Data Analytics

Leading a Data Team

(with people more qualified than myself)

A great joy in my career was having the opportunity to create and lead a data team from scratch. On top of that, with people who had more data knowledge than I did. A great opportunity to learn a lot.

To start a team inside a big corporation such as BairesDev was like managing to sell an idea or a project for people inside my own company, and receiving approval from the C-Level team was like having them "buy my vision". It's a good intrapreneurship story that could have been the sale of a large project to a client in my previous consulting life.

At that point in time, after 6 months from my first "data task", our team had grown from 1 to 4 people, and we went from 1 to more than 100 charts in various dashboards, visited monthly by more than 70 people from BairesDev and other hundreds of clients.

As I mentioned in more detail in the Product Management section, I joined BairesDev to work as a Product Manager on a new project. However, the business grew quickly and as there wasn't a Data Analytics team in our WinDifferent business unit, I had the opportunity to acquire additional responsibilities.

Key drivers behind the need for a team were customers wanting information about how their marketing campaigns were going, and problems with each team/manager looking at information from different data sources, which did not communicate with each other and generated distrust about which information was correct, harming the decision-making process.

Something I have always believed in was shaping my career and being the protagonist of my own story.

If I could share a piece of advice with younger colleagues, I would say never expect that someone or life will magically put you in the place you want to, and always ask, suggest, or act to put yourself in that position.

After all, if you don't say what you want, how will other people be able to help you? If you think no role suits you, why not create your own role and scope? Pre-defined models don’t always make sense.

It was with this in mind that I decided to take the lead and use the data knowledge I had learned during my time at Red Ventures to help with some of BairesDev's data needs. Far from being an expert, I was just someone with basic knowledge - but enough to make it happen and solve the main pains we had at the time.

Luckily, I had the trust of the CEO of our business unit at the time, who, when he realized that I needed data that was somehow related to my product, tasked me with "setting up a dashboard" for the executive team.

I still remember to this day the moment I fell into a Zoom room not only with the leaders of WinDifferent, my business unit, but with several C-Levels and founders from BairesDev itself, asking me if I was the guy who would solve the dashboard problem for them.

It was literally me and all the C-Level team there. I was very honest about the fact that I had other responsibilities and that I could only help them so much, but I emphasized that I would very much like to do so.

At that moment, the data team was informally born: they lent me a Data Engineer, who was very good with data modeling and already had the company's data context. We formed a nice duo, as I had great experience in collecting requirements and creating visualizations, but it would take me longer than him to build the data models and the C-Level team needed it for yesterday.

The result? The executive team was happy with the new dashboard and they decided to ask us for more things. And more things, and more things. At some point, we had people from all over asking us for information, some even outside our scope or from databases in our domain.

I spoke with my director, and we agreed to add a third person to the team, on loan from the Data Platform team. At this point, we also had an additional fourth person who worked on some sporadic tasks.

The requests continued apace: create new dashboards, add new metrics to existing dashboards, create filters, make dashboards faster, and add information that was previously non-existent in the database.

We did all of this in Tableau, from scratch. Sometime later, we started to recreate all the dashboards in Metabase, as per a top-down request given the difference in invoice sizes and the preference from some influential people on the team who liked the latter better.

The rework was significant, as by this point we already had numerous active users and hundreds of charts. However, I decided to frame it as an opportunity to learn about yet another BI tool, adding to the pile I've already worked with. Tableau, Metabase, Superset, Looker, and Google Data Studio (Looker is by far my favorite in case you are curious).

As the needs were growing, we brought in a fourth person, this time by opening a position and patiently hiring until we found someone we knew would be a great fit.

And, during an organizational restructuring, I brought my managers a document describing the team's mission, objectives, indicators to monitor our progress, and a plan of initiatives to be carried out in the coming semesters. With that, the Data Analytics team was finally officially born with 4 full-time members.

In the Data Engineering section, I tell a curious story about my time at BairesDev/WinDifferent, in the early days of our Data Analytics operation, when we saved a highly critical dashboard from exploding on the CEO's birthday! (Certainly, the most unforgettable gift he wouldn't want).

Many other adventures followed this one, but there is no particular one I would like to mention here. For me, the key thing here is that this experience gave me confidence in managing people even when I'm not fully aware of the technical details of everything that has to be done - which was a major concern for me at a certain point in time.

It was a great leadership exercise as of all the people on the team, I was without a doubt the one with the least data knowledge. However, this was never an issue and the team worked in an extremely productive and complementary way.

Because I was the first on the data team and had a lot of business context, I was able to help the team focus on the most urgent and important items to unlock other teams, generating value for the business as a whole. In addition to this, I also had a good database context, which was simply chaotic and undocumented, with hundreds of duplicated, deprecated tables and columns with missing or inconsistent information.

Having this context helped me speed up new members' onboarding, giving them information in seconds that took me days to figure out, and also in creating accurate, well-described data requirements.

At times, when the team was bogged down, I performed some tasks myself as I enjoyed getting my hands dirty and making things happen.

To sum up, creating your own career opportunities and surrounding yourself with people who have more knowledge in a specific area is a great way to develop both leadership and technical capabilities at the same time!