Quotient: Optimizing advertisers' and retailers' digital campaigns with an integrated solution

About Quotient

Since its launch in 1998 (when Quotient was Coupons.com), the company has been working to build stronger connections between advertisers, retailers, and consumers. Today, it offers proprietary technology with omnichannel digital marketing capabilities designed to drive sales and value through compelling consumer experiences.

Industries: Technology
Location: US

Tell us your challenge. We're here to help.

Contact us

Quotient consolidated all applications across the organization into Google Cloud to improve the scale and efficiency of its proprietary data and analytics capabilities.

Google Cloud results

  • Ensures enough resources to complete pipeline and batch jobs, including ad hoc requests from users, maintaining 99% of SLAs
  • Processes all in-memory cubes in under an hour, a 30% improvement from previous deployments
  • Empowers users to generate audience segments for marketing campaigns with 35% improvement in performance

2PB of enterprise data moved to BigQuery in six months

Standing out and building loyalty with consumers is tough in a competitive digital landscape. But with these challenges come opportunities to evolve marketing that is more aligned with what consumers want. This is where Quotient comes in. The Quotient platform helps consumer and packaged goods (CPG) retailers access the data insights they need to plan and measure marketing campaigns, ensuring they reach and engage customers across channels, at the right time, in the right way.

"Quotient's solutions help brands to target timely, personalized offers to consumers, which can drive engagement, loyalty, and ultimately, sales," explains Shanmugapriya Panchatcharam, Senior Director of Engineering at Quotient. The platform is powered with consumer spending data, location intelligence, client competitive analysis, and purchase intent data, which help to inform well-rounded omnichannel campaigns.

"Consolidating all our workloads on Google Cloud streamlines our pipelines and helps us achieve better data governance across the company."

Rashmi Joshi, Technical Lead of Data Engineering, Quotient

But as a fast-growing, global company, Quotient found that managing all this data across siloed environments (acquired through mergers and acquisitions) was no longer appropriate or productive. In 2022, seeking infrastructure uniformity, it migrated to Google Cloud.

The goal was to reduce data movement between systems by giving global teams easy access to the same environment. "Consolidating all our workloads on Google Cloud streamlines our pipelines and helps us achieve better data governance across the company," says Rashmi Joshi, Engineering Manager of Data Engineering at Quotient. "With Google Cloud, we can configure our data warehouse in multi-region datacenters while ensuring that data is safe and secure and can be recovered easily in the event of a disaster."

Improving performance and latency with an integrated infrastructure

In six months, Quotient migrated all its data warehouses, as well as internal and external reporting data applications, to Google Cloud. The company chose BigQuery to store its data warehouse with analytic capabilities built on top. Staging data is maintained in Google Cloud Storage, while Quotient's end-to-end data governance engine is powered by Dataplex's Data Catalog service. Meanwhile, Google Kubernetes Engine manages containerized applications.

Quotient is leveraging BigQuery workload management and slot reservations, assignments, and idle slot sharing. This approach dynamically manages resource allocation and provides guardrails, enabling Quotient's teams and processes to increase the efficiency of overall utilization of their environment.

This also helps ensure that Quotient has all the resources it needs to complete pipeline and batch jobs, including ad hoc requests from users. As a result, "we're able to maintain 99% of the SLAs because we have enough resources available for users to do their reporting with no delays. Users are able to execute their reports in a much timelier manner," explains Sravan Ankam, Senior Engineering Manager for the reporting infrastructure at Quotient.

With servers spread globally and teams working in silos, it was challenging to extract, transform, and load (ETL) data, since ETL jobs took almost a day to complete. But this additional time, alongside latency and query performance issues, has been significantly reduced. Sravan Ankam, Senior Engineering Manager for the reporting infrastructure at Quotient, explains why: "We extensively use BigQuery's idle slot sharing which has enabled us to finish our batch in-memory cube processing in under 60 minutes. That's a 30% improvement from our previous deployment, which is tremendously helpful because we now have enough additional runway for execution of user subscription reports and self service dashboards."

Exceeding customers' requirements with a data mesh architecture

Quotient is now exploring the possibilities of a data mesh architecture. The data engineering team is using Dataplex to define business and technical metadata to make it easily available for end users.

"We have built a highly targetable audience solution and segmentation process. Users can now generate target audience segments for marketing campaigns using brand-purchase matrix/custom queries experiencing a 35% improvement in performance."

Atul Agrawal, Senior Engineering Manager, Quotient

"We created an enterprise data warehouse with 2PB of data in BigQuery, which has exceeded our customers' requirements. This migration reduced data movement across multiple cloud vendors. It was challenging to extract, transform, and load (ETL) data for aggregates like basket, shopper trend and client-competitor analysis. With the help of sharing slots, Quotient was able to fully utilize all virtual CPU compute resources and processing time reduced by 40%," says Rajesh Venkatesan, Engineering Manager at Quotient.

Atul Agrawal, Senior Engineering Manager at Quotient, explains: "We have built a highly targetable audience solution and segmentation process. By understanding consumer's preferences, and purchase behaviors, we categorize customers into different segments based on their brand loyalty and their frequency of purchasing. Users can now generate target audience segments for marketing campaigns using brand-purchase matrix/custom queries experiencing a 35% improvement in performance." Quotient is now delivering 3 billion targeted or personalized monthly offers for different retailers which enables consumers with personalized savings.

Continually innovating a broader portfolio of capabilities

Quotient aims to use more Google Cloud solutions to power its ML workloads. "We have implemented several ML algorithms using on-prem infrastructure, our previous cloud vendor didn't have similar capabilities that BigQuery offers out-of-the-box. We're now looking at the BigQuery ML stack and Vertex AI to power the ML pipeline," explains Panchatcharam.

"With Google Cloud, we lowered infrastructure costs and improved the performance, governance, and efficiency of data products and teams. We're excited to continue to innovate on our existing solutions and prototype new ones for our customers."

Niranjana Kakumanu, Senior Director of Data Platform and Architecture, Quotient

"With Google Cloud, we lowered infrastructure costs and improved the performance, governance, and efficiency of data products and teams. We're excited to continue to innovate on our existing solutions and prototype new ones for our customers," explains Niranjana Kakumanu, Senior Director of Data Platform and Architecture at Quotient.

Tell us your challenge. We're here to help.

Contact us

About Quotient

Since its launch in 1998 (when Quotient was Coupons.com), the company has been working to build stronger connections between advertisers, retailers, and consumers. Today, it offers proprietary technology with omnichannel digital marketing capabilities designed to drive sales and value through compelling consumer experiences.

Industries: Technology
Location: US