection of CTV and mobile data where we apply the nearest neighbor, lookalikes and another predictive modeling to various attributes. When we look at the data collected from CTV/OTT and mobile, we analyze various attributes such as app ecosystem, technographic, brand visitation, purchase signals, and census data and further verify with trusted 3rd party sources. From there, we apply machine learning algorithms such as the closest neighbor, lookalikes, associations, and topic modeling to create clusters, user profiles, and audience segments that all connect back to our proprietary 55MM household graph.

Weisler: How does Sabio differ from other platforms in the CTV/OTT space?

Rahimtoola: Sabio Holdings is the only company outside Roku with a complete end-to-end CTV/OTT technology suite of services. We can partner up with content creators and launch apps along with creating and distributing add breaks in content with our newly acquired Vidillion SSP (Supply Side Platform) acquisition, monetize them via our Sabio DSP, all while providing differentiated non-panel-based analytics via its App Science platform.

Weisler: How do you use your data to inform political campaigns?

Rahimtoola:: We enhance National voter files with insights that help identify a political party’s viewing habits, app behaviors, points of interest, and key issues that matter to them most so that they can use this audience insights to inform their future campaigns, validate they are reaching their audience, and optimize in real-time.

Weisler: What were some of the trends in political advertising leading up to the midterms?

Rahimtoola:Looking back on November 8^th, we noticed the most notable political advertising trends being the increase in growth in CTV and OTT. The political sector has always been reluctant to buy into CTV, but compared to 2018, the shift has been dramatic. Prior, most political ad spending has been programmatic. We also saw an increase in the use of QR codes, and interestingly, there’s been a multicultural shift. In previous years agencies said they didn’t have a big enough budget for the Hispanic demographic. Still, there has been an increase in the general market putting money into the multicultural market.

Click the social buttons to share this story with colleagues and friends.
The opinions expressed here are the author's views and do not necessarily represent the views of MediaVillage.com/MyersBizNet.