How can you take billions of lines of log data, explore it to identify those hidden gems of information, and act on those insights instantly to improve your customers’ experience, enhance your products, and maintain a competitive advantage in a digital-first world?
This past week at PASS Summit 2018, Taboola, a discovery platform, shared how they are tackling this with the help of Azure Data Explorer, a lightning fast service optimized for data exploration. Azure Data Explorer allows users to get instant visibility into very large, raw datasets in near-real time to analyze performance, identify trends and anomalies, and diagnose problems.
Taboola delivers personalized recommendations to more than one billion users every month, rendering 3 billion web pages daily on many of the web’s most innovative and highly visited websites. Thousands of leading publishers, marketers, and agencies utilize Taboola’s predictive technology to increase user engagement, monetize their traffic, and distribute their content to engaged audiences across the web. Ensuring customers achieve maximum success from their campaigns is crucial to Taboola’s success.
Taboola’s data generates more than 22 billion records of edge delivery logs daily. Taboola needed a way to quickly, simply, and cost-effectively analyze this edge delivery log data to continuously improve their service with reliable, top performing content recommendations. Its previous solution consisted of several technologies stitched together from different vendors to store, manage, and query their data. This solution was complex and costly to use.
Following a demo of Azure Data Explorer and some testing using a days’ worth of log data, Taboola found that they could get fast answers to in-depth queries on the data.
Taboola relies on Fastly to deliver fast, secure, and scalable online experiences with its edge cloud platform. Utilizing Fastly’s edge cloud to stream the real-time logs directly into Azure Blob Storage, Taboola configured their Azure Data Explorer service to automatically ingest that data. This solution outperforms their previous solution with a faster update time and an intuitive interactive interface.
“The solution was so simple that we were up and running in a week, ingesting and analyzing 17 TB of data per day,” says Vice President of IT at Taboola, Ariel Pisetzky.
Fast time to business value
Taboola has immediately been able to put their insights to work to improve their customers’ experiences. Taboola is now able to more quickly analyze performance, identify trends and anomalies, and diagnose problems in near-real time. One such example is how Taboola was able to instantly identify a problematic latency issue impacting the speed at which content was being served to their customers, and quickly make the correction.
“The ability to do this in near real time, running queries constantly against the system is something we could not do before,” says Pisetzky.
Another example is how Taboola is leveraging these instant insights to help Taboola’s R&D team make advances in their product, influencing the design of new algorithms, and the generation of new ideas.
The fast deployment, short learning curve, and empowering pricing model are just some of the contributing factors that Pisetzky shared as to how Azure Data Explorer is helping the business. While just a few weeks in deployment, Taboola is already able to estimate savings of an entire headcount with Azure Data Explorer versus their previous solution – a head count that can be re-allocated to driving business innovation!
To find out more about Azure Data Explorer you can: