The way we structure teams @ Data
Find out how our Data team evolved and what we do to continue growing
Although the structure of a team depends on the size and maturity of the organization, in most companies we can find a Data team typically composed of the following profiles:
In our case, the team is very heterogeneous: we have physicists, social communicators and accountants, among many other profiles. We believe that this diversity of backgrounds helps us see the challenges we face every day from a unique perspective where synergy and collaboration take center stage.
As a company, etermax has been growing by leaps and bounds; when I joined in March 2017 we were approximately 220 people and today, May 2022, we already exceed 500. From Data, we reorganized and managed to grow 4 times in headcount during this same period. These changes were accompanied by new challenges and more data!
From the very beginning, we understood the need to have a decentralized team. In other words, we wanted to assign analysts and data engineers to the different product squads or business units, so that they could become a part of the team. Personally, I consider that this was a great success and it is one of the objectives we have been working on the most lately.
Did it work well from the beginning? No, but we had several learnings that helped us to continue iterating the idea. It is difficult to find the sweet spot between having a centralized Data team that provides internal consulting, and having a group of analysts that works directly with the business and has little daily contact with the rest of the Data area.
In our case, part of the team works in a decentralized way and another part works in a centralized way. What worked best for us was generating synergies between both of them, participating proactively in these teams from the outset: contributing ideas to the product roadmap, proposing business opportunities, validating them through AB Testings and helping the team to continue iterating on them.
Is there an example where this decentralization stands out? Yes, sure! A few months ago, with part of our Data Analysis team, we increased our ARPDAU (Average Revenue per Daily Active User), one of our key metrics, by 2 digits. This was the result of joint work between different profiles: Product Owners, Software Engineers and Data Analysts. To achieve this goal, we proposed different strategies that we iterated with the Product and Engineering teams and chose the best alternative based on cost-benefit. Once the best proposal was reached, we proceeded with an AB Test to validate the hypothesis. This time, the result was widely favorable and we were able to move forward towards production. However, none of this would have been possible without this great teamwork. In other words, it would have been very difficult for Data alone to understand the cost of developing a functionality in terms of software engineering or even understand the priority in the product roadmap.
We understand that our team structure is dynamic and it is natural that it evolves over time to adapt to new business needs. Here are some images with the evolution of our team structure from 2017 to 2022:
At first glance, you can see how we have deepened decentralization and created teams that intend to become as professional as possible in the development of skills the Product and Business teams need. In order to accompany this growth, it was necessary to rethink our structure. Regarding this, we worked on 2 core concepts:
● Have a team as horizontal as possible so as to be agile.
● Create a leadership structure made up of leaders and managers which enables us to continue escalating as a team.
Whether centralized or decentralized, I think one of the most important points that helped promote and consolidate these changes was the change in mindset. Mainly the transition from being a support area which provided ad-hoc solutions to become a team that proactively co-creates products and solutions with various teams
What has been working for us lately? Understanding that within the teams it is more convenient to have several points of reference in different subjects. In this way, we can specialize in particular knowledge, whether technical or business-related, and thus be more efficient. For example: we have Data Scientists who are experts in NLP and help us improve the quality of the questions we serve our users on Trivia Crack. We also have Data Engineers that work closer to business-related teams and have solid abilities in User Acquisition.
What points do we have to continue working on? At the moment we are mainly focused on continuing to accompany the product teams and producing an impact from that place. However, we also need to work on the improvement of Data Governance and Data Quality. Today, the whole Data team shares both responsibilities and all our members work on them. In the future, hopefully we will have teams with specific profiles that can handle them.
Would you like to be part of the etermax Data team? Find about our current job posts in this link!
