in development time
decreased from one week to 30 minutes
BOLD is transforming the career industry through the creative use of technology. They work with job seekers and recruiters to connect the best people and companies. Although they were surfacing their data with an existing tool, the visualizations weren’t what they needed. Head of BI and Big Data, Balaji Jayapal, wanted to find a better way to manage his 2TB of transactional data. Sisense was implemented in a number of creative ways from system monitoring for variances outside of standard deviation to reducing the number of credit card chargebacks.
It’s extremely easy to learn. If someone has experience in BI, even very basic, it’s very easy for them to pick up Sisense and start building dashboards very quickly.
The Opportunity: Tweaking The Programs
There is a huge volume of data collected by Bold. They currently have 60TB of data and are actively analyzing 2TB. They provide subscription-based services, which include a resume builder, cover letter builder, interview prep, job postings and worker postings. Each subscription has different frequencies and levels that need to be tracked. At a basic level, they want to see what subscription types are getting renewed the most, what subscription products are getting purchased the most, and what is the most effective model for connecting employees to employers, so that they can fine-tune their builder tools for maximum efficiency. Balaji was looking for a way to better visualize their transactional data to get those insights.
The Search is on
Bold’s current solution was not something that they wanted to continue using and, even though Balaji had worked for that company prior to being hired at Bold for his knowledge of the tool, it was also something he felt wasn’t working for their needs anymore. He did an extensive search of available tools, then tested several on live marketing data, and showed them to various affected departments to get buy-in. When it came to the crunch, it was the visualizations and ease-of-use in Sisense that made the deciding difference, and the purchase was made.
Sisense came out on top because it was easy to develop.
Profit in The Strangest Places
Balaji had a very interesting case with his Platform team, which had to do with credit card chargebacks in the subscription business. The credit card company charges a fee for every chargeback, which can get expensive, so keeping those to a minimum is significant to the bottom line. The first 5 digits of a credit card is called the “bin”. This identifies the type of card, which they then use within Sisense to look for chargeback ratios and find which ones are the most susceptible to problems. They communicate this to the developers who can then decline the transaction. The amount of money they get from the subscription isn’t worth the cost of the chargeback, so the user is prompted to use a new credit card or they can just bail out of the purchase. Each prevented chargeback is worth about $30 to the company and they’ve been able to drop the chargeback rate dramatically with this new process. This data granularity didn’t exist before.
With 20 Elasticubes and 50 dashboards deployed, Bold is using Sisense across their transaction data, business data, e-commerce data, and marketing data, each with their own set of dashboards. Balaji has made some very creative dashboards, including a macro-analysis system. Using their ETL tool with Sisense, they are doing hourly analyses of their transactions and applying standard deviation to compare
to the prior six months. If there is any variance beyond +/-2 then the system will send an alert to predetermined employees who can then login to Sisense and see what is going on by drilling into the details.
Elasticubes are now being updated with a frequency of between 6 and 24 hours, which provides a level of visibility that wasn’t possible before. The C level executives are loving the product and using it extensively. Whenever a new department is shown Sisense, they can sit down and create a dashboard in a live session against their data, creating a lot of excitement. Balaji is working on a new dimensional model and is anxious to go after the other 60TB of siloed data to finish a full deployment across the company.