There's no contradiction between privacy & performance: Saugato Bhowmik, Meta

Meta's Saugato Bhowmik maps out the path forward for personalized and relevant advertising in the privacy-first age where user concerns and new systems such as Apple's ATT are on an upsurge.

Personalized advertising in the social media landscape is going through a reform with the rise in data protection laws, privacy-first systems, and user concerns increasing by the day. Saugato Bhowmik, Director (CPG, D2C, Auto), Meta, outlines the path paved through technologies such as machine learning and new advertising models built at Meta that show relevant advertising but not through any personal-level user data.

Saugato states, "All digital players are looking to ensure privacy-first performance play". Here he talks about building a system and new experiments that utilize data shared by advertisers, in a privacy-first way, whilst still showing relevant advertising, and helping advertisers deliver performance.

There's no contradiction between privacy & performance

Saugato Bhowmik

Privacy-First Advertising

He discusses building a system that gets better with every passing signal and adding multiple different features that businesses can use to ensure optimization is richer, even in the absence of personal-level data.

Also Read: Meta isn’t just where users discover brands & products, it’s where brands & products discover customers: Arun Srinivas

Custom Events

Meta is also building a lot more custom events based on what is valuable to the businesses. These custom events go deeper into the consumer journey and engage consumers who are already present in the funnel. These events would be generic in nature but built on industry-specific insights.

He shares an example of brands from the Media & OTT industry looking to increase subscriptions, the event would tap consumers who have stayed within the funnel and opted for a free trial.

Cross-Channel Optimization

To better inform the systems, Saugato shares that the system will pick up more signals across the website and business app, and mix that for better optimization. He mentions India is an app-first market but also heavily relies on traffic generated on websites. For data signals to be aligned, it considers both app and website signals.

Post-Conversion Event

"Adtech industry grew during the unfortunate period of COVID-19 but is here to stay," says Sougato. He further outlines an ad tech tool that enables a post-conversion event that provides specialized support crop optimized support for consumers who already purchased a service or product. This is currently in the beta phase.

Privacy-Centric Framework

He discusses a measuring and reporting framework in a privacy-entric manner, which has been developed to consider changes in the landscape such as the ATT update by Apple. Explaining it further, he states that the framework "strips the data, delays the reporting, and aggregates all the data".

Conversion modeling is an AI-based model that generates probabilistic outcomes to report the impact of the campaign. The data received is aggregated data that advertisers provide.

"There's no personal data being accessed. We take the data and put it in our own conversion model that uses ML, and improve the data so that the reporting is done in a privacy-safe way, he adds. "We're no longer receiving personal info," he says, and it is transferred via a direct server-to-server connection.

Meta's Saugato Bhowmik spoke at the Data Unplugged Conference by MMA India.