What it took for Bigbasket to build a great data science team, CIO News, AND CIO
And to get the most out of AI and ML, a business needs a team of data scientists.
For Subramanian MS, head of category marketing and analytics at Bigbasket, a good data science team needs to bring together a set of complementary skills to deliver data science work products.
“Developing and delivering a data science work product such as a smart shopping cart or recommender system requires a combination of skills – domain knowledge, technical knowledge, engineering knowledge, and analytical skills. Therefore, a good data science team must bring together skills spanning multiple roles. Based on the data science problem being solved, smaller teams should be created with relevant skills from the larger team,” he said.
Since its inception in 2011, Bigbasket has hired and worked with numerous data scientists to build an effective and efficient data science team. But with changing hiring dynamics and the employee cycle, it has now become nearly impossible for a company to retain employees for longer. To solve this problem, Bigbasket took a three-tiered approach to investing in its data science team.
The first layer, as Subramaniam mentioned, is to offer them challenging roles. Most data scientists love data and statistics, they enjoy working on complex and important issues, so they like to be challenged in their work with different projects and business issues.
They should have the opportunity to work in business functions such as marketing, merchandising, or design in order to understand the business as a whole.
The second layer is to invest in employees as individuals. It is important to invest in their development and to advise them in their professional role so that they can continue to learn and grow. And a few Coursera courses can’t help a company improve the skills of its data science team. Therefore, there must be an opportunity for role-based learning available to them.
This is what Bigbasket has recognized and worked on and has therefore moved to a personalized development program for its employees.
“We recognize that the era of unified, multi-hour/day classroom development programs is over. With the advent of on-demand learning opportunities available, we have moved to personalized development programs tailored to each individual’s interests. We build a catalog of courses, learning sources and work with team members to build a development program in several periods. While this provides a formal training path for skill development, we also encourage team members to work on cross-functional projects to grow by observing and learning from experts within the organization,” a- he explained.
The final layer is to reward them and recognize their work. Framing rewards, bonuses, and ESOPs that match the industry phase helps them understand that the company values them. But more than the compensation, sometimes it is important to recognize their work. Leaders need to ensure that data science team members feel the work has an impact on the business.
With attrition at an all-time high in the industry today, it’s up to the company how it views the situation and how it invests in its employees. Although attractive entry bonuses and high compensation are a part, the most important thing that keeps an employee in the company is when they are able to find purpose in their work.