The assumption that data analysis requires a degree has slowed a lot of careers that didn't need slowing. The reality in most organizations is that the people doing the most useful data work are not data scientists with PhDs. They're operations managers, marketing coordinators, and product owners who learned enough to ask better questions and find their own answers.
What "data analysis" actually means at most jobs
In most non-technical roles, data analysis means pulling a report, understanding what it's showing, spotting something unexpected, and doing something about it. You don't need a statistics degree for that. You need to know how to structure a question, where the data lives, and what to do when the numbers don't match your expectation.
The skill set is: comfortable with spreadsheets, able to build a basic pivot table, knows what a percentage change means, can make a chart that actually communicates something. That's it. A motivated person can get there in four to six weeks of focused practice.
The fastest path is through real data problems
Courses that teach Excel functions in a vacuum don't stick. The ones that stick start with a business problem and work backward to the skill. "Here is a sales dataset. Your job is to find out which region underperformed last quarter." That kind of structure builds intuition, not just syntax memory.
The best self-learners pick a dataset from their own job and immediately start trying to answer questions they already care about. The learning is faster because the motivation is real and the feedback is immediate — either you can answer the question or you can't.
Where to start if you're starting from scratch
Begin with spreadsheets. Not because spreadsheets are the best tool, but because they're everywhere and the concepts transfer to everything else. Once you can manipulate data in a spreadsheet — filter, sort, aggregate, visualize — picking up SQL or Python is significantly easier because you already have a mental model for what you're trying to do.
From spreadsheets, SQL is the next most universally useful skill for non-engineers. Most business data lives in a database somewhere. Being able to write a basic query changes what questions you can ask without waiting for a data team to get back to you.