Tremor is a leading company in the field of advertising and marketing. It offers a publishers' automatic data-based e-commerce platform. Its software platform is designed for various media and appliances (i.e., mobile, smart-TV, CTV, etc.) and aims to optimize commerce between the publishers-buyers and the sellers.
Tremor’s solution was developed in order to provide an end-to-end solution. With its superior video offering, Tremor Video provides effective ads & give access to unique audience builds exclusive for your brand that brings video advertising to new heights, making them effective, and measurable.
The need: Big-Data Solution
Twingo and Tremor‘s partnership started in 2014, when Tremor was still called TAPTICA.
It was when Tremor needed a Big-Data solution to support its ongoing growing data requirements in order to provide analytic-based answers with fast response time for its various clients.
After the diagnosis process and a complicated POC, that included tests for fast load times and a quick analytical query response, Tremor chose Vertica as its Big-Data analytics solution. Today, over 2 PB of Data are stored at Vertica and used for efficient analytics.
Twingo’s involvement in the process continued in the design and implementation of additional data features, various ETLs and data pipelines. Suffice to say that Vertica is still Tremor’s choice.
The next challenge: BI system
Through mutual discussion, Tremor set two main objectives:
1. Allow users self-service abilities
2. Improving Dashboards performance
Tremor was then using an unsuitable BI tool that did not match its business and technical requirements, or its customers’ SLA.
“Issuing reports was so complicated, it was near impossible. We realized the first solution we needed to address was our ability to create and produce dashboards and analytics that are fast, and easy to access independently, it was not possible with Tremor’s then current system. It also lacked a fast development rate and was not sufficient for Tremor’s business requirements.”
“At that point, creating a dashboard could literally take between three days and up to four weeks. Processing capability of on-going data was restricted in size, and in some cases, the data volume was too big to handle, resulting in product returning errors.”
“Almost everything we wanted to do, was either too complicated or impossible within the limitations of the solution tool”, explains ChenTal, Tremor’s Business Intelligence Team Lead.
Twingo was again summoned to challenge the task. The objective was to identify Tremor’s business and technical requirements and provide the right tool to comply. “We focused on our major requirements and evaluated them: development speed, response rate, data freshness and support options for all its clients.”
Solution search and evaluation
Together with Twingo, our Big-Data partner, we started sifting for the right solution.
One that will fit our requirements, improve our analytical abilities, and achieve our goals.
After compiling together a list of products that might act as solutions, we came to the decision that Looker GCP meets all of our requirements.
The minute the user wishes to see the data, Looker runs a live-query of the Database and extracts the data. Together with Twingo’s BI developers, we started the POC, which was part of the actual project. They were always hands-on, guiding us how to use the tool effectively, and optimize it, as well as instructing the team. They were an essential part of the deployment team and did so very professionally.
Looker GCP enables us to improve our working methods. Before the switch, it took three developers working simultaneously to create dashboards. Now, we have over 30 users that can simultaneously create dashboards.
What makes Looker GCP even better is its user-friendly interface, ontop of its ability to explore the data and drill down. These features save a lot of time and allow users to rely on analysis and base their decisions accordingly, which is the system’s primary goal.
Today, all of Tremor’s employees, from C-level to analyst, can independently explore Tremor Data, giving them the capability to make data-based decisions making us true to our goal as a data-driven company.
“Our performance issue was solved far above our expectations. An enormous Buzz flooded our international branches, and we passed the challenge of adjusting the BI tool to our architecture with flying colors. Looker and Tremor’s architecture are 100% native to each other. They extract data from 90 billion records, and it returns within only 3-4 seconds. It takes 5 seconds from the moment Looker retrieves the Data, until it is presented on a color-coded dashboard with complicated graphs – it’s impressive.”
