Dream11 built a funnel analytic tool for user behaviour at a large scale: Abhishek Ravi, CIO

With the Tata IPL currently going on, there is also a plethora of games available for users to download and play.

Dream11 is an Indian fantasy sports platform that allows users to play fantasy cricket, hockey, football, kabaddi, and basketball. In April 2019, Dream11 became the first Indian gaming company to become a unicorn.

The users can actively participate in real-time athletic events and demonstrate their sports knowledge. However, ensuring that so many people enjoy the best possible experience on Dream11 every day can be difficult.

Funnel Analytics is one of the most typical behavioural analytics requirements they have at such a large scale, in order to understand user behaviour and preferences. They solved this using their in-house Data Platform, which collects, processes, and serves terabytes of data per day.

In this case study, Abhishek Ravi, Chief Information Officer, Dream Sports, tells us more. Also tells what it’s like to be a CIO. Excerpts: 

DQ: At the outset, please tell us about the business problem facing Dream11.

Abhishek Ravi: Dream Sports’ flagship brand Dream11 hosts 120 million+ users on its platform on IPL match days, and gives them the option to explore thousands of fantasy sports contests across a variety of sports. 

Our users can actively engage with real-life sporting events and showcase their knowledge of sports. With millions of users logged on the platform at the same time, it can become very challenging to provide seamless, best possible experiences on the app every day.

To address this, we continuously experiment with multiple features on our app and understand user behaviour such as time taken to navigate through different screens, to complete the journey, total time spent on a particular screen to see user navigation patterns from conversions/drop-offs stages and more. We strive to provide the best possible user experience. 

In this journey, our hero is DataAware -- a self-service funnel analytics tool used by the tech and product teams at Dream11.

DQ: How confident were you of the DataAware business solution?

Abhishek Ravi: We follow hyper-experimentation that is based on HEAL -- Hypothesis, Experiment, Analysis, and Learning. Everything at Dream Sports or Dream11 has to run  through this HEAL process. This enables us to fail fast, learn and build unique features for our users. 

We have also democratized data and its analytics to the extent that everyone at Dream Sports, with secure access control, can plan, test, or experiment with features backed by data.

We work in a cohesive pod structure, called Dream Teams, which includes 11 to 15 people across tech products, design, customer support team, etc., to ensure that every business problem we address has the best of both experiential and experimental minds, thus building the best product for users.

DQ: How did you all go about identifying the various needs? 

Abhishek Ravi: To ensure a seamless user experience for everyone who logs on to Dream11, one of the key requirements for us was to understand user behaviour and preferences. This involves mapping and analyzing a series of events that a user performs when he/she logs into the app, a journey that starts with user engagement in a mobile app and ends in joining a contest.

Multiple Dream Teams were involved in the entire process to ensure that users get the best possible product experience from our end.

DQ: The build/creation process needs time. How long did you take to develop this?

Abhishek Ravi: On our journey to understand common behavioural analytics, we developed an in-house tool DataAware — a Funnel Analytics tool that we use to recognize and know user behaviour trends at the platform. The process involved multiple steps from data collection to storage, processing, and integrating application programming interface (API) and visualization layer to ensure dynamic query generation.

After multiple rounds of developing and testing, we built DataAware with the features that were previously missing in our pre-existing analytics tool. Some of the key features include:

* Ability to provide a user interface to select the event sequence interactively
* Ability to apply filters on event properties
* Auto-suggestions on events filter properties
* Define conversion window
* Date Range — the ability for users to analyze across days, weeks, or months 

DQ: How has the project transformed the business? Has that also led to you identifying some additional opportunities to pursue?

Abhishek Ravi: Building our own in-house funnel analytics tool was a resounding success for us. It has helped bridge all the gaps that were previously missing in our pre-existing analytics tool. 

With the help of DataAware, we are effectively able to calculate conversion rates on specific user behaviours, in the form of a sale, registration, or other intended action from an audience. It also helps in better understanding drop-offs in navigating through the app and taking appropriate actions to increase conversions in Dream11.

DQ: Do you have any advice or any best practices to share with the other CIOs? They could be having some business challenges.

Abhishek Ravi: It is very important to let your team experiment and implement decisions based on discussions, surveys, and feedback. This helps people challenge the status quo day-in-day-out and make room for innovation. Any business problem can have multiple solutions, but you will be able to find those solutions only when you allow your team to experiment and innovate.

DQ: What does a good culture fit look like in an IT department? How have you cultivated it in the places where you have worked?

Abhishek Ravi: We value our employees and promotes a transparent, in-person, and open culture. We are guided by our five cultural pillars -- Data-obsessed, Ownership, High-performance, User-first, and Transparency. We encourage Sportans, our employees, to put our culture first. 

Over the years, we have discovered that the one thing that scales with an organization is culture. The hyper-experimentation-based HEAL program, in which Sportans are encouraged to experiment, fail quickly, learn fast, and move forward, is one of our organization-wide cultural practices that enable Sportans to perform their best stress-free. 

Instead of following the HiPPO (Highest Paid Person's Opinion), we encourage them to give their ideas and feedback and also take greater ownership of their work. We make decisions and adjustments as a group based on conversations, surveys, and feedback from all Sportans.

We believe that if our Sportans feel included and are welcomed to be themselves at the workplace, we can automatically reduce the level of stress or burnout that they may face.

DQ: What did you undergo to change the entire team's skill mix and culture? 

Abhishek Ravi: Our Dream Teams involve a mix of fresh experimenting minds and experienced engineers, and this helps us to not only go above and beyond the mainstream ways to solve a problem, but also test whether the product is sustainable and if it will work for a larger audience. 

Our early adoption of cloud helps our teams to quickly test out new features, scale tests in load/stress environments and drive maximum efficiency. These dream teams involve members from our legal and policy teams to ensure that our product is in tandem with the required regulatory guidelines.

Skill management benefits: 

* Opportunity to learn and work in an exciting, niche and fast-growing sector:Dream Sports is a great place to work that offers a confluence of sports and technology. We also heavily invest in mentoring our team members as we truly want them to grow and expand their skillsets.

* Opportunity to train and work with the latest tech stack: As a data-driven team, we work on the most cutting edge technology and build many home-grown solutions that address key product-related challenges such as user experience analysis, scale management, mobile app automation, security and FairPlay, etc. 

We work on a scale and user concurrency that is amongst the highest across apps in the tech industry. At Dream Sports, even the newest and youngest of our team members are free to raise their needs and concerns in terms of professional and personal development. 

We also help and encourage Sportans to take every opportunity possible to upskill and learn the latest technology. For example, towards the end of 2021, we sent select team members to London for an immersive training session with Scala.

* Learning wallet: We believe that Dream Sports is a great place for on-the-job learning, and this needs to be supplemented by the pursuit of learning new concepts that every Sportan can apply at work. Sportans can avail of this benefit for every financial year of which Dream Sports will bear 80% of the costs. 

Our 'Learning Wallet' allows Sportans to take charge of their own learning, in and outside of their job profiles, to enhance their knowledge, work in great synergy with other clubs (departments) and excel in their careers! They may take up any development program(s), training/certification program(s), seminars, books, ebooks, webinars, or any other learning initiative. The learning wallet benefit includes all expenses related to the Sportans’ learning initiative, including but not limited to their travel and accommodation, along with the tickets for conferences, workshops, seminars, etc.

Change management benefits:
* Proximity to the Stadium: We are committed to providing flexibility and work-life balance to all Sportans. Especially in a city like Mumbai, where daily commute and long distances prove stressful for most people, our Sportans have the reassurance of this benefit, which enables them to live in their ideal residence, while remaining close to the Stadium. 

Under the benefit, Dream Sports pays up to two-thirds of their rent and assistance in rent deposit payment so they can live within 30 minutes of the Stadium and commute to work easily. The benefit is available to Sportans even during the pandemic when they are working remotely, without worrying about rent and accommodation expenses.

* Relocation benefits: We recognise great talent from across the country and provide complete relocation benefits to Sportans and their immediate families, so they can join us in Mumbai. This includes taking care of all expenses for their relocation, accommodation in five-star hotel properties, while they look for their dream home in Mumbai and assistance in finding the same near the Stadium.

* Notable ongoing employee engagement initiatives include internal quizzes to test sports knowledge, opportunity to join exclusive sessions and clubs, monthly experience vouchers for employees and their loved ones, one-on-one sessions with physiotherapists, in-house massage sessions twice a week with a professional masseuse, access to healthy meals throughout the day, opportunity to meet and greet sports stars, and the opportunity to avail exclusive match tickets.

DQ: What are some of the technologies that you considered to enable the digital transformation?

Abhishek Ravi: Our core purpose behind any tech innovation at Dream Sports is to consider every opportunity to ‘Make Sports Better’ for everyone. Some of our recent tech innovations include:

* Push Architecture: We are putting in a lot of effort to use this across our product to become as real-time as possible.
* Use of concurrency prediction model to predict hourly concurrency on the 
Dream11 platform.
* FENCE (Fairplay Ensuring Network Chain Entity): Fraud detection system to identify and mitigate users creating multiple/ duplicate accounts on the platform to abuse the referral or promotional cash bonus schemes.
* Funnel analytics: It is a tool that recognizes user behaviour trends. It has proven to be a resounding success and helped us bridge the gaps that were previously missing in our pre-existing analytics tool.
* Data products: Build self-serve data products on top of our data platform from acquisition to serving and enable taking seamless data-driven decisions for all our teams.
* Mobile app automation: It helps in increasing the efficiency, reducing the execution time, performing feature regression, and releasing it for Android and iOS platforms.
* Torque: During big-league events such as IPL or World Cup, one of the key challenges is to serve more than 5.5 million concurrent users, and offer them a seamless experience in traffic. For this, we automate the entire process and named the framework Torque.

DQ: Elaborate on the Torque.

Abhishek Ravi: During IPL 2021, we had updated Torque to include new changes, introduced a new tool, to mock those dependencies which do not affect our transactions directly and helped us in simplifying the complex setups that are needed for tests.

To ensure a high-quality user experience during peak traffic, we also load/stress test every feature that is released for a smooth, performant and resilient user experience at scale. We have a testing framework that simulates any kind of traffic load with real life-like patterns. This gives a high degree of assurance that the backend would behave exactly as expected.

DQ: Finally, can you touch upon the methods used to build DataAware?

Abhishek Ravi: At our scale with 16+ terabytes of data per day, we had to go through a few experiments based on HEAL (Hypothesis, Experiment, Analysis, and Learning).

Shortlisting the right mix of tech stacks was key and the overall approach was to break down features into iterative releases and continuously gather feedback to build on top.

Additionally, as a data-obsessed team, data observability helped us to understand what's working and what's not. We tracked the usage of DataAware on a day-to-day basis and continued with incorporating the feedbacks both from internal and external stakeholders.

DQ: How was the team organised and structured for building and operating DataAware?

Abhishek Ravi: As we work in a Dream Team structure, having the best minds with the right skills is required to deliver DataAware end-to-end. We formed a team consisting of data engineers, analysts, front-end engineers, and product managers who worked as a mini startup and continued to enhance DataAware by incorporating user feedback.

DQ: What has helped you the most in your career?

Abhishek Ravi: I strongly believe that skill can only take you to an initial start. Attitude is what sets you apart. Having an open and learning attitude helped me immensely to grow in my career and model my team to achieve greater unthinkable feats. They always challenge the status quo with humility and that makes things incrementally better each day.

DQ: So, would you say that the mandate of the CIO role has changed in the last decade?

Abhishek Ravi: CIO leaders have grown beyond their technological expertise. They have invested in their soft skills to be more approachable for their team. CIOs can always employ people with deep technological experience, but with greater emphasis on soft skills, they are also able to convey the benefits of technology strategy in business terms. Hence, in the last decade, CIOs have become more empathetic and motivated to pursue new business opportunities.

There is a significant skill gap to overcome, as well as a new literacy that must be cultivated. It is crucial to find strategies to encourage individual knowledge in each IT discipline, while also aligning it with the business and technology goals. 

At the same time, CIOs and their teams will never be able to master all of the technology specialities. They must keep focused on the digital transformation goals to avoid becoming distracted.

DQ: Are there any roles or skills that you were finding the most difficult to fill?

Abhishek Ravi: The tech industry in India is rapidly evolving, resulting in a scarcity of talent that is up-to-date and skilled in the latest technologies. While young tech professionals can help fill the gap to some extent, experience in solving technical problems is just as important, if not more so. As a result, having the correct mix of expertise and experience, as well as a desire to learn and explore, is highly sought after.

While there are many people that have gained expertise in AI/ML/Data in the last 7-8 years, implementing such solutions at scale remains a challenge. We have a severe shortage of people and the ability to scale AI solutions. We search for people who can strike the perfect mix between delving deep and keeping up with the speed with which our products change at Dream Sports.

DQ: Would you say that classification of IT and digital assets are changing in organizations in this era of everything software and as-a-service?

Abhishek Ravi: Companies are transforming the way they work, communicate, create products and services, and deliver them to their customers as a result of digital transformation. 

Software-as-a-service (SaaS) enables businesses to provide their products and services to clients in a streamlined manner. Customers can receive products and services via the Internet, resulting in convenience and accessibility.

Installation, patch management, version control, bug effects, and user access management are some of the operations that are almost obsolete in the SaaS environment as the ITAM (IT Assets Management) focus has switched to monitoring software assets, rather than managing software assets. 

While the focus to optimize costs and maximize the use of IT assets remains the same in the SaaS environment, the process of measuring IT usage has changed dramatically. The most noticeable change is driven by ITSM (IT Service Management). It is a platform of service, rather than support, and hence, crucial in monitoring the organization's cyber-safe practices, as employee training is a critical component of managing cyber-security threats.

DQ: You also have an aggressive mobile strategy. How has that been integrated with the IT infrastructure?

Abhishek Ravi: Generally, a minor update to any service or app might take a long time and effort to implement because it necessitates testing on both platforms, i.e., Android and iOS. Every time a new feature is developed, it necessitates regression around the core flows, resulting in lengthier release cycles. 
As a result, we developed a Mobile App Automation Framework that would aid us in enhancing productivity, reducing execution time, re-usability of test code, and lowering maintenance costs.

Java, Maven, TestNG, Appium, Jenkins, Android Studio, and Xcode were the tools and technologies that we used to achieve this approach. Further, the emphasis will be on developing the ability to capture per-screen application programming interface (API) calls in order to monitor and analyze performance impact, as well as improving the event capturing mechanism and performance-related metrics such as memory, CPU, network, and battery.

DQ: Would you give any advice to the aspiring IT leaders?

Abhishek Ravi: One must accept that the role of an IT manager has shifted. Tech teams are always under a lot of pressure to stay relevant in the market as technology evolves. Following the pandemic, companies have placed a greater emphasis on mental health and employee well-being. 

As IT leaders, we must recognize that understanding your employees' skills  is very important, alongside your company's needs. In order to lead, one must have compassion for others. Recognize the power of people. Always emphasize people in everything as they will be the most valuable resource. Be fair and accepting of flaws.