AppsFlyer launches ChatGPT plug-in for SQL queries 

AppsFlyer introduced a new feature in its data clean room. This feature lets marketers ask business questions using everyday language, instead of needing to use SQL for data analysis.

AppsFlyer is testing this feature with partners like the Hopper travel booking app and a few other apps in areas like transportation, gaming, financial technology, and food delivery.

Usually, when marketers want to get information from their clean room data, they have to use predefined reports that come out daily or weekly. These reports only show what the reporting tool’s interface allows. Otherwise, marketers might have to get help from data scientists and business intelligence engineers to analyze data, but these experts aren’t always available because they’re busy.

For people who don’t know how to code, working with a database is usually not something they can do on their own. However, marketers simply wish to explore their data at times, according to Edik Mitelman, who is in charge of AppsFlyer’s Privacy Cloud platform that focuses on privacy-focused measurement, activation, and reporting tools.

“They want to discover new things and access data beyond what they could get from a report,” Mitelman stated. “That’s why we launched dynamic query capabilities, so that marketers can tinker with their data and get the insights they need.” 

A little over a year ago, AppsFlyer introduced its data clean room as part of the Privacy Cloud platform, starting with a simple reporting interface. Over time, based on feedback from customers, they have gradually added more advanced analytical capabilities. However, according to Mitelman, the significant enhancement comes in the form of enabling marketers to personally ask questions directly to the system.

“You can run any complex model ad hoc on the data, and this will give you results,” Mitelman said. “Just ask a business question, and it will answer you.”

The dynamic query tool is powered by OpenAI using a ChatGPT plug-in. Marketers can log in and communicate with the system using any language supported by ChatGPT. And, they can perform various queries, ranging from simple ones to more complex ones. To address potential inaccuracies (where large language models might produce incorrect information), AppsFlyer allows users to edit the output before using it for a comprehensive analysis.

All queries are performed within the system, ensuring that no data leaves it. This means that while AppsFlyer’s model learns, private information is not exposed online or used to broadly train OpenAI’s tools, which is essential for maintaining the integrity of a data clean room. 

It is designed to work with any cloud provider (aka cloud agnostic) , making it adaptable to various cloud warehouse partners that marketers might use. However, there’s still an issue of interoperability within the data clean room category. According to Mitelman, it’s complicated and inconvenient for advertisers and publishers to collaborate with multiple data clean room vendors without integrating all of them.

AppsFlyer’s Data Clean Room is the initial step in a series of technological advancements under the AppsFlyer Privacy Cloud initiative. This innovative tool empowers app developers to create insights from their data in a private and secure manner. It allows for the analysis of first-party data, attributed marketing data, in-app events, and third-party data within a privacy-conscious environment. Therefore, app developers can precisely measure and make more informed decisions regarding their marketing campaigns. This leads to improved return on investment while safeguarding customer privacy. This capability drives growth in marketing and overall business endeavors, benefiting both supply and demand sides. It fosters confidential data collaboration and provides actionable insights that directly influence the business’s overall performance.

Written by Gizem Yılmaz


Leave a Reply

Your email address will not be published. Required fields are marked *


The 7 Best Shopping List Apps for Android

9 in 10 CTV ad dollars will be programmatic in 2023