Deciding which metrics to be focused on in data analytics and the roles
From August 2022 until May 2023, I am assigned to the Subscription team at the company I am working at. This service is a brand-new thing as part of the company’s grand plan to achieve positive revenue as soon as possible and increase the “loyalty” of the users at the same time. To be honest, it’s not my first time being assigned to handle an analytics team to provide monitoring, reporting, and data analysis for a new feature in a company. Based on my experience, the most difficult part is not the things I have mentioned above, but deciding which metrics to be focused on.
First case: A small e-commerce website.
In 2014, I got hired as the first data analyst ever hired in a company with an initial “BL” in Indonesia. As a data analyst, I’m very passionate to start my journey, exploring the datasets, generating actionable and impactful insights, and presenting recommendations to stakeholders. In short, our north star metric is Gross Merchandise Value (GMV) with a definition, “Paid transaction gross value in our platform.” Gross means includes any additional cost other than the items paid, such as shipping cost, payment unique code, discount, and so on. It will be called Nett Merchandise Value (NMV) if it does not include the small things.
The first thing thought by me, as a newbie data analyst, “Okay! I will explore the datasets, looking for any characteristics and correlations to the GMV, after that, I will give you the findings and recommendations!”
This is when I learned that it could not be done like that. There must be a lot of variables and factors that affect the GMV. How could I know that my findings and recommendations are valid and actionable if I don’t know what is “the cause”? This is the start of my journey in learning about the difference and relationship between correlation and causation.
As an illustration, as a data analyst, if it’s not yet established, we have to do this before looking for any insights from the data stored by the company. The picture above is oversimplified, it will be more complex when we start to break things from the perspective, let’s say we are assigned to Marketing, which usually focuses on awareness and acquisition, the north star metric should be not GMV, because the direct impact caused by any activities of the Marketing team, can only be felt until the acquisition phase.
Based on image 2, we can see that the method is similar, but the value-focused is different. Simple logic: if we get a higher NRU for the company, we will get a higher GMV with the existing CVR-O.
Scopes of data analysts
That’s why some companies specifically mentioned the “scope” of the data analytics we will do in the current job market. Some of the roles are:
- Marketing Data Analyst / Digital Marketing Data Analyst
- Product Data Analyst
- Sales Data Analyst
- Business Data Analyst
- Customer Relation Data Analyst
- Fraud Management Data Analyst
- Experimentation Data Analyst
- Pricing Data Analyst
- Forecast Data Analyst
- Data Analyst, generalist.
Based on the list above, do we have to focus on the specific scope as a specialist or a generalist? In my opinion, it depends. If we are passionate about data analytics AND marketing things, I think we have to pursue the specialist, so we focus on marketing data analytics, because some of the tools, methods, and other things are very different from the other scopes.
Using Google Trend, for the past 5 years, we can see a difference between each role, it can be used by us, data analytics professionals, to decide what things to learn in the future so we can keep updated with the market needs. But if you have an opinion to be a generalist, it’s also a great thing, it will be harder I guess because we have to keep updated with a lot of tools, methods, business and domain knowledge, to make the hiring manager confidence with our ability when assigned to a specific scope in the company.
In the next posts, I will try to share and elaborate on some of the roles of data analysts listed above. My goal is for all of us to know better about the data analyst and know what expectations we can exactly hope from them.
