Data Roles - Why Data Scientists & Analysts are like Doctors & Nurses
Disclaimer - These views are based on my five years experience as a data analyst in two tech companies.
The modern data industry is an evolving one filled with many confusions. A common confusion is on the difference between a data analyst and a data scientist. For this, I like to draw parallels about their differences to the roles of doctors and nurses.
The Analogy
Doctors and nurses, like data analysts and data scientists, have similar foundational skill sets.
However, doctors and nurses provide very different services in the medical industry, just like data analysts and data scientists in the data field.
Nurses and data analysts focus on day-to-day service provisions ( e.g. taking care of patients / working with stakeholders ), while doctors and data scientists work on more advanced interventions ( e.g. surgeries / predictive modelling ).
Doctors and data scientists can do some nursing and analytics tasks, but because nurses and data analysts do these tasks more often, they can be more proficient in in those tasks than doctors and data scientists.
Doctors generally earn more than nurses, just as data scientists generally earn more than data analysts.
Doctors and data scientists are more prestigious roles, while nurses and data analysts are seen as non-core “support” roles in some companies.
No analogy is perfect
Unlike the medical industry, our data industry does not have a central body to officiate certifications and titles to data professionals. Hence, companies have the liberty to hire or rebrand their data analysts into data scientists, all while not making any significant change in their job scopes or remuneration packages. To be clear, the “data scientists” I mention here are sometimes called “data scientists (machine learning)” or “machine learning engineers” in some companies.
Many data analysts also see becoming (machine learning) data scientists as a career promotion. They see this as a way to get more challenging ( technical ) problems, and a very direct way of boosting their salaries. However, I think most nurses don’t aspire to become doctors. And just as the experienced nurses provide great value to our medical industry, senior data analysts are also important in setting the tone and foundations of an organisation’s analytics culture and capabilities.
Aspiring data professionals
A ( machine learning ) data scientist requires a lot more technical knowledge in areas like engineering systems and working with production-grade codebases. Among their tasks, data scientists will at times need to create and deploy large-scale prediction models as real-time solutions to their company users. In other instances, data scientists need to set up and test hypotheses to understand the underlying impact of certain product and business decisions ( This is at times done by product data analysts on new product features too ). All these are very challenging technical problems.
On the other hand, data analysts work more closely with stakeholders to figure out how analytics can be used to amplify the capabilities of different functions. On some days, data analysts might need to work with stakeholders to define the correct metrics that accurately measure the performance and value of a new product feature. Once that is done, the data analyst has to then work through their company’s data marts / data lakes to actually calculate the performance metrics that needs to be monitored periodically. When analytics break, the data analysts also need to figure out how to troubleshoot the issue, which could involve working with frontend engineers to figure out if the app events are firing properly, or working with backend engineers to figure out any potential server issues. While these problems look less technical, they are not any less challenging, and definitely, their impact may not be any less either.
Business owners
While not perfect, the doctor-nurse analogy reminds business owners to be mindful of the analytics roles and tasks that their organisation actually needs. A smaller business owner who insists on replicating the predictive model from large tech companies is akin to someone wanting the medical treatment his friend had, without understanding how that treatment affects his own body condition. A small firm with little or messy data will not benefit from forcing a data scientist to build a predictive model, because a quality, useful prediction model is still dependent on the organisation having enough, quality, labelled data sets.
This analogy only covers a subset of data roles, as there are more technical data roles like data engineers and analytics engineers, and less technical data-related roles like business analysts and analytics end users who consume data for their work.
Nonetheless, I hope this analogy helps a business owner understand the importance of hiring the right people for the right ( data ) tasks, so that their organisation can get the most immediate impact out of their data efforts. The business owner may first need to improve the overall quality and volume of the data within the organisation ( The Sushi Chef Analogy ), or to hire someone to better orchestrate their current analytics workflows ( The Orchestra Conductor Analogy ). Or what the business owner needs is to train their stakeholders to be more data literate. Not every data solution is about deploying an expensive prediction model.