About project management

During my last months at Mercado Livre, I experienced ups and downs as we all do throughout our careers. One of the hardest points for me has been the lack of a Project Manager figure among us. This is a position and responsibility that falls therefore onto the shoulders of the Project Leaders (EMs being among those).

The fact is, I have worked for years with them but never really as one of them. I know about Jira, Agile, Kanban, sprints, roadmaps, and so on but never have I had the opportunity to handle this on my own.

In 2022, we managed to deliver quite a few projects but not in the most efficient way. Feedback on this issue has been recurrent in my 1:1s, and I decided to dive deeper into the topic to shake things up and improve our workflow.

This post is, (hopefully), the first of a series around the project management topic from the eyes of an EM (or PL).

2022

During the last months, my team and I used a simple Kanban perspective for task management tracking. One backlog, one mega sprint (3 months), and some tasks dragging over the columns of our board. No refinement, no planning. Very few retrospectives. It has its advantages but from what I saw also several drawbacks.

We did not allow time for any refinement or grooming. It locked us into a task-pushing mindset, more old and mechanical than modern and dynamic. The massive retros we did, took a large amount of time and were not followed by actual actions. Finally, the lack of overview between main projects, side projects, and small side tasks over time had us lose a global view of where we were and still had to do at specific points in time.

23Q1

The solution I am deploying in this 23Q1 is a classic Agile one, based on short sprints (2 weeks) with one planning and one retro/review. On top of this, each project will have one (or several) grooming task(s). Those are sessions dedicated to asking tough questions, talking to external teams, laying down our roadmap, and dividing with child tasks as much as possible so that the sprint-planning sessions focus on simple task-picking based on priority. Of course, we will have ongoing larger tasks that persist from one sprint to the other.

At the end of every two weeks, we plan to do a retro of what was delivered or not, and a review of what worked and what did not. Tasks and management. On my side, I still have to plan for a follow-up on the actionables over those feedbacks. Since they should come in small quantities, I hope to manage them well.

On the side, we have registered the whole process update on an internal document and try to keep everyone engaged in writing thoughtful descriptions for each task.

The last important item I can remember now is that we will try to track as much as we can "intruder items" during our sprints. Those little things that come out of the blue and delay projects. Let's see if we can spot patterns within intruders or even defects in our methodology and improve.