I started my journey as a product analyst eight years ago. For almost three years I’ve been interviewing candidates for a product analyst position.
Within this period, my image of the perfect product analyst has changed dramatically.
So, let’s begin from the start.
Three years ago, I was working in a middle-size SaaS company with 3 analysts (including me). At the beginning of my “interviewing” journey, I was looking for someone exceptional with exact skills that analyst should have to work in this exact company.
Based on that my “fantasy”, I had a bunch of requirements for the person that I wanted to work with:
3 + years of experience with a SaaS business model,
in-depth statistic knowledge (especially A/B testing, hypothesis testings, probability density etc.);
metrics (retention rate, ARPPU, ROI, AOV etc.);
MySQL;
master degree in Statistic, Mathematics or similar;
advanced Tableau.
With that saying, I also should admit that I did not work in some fancy company, so not a lot of people desired to work there. So, I think you already figured that I did not get a lot of CVs that suited my requirements.
That’s why I’ve started interviewing candidates that were at least close to them. To my surprise, I understood that my requirements list was not so good.
After conducting 30+ interviews, I formulate my “4 tips rule” that helps me to identify the great product analyst.
Curiosity
You cannot be a great product analyst without curiosity. In my list, it’s the most important thing. It would help if you were interested in the product. Without that, your analysis will be pointless. It’ll be hard to generate hypotheses. You’ll be doing only those thing that product managers ask you (hope, you do not want to do only that). I’m confident that curiosity will help to make your life more enjoyable, exciting and fast improving. It enables you to find the deep-down truth about things.
The good thing is that you can develop this skill in a few weeks. Stay with me, I’ll not ask you to sign up to tons of online courses. It’s super easy.
Next time when you’ll be scrolling Facebook feed, ask yourself:
What is the value of Facebook for users?
Why did Facebook add new reactions on posts?
Why did Facebook add stories?
What metrics will you use the effectiveness of those features?
How would you improve Facebook?
How would you measure those improvements?
You can play this game with any product that you use.
So, as you can see, it’s simple as it sounds.
SQL
I cannot stress enough the importance of SQL. It’s the primary language to manipulate the data. All big companies (Facebook, Google, Booking, etc. ) requires SQL.
If you are not familiar with SQL, it should be the number one task in your “to-do list”. You can start from learning the basics (JOINS, SELECT FROM WHERE, GROUP BY, HAVING) and then add window functions (RANK, LAG, AVG, etc.).
A lot of free online courses have already built-in compiler. So, you do not need to install any soft to learn it.
You can try free w3schools. Or if you want to spend some money to have the motivation, you can try DataCamp.
Do things fast but with a proper level of attention to the details
I understand that it’s hard to measure this skill. It’s crucial to know when to stop improving your level of accuracy. The cost of knowing that the population of the world is 7 812 867 865 is much more expensive than the benefits (sorry, I made up this number). But for some medical analysis, the level of accuracy would be much higher. The cost could be people’s lives. So, you always should know what the proper level of accuracy for the particular task is.
Maybe on that part, you’re asking yourself: about what level of accuracy she’s talking? It’s about analytics, we should have 100% accuracy.
Sorry, but I need to upset you. I worked at 5 companies, and none of them did not have perfect data. Every time there are some bugs, some things that we cannot measure. So, you need just to live with that.
For developing this “understanding”, you should ask a product manager how important the level of accuracy is, how he or she will be using your analysis, what is more important: time or accuracy.
Visualization
I’ll not tell you that you should know Tableau or Looker or Power BI or any other visualization tool. It does not matter. You need to know how to show the story you want to tell. It includes understanding what exact insights show, what type of graph to choose, how to not overload with tons of data. Before any visualizations, you need clearly understand the goal of the report and who will use it. Only based on those answers, you could start to think about what and how to show data.
For improving your visualization skills, I can recommend you to ask more questions to yourself. e.g. If someone shows you this dashboard without any explanations, is it easy to understand it, is it shows what you wanted to show.
For inspirations, I recommend you to follow Ryan Sleeper. You can also check Ryan’s “Following”.
He also has a cool website where he posts a lot of tips about Tableau.
In conclusion
You can see my list of requirements changed to only 4 skills.
I cannot say that it’s only those 4 things you need to have to be the best product analyst. But I can guarantee you that that set of skill is going to create an environment that will speed up your improvements in many time.
