Self-serve is about empowering your coworkers

10 days ago, I had the privilege of hosting Jean-Mathieu from Datadog in our webinar. We discussed the Analytics strategy he put into place to support the company stellar growth, from 200 to 5000 employees. Click here to access the replay!

Most of the time, when we think about self-serve for analytics platform, we're looking at it from the user's point of view, not so much from the data creation side. Jean-Mathieu and his team built a data infrastructure that could handle over 5000 employees - that's no small feat! One of the pivotal factors he attributed to the analytics team's ability to support this growth was the time and effort they put into making sure other engineering teams could easily push data into the platform.

This sparked a thought in my mind.

If dashboard tools are indeed an extraordinary means to liberate data analysts from the incessant day-to-day queries from business users across all entities, what is the equivalent of letting engineering teams create their own pipeline for the last mile analysis?

I think training and documentation are key. Here are some more actionable suggestions that data analytics teams can use to remove the bottleneck effect.

Data is a skill, not an innate trait

I've frequently encountered data teams expressing frustration over business users' apparent inability to effectively utilize data.

It's crucial to understand that data proficiency isn't necessarily an inherent trait; rather, it's a skill that can be cultivated and honed.

In many teams, the catalyst for fostering a data-centric culture has been training. Fortunately, implementing an initial training program can be less resource-intensive (both in terms of $$ and time) than one might anticipate. However, it's a strategy that works effectively for small teams and scales remarkably well over time.

Here are two training session suggestions that could prove a very high ROI:

1. Bi-weekly group sessions for newcomers: This is an opportunity to introduce the data team, familiarize newcomers with the primary BI tool, and ensure everyone knows where to find and how to access data. Not only will this earn you empathy points from your business users, but it also ensures that everyone knows where to start their data journey.
2. Bi-weekly "Happy Hour" sessions: These are open forums where individuals can ask questions and seek clarifications. Participants can ask their queries in real-time, and solutions can be found collaboratively. This not only allows you to connect with your users but also enables you to train them by demonstrating solutions in real-time. It's akin to pair programming, but tailored for self-serve data scenarios. If you want a live example, you can check this Gojob story.

An example is worth a thousand words

In order to empower individuals to tackle more complex topics or questions, you should also create and disseminate more content.

Allow your business users and data analysts to navigate through a history of previous work and accompanying documentation. While the end result holds significance, the intellectual journey to reach it is even more crucial, as it equips individuals with the ability to replicate the process.

“If you give a man a fish, you feed him for a day. If you teach a man to fish, you feed him for a lifetime.”

This can be a 2-step process to ease adoption for you and your team:

First, ensure your work history is readily available in a searchable format. This could be a folder filled with slide presentations, a Notion page brimming with examples, or — even better — notebooks that encapsulate both data and context.

Second, allocate a few hours periodically to create and refine documentation and tutorials. This investment of time will compound over time, providing immense value. Remember, these resources don't need to be flawless; their existence is what's important. With this in mind, set aside dedicated time to create additional tutorials that enable individuals to access more complex data.

At Husprey, we opted to develop our Query History feature, enabling our users to search for specific sql patterns and duplicate queries. This feature both provides the technical examples and necessary context around to ensure users grasp the entire process.

Identify and nurture your data champions

As you progress, you'll begin to notice some individuals are developing a greater affinity for data exercises than their peers. These are your data champions.

These champions should feel as though they're part of an exclusive club where collaboration is not only effortless but also actively encouraged. Over time, they will evolve into your ambassadors within various teams, upholding best practices and handling basic day-to-day training.

As teams expand and the need for data expertise increases, it's not uncommon to see these data champions choosing to transition into data analyst roles. This transition is a win-win situation. You gain data analysts who already possess a deep understanding of your business, which is a perfect recipe for effective data analysis. Chef kiss.

A-tip: Consider sending out dashboard or query execution statistics and user rankings every quarter. A touch of gamification can go a long way in boosting morale and fostering a healthy sense of competition. For example, Datadog analytics team implemented badges in Metabase!

To amplify data usage, self-serve is indeed the linchpin.

However, to enhance the efficiency of self-serve, it's crucial to establish a virtuous cycle where analytics can be utilized daily without your constant involvement!

This can be achieved by training all users, sharing replicable examples, and nurturing the emerging champions within your ranks.

What strategies can you implement to ensure your data team isn't a bottleneck, but rather a catalyst for data-driven decision making?