What does a Data Analyst do?
Raybeam is always looking for good data analysts. We’re confident that with enough data analysts we could change the world. But because the job description seems vague and wide, candidates and clients both ask:
What does a data analyst do?
What skills does a good data analyst candidate have?
In this article we’ll explain what a “data analyst” does and doesn’t do, and the skills we expect.
What do they do?
Every company has people whose job is to combine data from different sources to answer business questions. These people aren’t considered technical by software engineers. But they still exercise a lot of creative, business and logical judgment, on every project. The people who do this work are called “data analysts.”
Data analysts figure out how to make new good data, specifically new records, out of a company’s raw data. Data analysts don’t analyze so much as they remix raw records into something new, closer to the business need.
You hire data analysts to break your data out of artificial silos. The ERP, CRM, ecommerce, PLM, marketing, web analytics, inventory and fulfillment, sales, and privacy systems all have their own picture of the customer. But they don’t all talk to each other.
Data analysts exist because:
You need to integrate your data before your company can thrive.
Software engineers have other work to do.
Data analysts glue your data into a coherent whole, building a “semantic layer” out of raw data that business people can use. Data analysts may need to merge two spreadsheets, ten databases, or five dashboards, but they won’t know until they take a look.
The people who do this job end up with a mix of skills, including data quality, data modeling, visualization, and business analytics. Hiring good data analysts requires not losing sight of the forest for the trees.
Very few data analysts learn all the skill-sets required to navigate the entire space between your applications and your business. Most learn a couple components well and consistently. But software engineers may try to build a suspension bridge when a rope swing is all you need to start. You hire data analysts to build the right-sized solution for the problem.
Data analysts are more than critical to the way organizations work.
What doesn’t a Data Analyst do?
Data analysts build an organization’s semantic layer. We expect them to develop repeatable, consistent and precise logical formulas - proofs - as quickly as possible. We don’t expect them to do more than that, because its a big job.
At Raybeam we hand the automation of those proofs off to engineers. We expect our engineers to build tools and platforms to make semantic layer construction easy. Data analysts then are free to build the semantic layer as fast as they can, with as little friction as possible.
We think this division of labor lowers the daily stress on data analysts, and they get the best use of their skills. Data analysts find the logic, and data engineers automate the logic.
What does Raybeam look for in Data Analysts?
At Raybeam anyone who has one or more of these skills could technically be a “data analyst”:
Visualization and report development
Business, statistical or other quantitative analytics and storytelling
Data science
Data modeling
SQL integration script development
Data quality analysis
Database discovery
Data architecture
A senior data analyst has more than one of those skills, plus in-depth experience in one or more and likely a preference for at least one. They may also have some data engineering skills, and that’s even better. Very few people end up with decent skills across the board. The more experienced analysts have at least tried and failed.
We also look for the willingness to try and fail. None of us are born knowing how to merge datasets, or explain or visualize data so someone can use it. Good data analysts are confident they can make datasets come together, or explain why not. They may eventually be wrong, but a data analyst is willing to try to find a join.
A third feature-set we look for in data analysts is an opinion about what works and what doesn’t. Successful execution requires we optimize business impact, reusability, and feasibility. Semantic layers are made up of lots of individual decisions about business logic, reconciled into a consistent whole. Good data analysts can execute the big picture, while understanding who’s using their data right now.
Come work for us
Data analysts play a foundational role creating and using the semantic layer in every business. We need them to glue together data sources for business users.
They may have to do all sorts of creative, logical, technical, and business analysis work to meet their user’s expectations. The integrations they drive hold companies together. At Raybeam we know they’re critical to an organization’s success, and we work to ensure they can focus on their best work. Come work for us.
