4 ms·

A data scientist is two things: 1. A business analyst who has traded Excel for a Unix command prompt and a scripting language, because (a) there's more data no

A data scientist is two things: 1. A business analyst who has traded Excel for a Unix command prompt and a scripting language, because (a) there's more data now than before, and (b) data is harder to clean because of (a). 2. A mythical unicorn creature that has spawned a conference, multiple books, and thousands of page views for O'Reilly, Business Insider, Forbes, etc. They're all "data-driven organizations" that double their page views any time the word "data" is whispered. As a result, they parse shitty articles into four ad-driven pages and publish books on "Data Mining Your Cat" or "Machine Learning Models for Organic Dietary Schedules." Here are some better questions that will save you some time: 1. What is this article? Buzzword soup mixed with terrible journalism. 2. How do I become a data scientist? Don't--instead, build a good foundation in statistics and keep coding. You'll be happier and wealthier for it. EDIT: removed a bunch of snark. Not all of it, obviously. ;) EDIT 2: added some snark back. I really hate this "data scientist" meme.

_dps 15y ago
I second your snark, and also find "data scientist" to be analogous to "food cooker", but I've also come to realize that there is a need for branding whatever-it-is as separate from "business analyst" because otherwise expectations from non-technical customers get very confused. I interact with many buyers of buzzword-laden "big data" (snark on that one reserved for another day) and "data science" offerings. If you tell them you're a business analyst they will make many hidden assumptions based on the overall population of people using that title. Among those assumptions: 1. The business analyst will probably know SQL, but if we want to do something like analyze our query logs it will cost us developer time. 2. The business analyst will have domain knowledge about financials and maybe user engagement, but probably not anything else. The business analyst will not be able to tell us, for example, the characteristics of our 99-th percentile slowest page loads. 3. The business analyst is not hip and modern, because everything is about the "big data" these days. Realistically, the language of modern business data applications requires a name for someone who is not just a business analyst as we knew them from the 90s. The people offering the services need a name to differentiate themselves when applying for jobs, and the customers need a way to indicate that they are not looking for "just a business analyst." We seem to have ended up with a rather foolish bit of terminology to serve this need, which is not surprising given that it emerged to facilitate an asymmetric market of the clueless looking for the knowledgeable.

_dps 15y ago

I second your snark, and also find "data scientist" to be analogous to "food cooker", but I've also come to realize that there is a need for branding whatever-it-is as separate from "business analyst" because otherwise expectations from non-technical customers get very confused. I interact with many buyers of buzzword-laden "big data" (snark on that one reserved for another day) and "data science" offerings. If you tell them you're a business analyst they will make many hidden assumptions based on the overall population of people using that title. Among those assumptions: 1. The business analyst will probably know SQL, but if we want to do something like analyze our query logs it will cost us developer time. 2. The business analyst will have domain knowledge about financials and maybe user engagement, but probably not anything else. The business analyst will not be able to tell us, for example, the characteristics of our 99-th percentile slowest page loads. 3. The business analyst is not hip and modern, because everything is about the "big data" these days. Realistically, the language of modern business data applications requires a name for someone who is not just a business analyst as we knew them from the 90s. The people offering the services need a name to differentiate themselves when applying for jobs, and the customers need a way to indicate that they are not looking for "just a business analyst." We seem to have ended up with a rather foolish bit of terminology to serve this need, which is not surprising given that it emerged to facilitate an asymmetric market of the clueless looking for the knowledgeable.