A data scientist is a statistician who lives in San Francisco.

https://twitter.com/cdixon/status/428914681911070720

A data analyst is much more likely to be assigned to a business department andnbe the local Excel guru who cobbles together SQL against the enterprise data warehouse, or whatever passes for one. They essentially fill the gap between IT resources in terms of the business's data needs.

The analysis is typically the application of basic arithmetic under the constraint of gnarly "business logic," which is almost never logical.

This is not meant to be pejorative, but represents an amalgamation of the traits I've seen in those who have the title of data analyst at our client companies.

A data scientist tends to have an advanced degree or significant experience in applying statistical techniques to the corpus of business data.

To give an oversimplified example, the data analyst likely answers the question "who are our best customers and what are their traits?" whereas the data scientist answers the question "what traits of customers can predict their becoming a great customer?"

Since we all live in a bubble, I will share mine for context. I work for a BI and analytics consultancy that also has a data science practice and our clients are primarily mid-sized corporations.

My guess (and this is a reasonably educated guess): a data scientist creates models and usually has an advanced math or stats degree (or similar). A data analyst uses models created by data scientists, and often has a business/econ or other undergraduate degree.

Of course this is a generalization, but in my perusal of job openings over the past year, this seems to be roughly what companies mean.

They seem to be mostly self-applied labels you use depending on the audience, in my experience.

Not that a degree in a very specifically named program should be the one and only gatekeeper to an occupation, it's just that these seem to be fairly ill-defined titles that tend to be thrown around these days.

I'm kind of tempted to put "software engineer" on that list as well, though I think people tend to understand that it is more than just programming at least.

Data Analyst is a general run-of-the-mill person who has some understanding of Excel, SQL, massaging the data etc. Data Scientist is a buzzword these days but should in reality entail having an advanced degree in statistics or math, ability to reason about underlying data and associated models.

R v Excel