We Regret Being Data Analyst Interns
Who are we? What do we do as Data Analyst Interns? Why do we regret being Data Analyst Interns?
Hey, I’m Diana!
Full Name: Angela Diana Suwanto
Tribe: Account and Platform
University : Nanyang Technological University
Major : Mathematical Sciences (Statistics)
Before I started my internship, I have only understood that Data Analyst should be able to transform raw data into meaningful data. As I have worked on several tasks, I figured that the process is never as smooth as what I thought it would be. However, getting hands-on experience by working as a data analyst has taught me to be patient, thoughtful, and most importantly, perseverant.
Pain is never a burden, but soon to be an excitement.
Ola, I’m Denise!
Full Name : Denise Sonia Rahmadina
Tribe : Content
University : Institut Teknologi Sepuluh Nopember
Major : Informatics
At first, I wasn’t quite sure about what mindset and traits a Data Analyst Intern should have. But as soon as I was involved in a few tasks, I found that not only technical skills matter on being a Data Analyst, but one also has to be curious, persistent, and make efficiency their top priority. These are the things that I always keep in mind every time I’m faced with challenging tasks and I believed that these challenges are a way to make me grow.
We don’t grow when things are easy, we grow when we face challenges.
Hello, I’m Nala!
Full Name : Nala Krisnanda
Tribe : Communication
University : Universitas Indonesia
Major : Computer Engineering
My interest in changing data into something useful intrigued me to be a Data Analyst, although I only know a bit about what Data Analysts should do. After I got into the internship, I realized that not only technical capability is needed on being a Data Analyst, but also statistical skills and communication skills such as reporting & presentation. I’m excited to move forward to being a good Data Analyst intern and learn bazillion things from this experience.
THE HUSTLE AND BUSTLE
What do our daily activities look like? We have summarized our hustle and bustle in five steps following the Exploratory Data Analysis (EDA). Along with each step, we include their resemblance to the process of serving sushi as sushi prodigy. Let’s check it out.
1. Planning
What are the requests from the product and business team? What kind of information we (and they) would like to see? What are the tools needed to achieve those objectives?
We, as Data Analysts, are the strategic partners for almost all departments. Therefore, it is inevitable that we receive requests which contain information about what they would like to see; problem detection, the performance of a specific product, and many others. We have to align our goals (and plans) with theirs by deciding on the best tools such as analysis tools, visualization software, and many others.
It is analogous to a sushi prodigy who is researching for the customers’ preferences. Even though he (let’s assume that the sushi prodigy is a male, not intended to be sexist) also has his preferences, he still needs to align with his customers’ preferences to benefit both parties. Thus, in the planning step, he needs to decide what kind of sushi he would like to make and what the best ingredients are.
Thus, if we have had requests from other teams, we are supposed to determine if it is rational and executable. It is best to discuss it as much as possible with others. Discussion is the key to achieve shared goals. Keep a clear communication on the progress with others as well.
2. Data Extraction
Now, we have plans and goals. What are we going to do with them? Yes, let’s execute them.
Given a dataset, do we need to query them all or should we just extract some of the features (columns)? The cost came upon your actions. If you query all unnecessary features, not only it will be expensive, it will also mess up the view.
Similar to a sushi prodigy, after the planning steps, now he came up with a bunch of shopping lists. He encounters a dilemma: should I buy all fishes or should I only buy the fishes required by the recipe? We will definitely suggest him to do the latter. Why? It is clear that if he buys all the fishes in the market, not only it will be expensive, it will also go in vain.
So, how should we distinguish between important features and less important features? We should ask the respective requestor on what features would be of the utmost importance.
3. Data Cleaning
Maybe you have heard that real-world data are messy. They really are. Missing values, outliers, and inconsistent naming convention are some roadblocks that we encountered during the process. How are we supposed to treat them? The answer depends on some aspects.
The most common solution is to group the messy data together whenever possible and see what we can do about them. We might also remove them or set them aside first.
It is akin to fish that he bought from the market. Assuming that we are going to make salmon sushi, then we need its flesh. Should we remove other body parts? It really depends on a lot of aspects.
4. Data Visualization
With the assumption that the rice has been cooked before, the sushi prodigy needs to have a clear vision on how to serve it to customers. Even though it might taste scrumptious, but if he creates awful-looking sushi, the customer might end up throwing it away instead of indulging the sushi.
Visualization is as important as the other steps in data analysis. Creating an unbiased and appealing chart takes in a lot of effort but it is truly satisfying. A dashboard is a collection of charts, tables, and other information. If sushi is a chart, then a collection of various types of sushi is a dashboard.
Creating a dashboard is sometimes a roadblock for us. Small details such as fonts, sizes, and alignments could be major problems. Another challenge we face frequently is when creating an unbiased and unambiguous chart. We need to present the right visualization so that the previous steps of our work will not be unavailing.
In our opinion, the best solution to these roadblocks is to have your dashboard reviewed by other people. We could see their reactions and comments that will be absolutely beneficial for the improvement of our dashboard.
5. Analysis and Reporting
On the final step, we need to review the list of information, choose the right method (analysis, e.g. statistical analysis) to acquire the information and serve it. Sometimes, our analysis might contradict our expectations. Another time, we are unsure if we have analyzed it accurately and if the quality is preserved. Our best solution would be to trace back all the steps from top to bottom and ask ourselves if the result has explained all the information that we need.
THE REAL REASON
Plan, extract, visualize, analyze, validate, evaluate, and repeat. Isn’t it tiring? That’s why we said that we REGRET being Data Analyst interns. We’re kidding. The phrase is actually referred to as our sadness because we are just Data Analyst Interns at Tokopedia and the internship is ending very soon. Tokopedia has nurtured us in so many aspects. There are several other reasons which cause us regretting being only data analyst interns.
Firstly, Tokopedia provided us with endless opportunities to be involved in a team and contribute to some impactful projects. Our fellow teammates listen to our ideas and insights attentively during every meeting and weekly presentation. We felt that we are more than just Data Analyst Interns, but real Data Analysts.
Moreover, Tokopedia gave us a new family. All Nakamas here are so welcoming and really warm to us; we are considered as part of the family since our first day of work. Although we are here as interns, however, we feel attached to both our team and other teams. They are more like a friend rather than just fellow workmates.
Last but foremost, Tokopedia provided us a superbly comfortable, lively, and cheerful work environment. Unique office with open space, arcade games, work stations aside from our desks, mind-boggling events and seminars truly gave us a whole new adventure about work life.
Written with 💚 by Angela Diana Suwanto, Denise Sonia Rahmadina, Nala Krisnanda
