For a year, we’ve been working on building the first iteration of the data team at Alan. To decide what to build and how to build it, we first worked on the why.
Why does Alan need a data team? AI is not our core business today. Yet, we believe that Alan will not be able to fulfill its crazy ambitions without a strong data team. So why?
There are generally two types of missions that the data team fulfills: create Insights, and build Products.
Insights are about creating and sharing results, serving three purposes:
- Decision making: Across Alan, people need to quickly make impactful decisions. We create or empower the team to create the minimal results that make those decisions possible.
- Objective setting: At Alan, we use OKRs. We believe the quantification of success can be a powerful driver to align teams. We need to know what objectives are trackable, design the corresponding measures, and define the level on the measure that corresponds to success.
- Monitoring: Alan is growing really fast. We need key indicators that we follow regularly to spot any problematic trend.
Products are about creating value in the form of product features for our customers or tools for internal users. Features are typically developed in collaboration with the engineering team.
At Alan, we currently have 5 areas of application of our missions:
- Growth: From sales ops & support to marketing analytics and tracking of attribution.
- Product: Product decision support, user behavior tracking, data-powered features.
- Insurance: Monitoring of rentability, reports for clients, pricing simulation, fraud detection.
- Operations: Care (i.e., Customer Support) analytics, operation tracking, data-enabled processes (OCR, NLP, etc…)
- Business Monitoring: Definition of KPIs at all levels, monitoring, reporting.
Teaching how to fish
The role of the data team is to make sure that those missions get accomplished, but we can’t fulfill them alone. We believe that, with the right tools, the person who asks a question is often the best suited to answer it: it maximizes context and minimizes the need for communication (which usually incurs information loss). In this context, the data team has an additional duty: empowerment.
As we grow, we expect the data team to do a smaller and smaller part of the analytics work and focus on new, broader and more complex problems. This requires constantly building for the future so that what we do today can be done faster tomorrow, and potentially by people outside of the data team. We build tools, infrastructure, processes, and good practices to scale out our work.
Bring in the Full Stacks
After laying out our mission, we need to assemble a team ready to tackle these challenges. It’s time to decide what profiles we’re looking for!
The data practice and profiles can be partitioned in three fields:
- Data Analytics: The ability to connect needs with the data that should be created, transformed and processed.
- Data Engineering: The creation and maintenance of systems that handle data, at scale.
- Data Science: Advanced stats, modeling & machine learning.
To build the first iteration of our team, we’ve decided to look for full-stack data profiles. “Full-stack” does not mean being an expert in everything. Full-stack means being at ease with the full spectrum of the three fields above, having experience using a variety of skills, and being curious and eager to learn about the skills you don’t have.
Why? One of our main values at Alan is ownership. We want to give people problems to solve, not solutions to implement. Full-stack profiles help build long-term value by solving problems with maximum context. When you read the obscure documentation of an API to extract some data, it’s very powerful to know what tools or results you will want to build on this data, and why. Having one person own the full cycle creates solutions that are better suited to the problem and more durable.
Building a team with full stacks is a high-cost, high-reward adventure. Recruitment of full-stack profiles is extremely difficult, especially since the field of “data science” at large is fairly new. Yet, this has paid off greatly so far. The two main concrete advantages that we see everyday are (i) the adaptability of our team, which can stay nimble and contribute in very different fields across the company (ii) the long-term value that we create by building our core tools that help us scale our work and the work of every Alaner.
Change is desirable
This is our first iteration. We expect this vision to change as we grow in order to adapt to our new environment and learn from our successes and failures. Yet, we believe that clearly outlining this vision now is a key enabler to rapidly grow our team.
If this vision speaks to you, or if you want to help us design its next iteration, please drop us a line at email@example.com.
If you’re interested in technical subjects that make this vision possible, go check out our article on the solution we’ve built to share prepared data with the team.