In this post we will talk about at a high level what a prototypical data analyst might look like from a skills perspective. Very few ever actually reach this level. Almost everyone will have some different combination that can still work. The goal of this post is to help you understand what skills are needed and why so that you can gain insight into what potential employers might be looking for. In subsequent posts we will talk about how to land your first role and will also break down each category in more detail.
Technical Skills: This may be the most polarizing topic on this list. It is certainly the most tangible, probably the most glamorized topic, and in my estimation the most overhyped criteria for new hires. From my experience 98% of new hires are not prepared day 1 to contribute much technically. Even if they have had courses that required some technical skills most people can’t come in and take on brand new builds. Instead they need to be given established sets of code that need minor tweaks. Eventually they grow their skills with on the job learning but it takes time. For that reason technical skills aren’t as critical as people think. Yes having exposure helps. Yes being technically inclined is a must. But any good hiring manger understands that for an entry level role establish technical knowledge of their tech stack is not a much. Now when you start talking about mid-level and senior analyst roles it is a much different story, but for your first shot it isn’t a must.
Communication: There are a lot of components that fall into the broad umbrella of communication. Below are a series of questions to ask yourself. Each concept builds on the ones before it and from my experience represents the usual progression I see from data analysts. You likely can’t do one well unless you can do the parts that come before it.
Can you clearly and concisely get your message across?
Can you listen to others and truly understand what they are telling you? Can you ask clarifying questions even when it requires some bravery to do so?
Can you tailor your communication based on your audience? For example how you explain a technical issue to a senior leader is much different than how you might tell your boss. Can you stop yourself from trying to bring everyone into the weeds?
After an analysis can you sort through the noise and find the one or two important messages? Can you resist the temptation to show all your work and the other interesting things you’ve found? Most people can’t absorb more than one or two things especially if they are big ideas they need to act upon.
Can you weave your narrative into other components of the company’s strategy? Can you connect dots for others to see how your findings fit into the big picture?
Note: This one also requires a strong grasp of business knowledge and strategy
Data Savviness: A lot of people don’t distinguish technical skills from data skills. However from my experience they are different enough that many people have one and not the other. When I talk about being data savvy I am talking about an ability to understand the data and how you can manipulate it, aggregate it, or combine it so that it can be used to provide an answer to a question. Many technical people, especially those in IT who historically have been more focused on efficiently moving data, can do all of the technical things to manipulate, aggregate, or combine data but they need requirements from someone else on what the result needs to look like. They have the technical skills but not the data savviness. Data analysts don’t get requirements, they create their own and then manipulate the data. Can you understand what each row represents? Can you figure out how to fit mismatched pieces together? If I give you 3 related datasets all at different levels can you figure out the right level to be the base? Many times you must aggregate or duplicate before joining datasets. Can you think through the appropriate steps? Unlike the technical side where you can be taught the mechanics and they work the same in any scenario, there is no textbook way determine how the data need to be wrangled and pieced together. That is why this skill is a bit more art than science. It can be learned but more many people it does not come naturally.
Business Knowledge: Unless you come from the same industry then you will not have the needed business knowledge on day one. Everyone understands this and that’s OK. What we really mean here is can you, and do you, lift up your head from time to time and think about the big picture. For every job in every industry you must understand your own work to be effective, that is just a given. However think about your current or last job, even if it was something paying minimum wage. Do you understand how your role fit in with the other roles? Can you explain the business model to someone if they asked? Many people can’t, and that is one of the key reasons why they fail to advance their careers. To be a data analyst you must be able to learn both the micro and macro parts of the business. The data you are working with represents the business! Really think about this the next time someone asks you what you do for a living. Can you explain the business or just your own part? If you can’t or won’t learn the business model then you won’t success as an analyst.
Critical/Analytical Thinking: This one is non-negotiable. You are there to help solve business problems. These are rarely straight forward with simple answers. If they were the jobs wouldn’t pay so much. This can manifest in so many ways but I’ll give a few examples. Often times your business partners will have identified a problem but not necessarily a question, or at least a well formed question. They may say something like “Revenue is down, why?”. It’s so wide open that you’ll need to ask questions and come up with some hypotheses to investigate. It could be that you have incomplete or messy data. You’ll need to make decisions about how to fill in those gaps and determine to what extent the result is still reliable. You’ll have to make decisions about how to analyze the data. It is rarely just throwing it into some traditional statistical or machine learning methods. You need to understand what you need to analyze so that you can use the right method. While this might sound intimidating it can also be really fun! The work is cool, it isn’t life or death, and you are always learning. But for any of it to work you must be able to handle ambiguity and solve problems.
User Engagement: I could have lumped this in with communication but I really believe it is a whole other skill entirely. Think about the last time you had to go to the doctor. What was your experience like? If you had a doctor that didn’t make you wait, that seemed to genuinely care about you as a patient, and overall made your visit as painless as they could then you probably had a positive experience. If those things didn’t happen then you probably had a poor experience. Analytics is similar in a sense. Many people have no choice but to use your services. They need data to operate in their role, without you they will fail. However many people don’t understand data & analytics and don’t want to. This can lead to them being anxious. If you can give them a good experience, have the proper bed side manner, then you will gain allies and they will be must more trusting of your work. If you don’t do these things, if you ignore the people aspect and just focus on the data then you will never build those relationships. And guess what, there will be many many times when you miss a deadline or make a mistake. If you have strong relationships then your business partners are much more likely to forgive. If you don’t then they are much more likely to throw you under the bus. Unless you are doing consulting work that you sell to outside clients the analytics work you do will not be directly connected to revenue. You need your business stakeholders to vouch for you and the value you bring. If they don’t then you are in danger of having funding cut or jobs eliminated. The most successful data analysts are those who engage with their users, the ones who don’t generally don’t last.