What can you learn in a data-for-good coding marathon?


It was promising to be a beautiful, warm weekend in the city. I had signed up for the DataKind data dive. That meant spending the entire weekend nerding out on data to help non-profit organizations. Friends had invited me to go on hikes, to the beach, to drink beer… you know, fun stuff. Yet, I chose to be in a downtown office room full of other mostly strange people. Was it worth it? Read on to find out.

In a place so tech-driven like the San Francisco Bay Area, there is no shortage of all kinds of “hackathons”, or programming marathons. I really enjoy spending time exploring data and coding, but like everything else in life, we need a balance. That’s why I’m very selective about which events I participate.

Another day there was a marathon to work with weather data, the kind of subject that is interesting to me, given my natural sciences background. I showed up and realized we’d spend the (again beautiful, warm) weekend in a place with little natural light that felt like a dungeon. The participants were a bunch of dudes who may actually live in a dungeon, and who were there in great part because a cash prize was being offered. Nah, not worthy of a sunny weekend!

Perhaps for most people, these events represent opportunities to learn, network, and strengthen their projects portfolio. However, from my observations, programming marathons related to social good are different. People tend to join also because they genuinely care about helping non-profit organizations and their mission.

So, was it worth to join the DataKind SF event that weekend? Absolutely! Here is the feedback received from the organization that our team helped, called Concrn:

“We want to say a yuge Thank You to all the data divers on the Concrn team this weekend. The insights and products you built are amazing and have already helped us look into our strengths, weaknesses, and opportunities. We actually already showed our data yest to a few stakeholders who were really impressed. Thank you all.”

What can you learn, and what to expect from these events? Below I describe some key takeaways, which may apply not only to data dive marathons but also to full-time data analytics roles.

Sometimes, you must help people figure out what they need

This may even sound counterintuitive since these same people, the clients, were the ones who requested your help in the first place, but this is more common than you might think, even in the corporate world. The key is consistent, iterative communication. Otherwise, you may risk working to produce results that can be perceived as missing the point, not very helpful.

It doesn’t matter how much preparation was done, the data won’t be ready

People are usually eager to practice their modeling skills, but before that, you need the data to be preprocessed accordingly. It can be frustrating to realize that you may spend most of the time cleaning data, but this is, as we may not like to admit, a very relevant skill. DataKind is organized in such a way that they have volunteers cleaning and shaping the data ahead of the main event, at least just enough to enable more productive outcomes, but don’t keep your expectations too high on this one.

There are a lot of data problems that are small data problems

The data can be really small. It can have a lot of missing information. This is especially true for newer, smaller organizations that have no staff with data management exposure or established processes in place. You may not be able to do any good modeling with such data, but you still can help to uncover the signal from the noise. You can find such insights, create appealing visualizations, and use them to support the story. Often times, you don’t need to do anything too sophisticated to contribute to the cause or the business.

There will be people with all levels and types of skills, and that is great

I have participated of one of the first DataKind SF data dives, back in 2015, and I can tell that their workflow has greatly evolved. Likewise, I have also become more proficient since then. In 2015, I was more comfortable programming in R, and I had to really focus on what I was doing. By 2017, I had switched back to Python, had built some codebase that was reusable for the task, and had more realistic expectations about the work involved. Therefore, coding marathons can be a good gauge of the level of your skills, especially because you will work with a diverse group of people. In the last event, I was able to interact a lot more with the team, and still get things done. In fact, you will be amazed by what can be achieved in just one weekend!


The data-do-gooders of the DataKind SF Hackathon, April 2017. Photo by @DataKindSF

Overall, I believe programming marathons that promote social good are special and can be very worthy of your time when well-organized, like the ones from DataKind. The “special” part comes from the good energy that flows around the entire weekend. In general, it’s a friendly environment where you meet a lot of nice people. As a bonus, being able to help a great cause will undoubtedly make you feel good about your efforts.


2 thoughts on “What can you learn in a data-for-good coding marathon?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s