Data Analyst Interview with Ina. Data science from the start
Here at 365 DataScience, we believe that the data science careers such as the Data Analyst are understood through the experience of the analysts themselves. That is why we have a section devoted to interviews with professionals. By reading their life stories, we believe you will relate to one or several of them and get your career on the fast track.
Our first interview is with Ina. She is a Data analyst at an online video consumption platform and this is her story.
Could you briefly introduce yourself ?
Hey, and thanks for the invite. I’ve followed your website right from its inception and am pretty excited to be featured on it. I hope my story can be useful to aspiring analysts.
So, I am a graduate from a High School of Mathematics and a final year university student in Applied Mathematics. Currently, I am working as a Data Analyst at a start-up. The company is dealing with online video consumption and we have tons of data flowing to us at every moment. I’ve been on the job for almost a year now and am feeling more and more involved and knowledgeable. Instead of being in the field of mathematics, I feel like I am a part of the data science field now, so your target group, right?
Yep, that’s right! Okay, so data science became quite popular just recently. When was the first time you heard about the field and how did you end up working in it?
Okay, let me just clarify, for administrative purposes, I am still a student. However, I have finished my regular studies and have just one external exam to take in order to get the long-anticipated diploma.
I started looking for a job during the summer of 2016 and basically went onto a job hunting website and wrote ‘mathematics’ in the keyword field. The search engine spat out a bunch of positions — all of them were full of words I hadn’t heard before like: dashboards, BI solutions, QC analysis, ETL, data warehouse (I knew this one). During my interviews for these positions, they would always ask me what I thought the responsibilities would be. In the beginning, I was super clueless and would say something along the lines of: you get some data, you analyze it and pass it on to the next guy. Then they would explain the job to me, but to this day I think it was empty words. The HR people were very bad at explaining these positions.
During those interviews, companies would ask me questions about Excel, SQL, how to read a graph, logical questions, simple mathematics; or even to write a piece of code in whatever language I wanted.
So, finally, I interviewed with the start-up I am currently working at. The name of the position was ‘Business Analyst’ but is essentially Data Analyst. So, I showed them I could do many things (Excel, PowerPoint, basic programming) but still had no idea what the position was about. They told me that they needed smart people with certain capabilities who would be excited to learn the rest. At the end of the interview, one of the managers started writing on the whiteboard: A/B testing, video engagement metrics, conversions, gamification analysis, etc. as main responsibilities.
Out of the whole application process, two companies offered me a Data Analyst position on the same day — a big multinational company and a start-up. For me, it was a no-brainer — I went to the start-up.
Sounds like you found a place you would fit right in. Cool! So, what is it that you actually do as a Data Analyst?
Well, there is this big database. I run queries and create reports based on the output. I am doing everything on MATLAB and if you are familiar with MATLAB, you would know that the graphs are extraordinarily ugly. So, I take them, put them in a PowerPoint presentation with a detailed description and include my own recommendations. For example, this is strange; that is interesting; in my opinion this is because…, etc. Then I email the report to all managers and if they have further questions, they contact me. The better I get, the less they contact me, which makes me proud, but also a bit isolated, haha.
To get back to the question — all the data I am working with is internal and classified. However, I create reports for just about anything you can think of. And if I don’t understand a metric, someone explains it to me and only then I make the analysis. So far, I am analyzing whatever by managers ask me to. However, in the future I will have the opportunity to initiate my own research.
Is there some daily, weekly or monthly routine or any kind of hard structure?
I do not have any daily routine. I have a norm — I must produce a certain amount of analyses. They give me, for instance, twenty tasks at once and I have to create at least X reports per day. Apart from that, I have start-up specific duties, but they are not related to the Data Analyst position.
I suppose the only solid thing is that every month I need to update the same report that goes to the investors. Everything else varies.
You mentioned MATLAB already, but apart from that which tools and software are essential to your work?
There are three of them: MATLAB, Excel and PowerPoint (we have company specific software as well, but nothing data related). All the analysis is done in MATLAB (and sometimes Excel) but I report in PowerPoint as it has to be manager-friendly. I would say that my time is equally divided among those three but I enjoy MATLAB the most, as that’s the place where the data lives. Actually, some days I just use MATLAB all day, because it’s super cool.
Did your university prepare you for the job or did …
(she intercepted me at that point. The full question would be: Did your university prepare you for the job or did you carry out your own research? Did your employer provide you with some additional training? Or was it some sort of combination of all of those?)
100% not. The company taught me everything and whenever I need something I go online to look for it.
How is the position organized? Are you a part of a team or is the team yourself?
Everyone has this enormous desk and is sitting alone. I am the only Data Analyst but in my room, there are 5 different people. I would say that I am working directly with the three managers, so I don’t interact professionally with my other colleagues.
What about the working hours?
I work from 9am to 7pm. But it’s very flexible. I may come at 10am and stay until 8pm, or if I have a doctor’s appointment or something, I just leave and come back. It’s a start-up 🙂
Every job has its ups and downs; can you tell us what you find least and most pleasant in your job?
Least pleasant… there is this thing that my boss says only I can perform well enough, so he gives it to me. It is something he used to do on his own for years, but now I’m the person for it. It is super tedious but someone has to do it. I know he has more important things to do, so I do not moan about it as it is not that much anyways. It’s just boring.
The most pleasant thing about my job are my colleagues, for sure. Once again, it’s a start-up, so everyone is smart, hard-working, open-minded and we get along quite well. I would not trade it for a higher paying job in a big company, just because of the people!
This sounds super nice. To change the pace a bit, what is the most stressful situation you’ve been in?
It was this week, actually. One of the managers (the one I work with the most), went abroad and I could not communicate with him. I had to update a yearly report, based on a report that he created one year ago. I wrote the code, ran it and the data was extremely different and I mean… hundreds of times bigger or smaller than the previous report. As you can imagine, we had to compare the two reports and they were completely inconsistent. I started on Tuesday, my manager was gone Wednesday and Thursday, and my deadline was Friday. I couldn’t just call my boss and tell him: hey, boss, my code is not producing the desired result, what should I do?
It was up to me to fix it.
But the data was awful — full of mistakes, blanks, faulty data, etc. I spent the whole Wednesday fighting with it but it was futile. So… after work on Wednesday, I went to the best programmer I knew (a close friend of mine) and asked him to check my code. He confirmed that it was perfect, haha. The next day I reran the report, using my code for the previous period and it was consistent with my report, but had nothing to do with the previous one. I prepared the amended report, although I knew they wanted to see the same data. On Friday, my boss came back and told me it was strange ‘af’ and I was right. In the end, we decided that we have to use some different input, or manipulate the variables in some way. I am still working on it.
What made the situation truly stressful was that I had 4 days to perform a job that I could do for 1 day, but the data was wrong, and I was completely alone with no guidance.
Oh, wow. Sorry to make you relive this. How about a moment you were super proud with what you’ve accomplished?
Super proud is a bit extreme for short projects. Sometimes I feel proud. I would say it would be whenever I optimize the code of previous coders. There was this one time, where the code was running super slowly and you had to put a lot of input manually. I automated the whole program and felt super proud. I did not really need to do it, but wanted to. Now this is very useful for me, because I am getting faster at automating most of the processes I deal with.
Nice, nice! Okay, tip of the interview time. Is there a nifty tool or a method that you discovered or were introduced to in the course of the workflow which you now can’t live without?
The only thing on my mind right now is automation of processes. If I had to say one thing — just read about automation. It’s the best.
Every time we have those interviews, we finish with some nerdery. What is the one nerdy thing you would like to share with the world. Doesn’t have to be data science related.
As my company is in online education, I regularly stumble upon some interesting school-level math. Recently, I saw the Nine-point circle. The interconnection of the points in a triangle astonished me. Basically, the theorem states that the 3 midpoints, the 3 feet of the altitudes and the 3 midpoints of the line segments from the orthocenter to each vertex, all lie on the same circle. And that’s true for every triangle! Not what a Data Analyst would tell you, but still truly remarkable!
(according to Wikipedia, Euler and Feuerbach contributed, but Terquem finalized the theorem and its proof, therefore its name: Euler’s circle, Feuerbach’s circle or Terquem’s circle)
Originally published at 365datascience.com on May 17, 2017.