Introducing our three roles in Data & Insight at Oda

Have you ever wondered what the differences are between a Data Analyst, a Data Scientist, and a Data Engineer? Maybe you have also wondered what each of these roles do, and what they don’t do?

In this post we will introduce why Oda chose exactly these three roles in our Data & Insight discipline, along with how we define them and what sets them apart. In following posts we will go into more specifics about those roles, straight from the mouth of the people who do the job, some of our Data Analysts, Data Engineers and Data Scientists.

Why three roles?

Every person working in Data & Insight (and in Oda!) is unique and have their special combination of education, experience, personality, passions and strengths. We have education ranging from PhD in Psychology via Master in Space Physics to having worked their way up in domains like marketing. Some people started off as developers and others have more than ten years as consultants specialized in data. And all have experience across a bunch of industries, cultures and countries prior to joining Oda. The sum of that diversity is what makes us strong, and combining these backgrounds make even better solutions to the problems we solve.

We of course take the specific characteristics of each individual into account when assigning people to teams and tasks. Still we find that it makes sense to set some common expectations when we recruit for these roles and support on their personal development. We have chosen the three common roles Data Analyst, Data Scientist and Data Engineer as they describe people with similar responsibilities and expectations in meaningful groups.

We like that they all start with “Data” as we are all part of the Data & Insight discipline with the joint mission to realize the potential in data. It is also a plus that they are broadly used in the tech industry. Our recruitment process is tailored for each of the three roles. We have also formulated expectations for each of the three roles in our joint Product & Tech personal development framework, Compass, that we use in our structured personal development work.

Compass, Oda’s internal framework for personal development, has six steps and two tracks. Expectations have been formulated for each of our three roles in Data & Insight for all of these tracks and steps.

These two examples of tailored processes, recruitment process and personal development framework, are the main reasons why we have decided to keep it to three roles now. But what about specific roles such as ML Engineers and Analytics Engineers? We absolutely see individuals who might fit into such specific titles in these three groups, and really support that kind of specialization to be the best version of themselves. For us, that does not mean that we have to create more granularity in roles, recruitment processes and frameworks for personal development, as we always make sure to tailor when needed.

What the people in Data & Insight have in common

Data is key to everything we do in Oda, and for the organization to be a truly data driven, we need everyone to use data and insight in their everyday work. Obviously, this is not limited to our Data Analysts, Engineers and Scientists, but includes people like business controllers, product managers, performance marketers, senior management and many more. That being said, our Data Analysts, Engineers and Scientists are responsible for making this happen, and have specialized skills. The Data & Insight discipline is defined by the methods applied and the competencies required to use them. Thus, we can have a wide range of backgrounds in the same discipline. Further specialization within Data & Insight can be achieved both through learning a wider range of methods and going deeper into selected methods.

Now let’s look at some of the skillsets and expertise areas we have in common in Data & Insight. The three roles have different strengths and focus within these:

• Deep dive analysis and guiding decisions: We enable well grounded and data driven decision by using our specialized skills in a combination with domain expertise to find how we perform and underlying drivers. This understanding is used when we identify opportunities to improve, not only by building data products, but also any kind of improvements in the domain we focus on.
• Define and test hypotheses: Use of the scientific method is crucial to pinpoint the question we are trying to solve and find the best way to test it. A good understanding of the domain is also key to define the right hypotheses. How do we measure success/performance? A good statistical understanding is important to determine whether we actually can draw a conclusion or not based on what we observe.
• Data wrangling: To be able to work analytically, you must be able to understand the data itself and transform it into meaningful formats for value creation. This includes methods to discover and handle data quality issues, transformations, aggregations and feature engineering.
• Building, tuning and testing algorithms: What is the best method to use for each problem? How should we evaluate the model and how do we interpret the results? When do we need to re-evaluate the model?
• Building data platforms: Technical set up off a combination of tools and solutions to perform analytical task and make data available for a variety of users. Handling of regular and continuous flows of data into data models, algorithms and dashboards that run fast with the expected availability and freshness, and also engineering and handling complex data integrations.
• Communication, visualization and operationalization of results: Explaining the approach used for the entire organization and the consequence of the results. This also include good visualization of data. What is needed to create an impact and how to measure it?
• Enabling and training others: It is not the output itself that defines the discipline. This leads us to an important distinction of Data & Insight; we help create value from data, not only by improving the product itself, but also by enabling other parts of the organization. Thus, it is the Data & Insight team’s job to train colleagues in the ability to apply analytical methods, use insight tools and draw statistically sound conclusions.

What is the difference between the roles?

We repeatedly get questions about the difference between the Data Analyst, the Data Scientist and the Data Engineer. Here is a generalised visualization of how we have defined these roles in broad strokes.

A high level view of how the three roles have similarities and differences in their focus

We have explain these differences in more detail in following posts about each of these roles:

• Rohit Patil on being a Data Analyst (and manager)
• How it is to be a Data Scientist from Kate Kuzmina
• Hanna Heggheim Lee’s experiences as a Data Engineer

Go ahead and check out our open positions. You can also read more about how we are organized in the first post in our series about scaling insights.