Saul Steinberg, Untitled (Question Marks) (1961)
Since writing my post on data science being different, a lot of people have emailed/DMed me asking for advice on Python, data science, machine learning, and careers.
And, as David wisely notes, if you find yourself saying the same thing a lot, you should put it in a blog post, so here it is.
Hopefully this advice is useful to people who write to others on the internet asking for help.
Often, I’ll get messages like these:
“I’m getting into machine learning. Do you have any book recommendations?”
“I’m looking to switch into data science as a career. [5-10 paragraphs of background, including every course the person has taken and a detailed description of each job] Can you tell me what to do?”
“I want to learn Python. Can you give some links to blog posts that you like?”
I won’t be sending responses to questions like these. Why?
I get a lot of questions. I get at least 1-2 detailed emails a week, and at least 3-4 Twitter DMs. Many people in tech (and, similarly, in journalism or other high-visibility online fields), particularly those who have popular blogs, high Twitter followings, have volunteered as open source maintainers, and have written books, get similar amounts of mail - usually much, much, more.
It’s very daunting to open your inbox or Twitter DMs and have 5-6 notes of people who need your advice or input on something, on a regular basis.
Don’t get me wrong, I’m happy to answer career and technology questions and enjoy doing it! It’s one of the reasons I have my email publicly available, and why I’m on Data Helpers. I am really flattered that people think I’m an expert in anything other than hosing Python environments!, and I love engaging in the online tech community.
However, at some point, my ability to keep up with my inbox lags behind my ability to provide focused, concrete, helpful advice.
As Trey writes in his fantastic post,
First, let me make absolutely clear that it is flattering to be seen as someone whose opinions are valuable and it is certainly a position of privilege to be seen as a source of career advice. I’d also say that I’m probably the last person you should be seeking advice from. I’m stumbling around in the dark without a flashlight just like everyone else!
Second, let me just tell you how much dread that simple email can bring about. If I ignore this email, I’ll feel like a jerk. If I respond to this email, I’m almost guaranteed 30–60 minutes of meandering, unplanned conversation. Don’t do this! The person you are emailing is a) busy, b) not invested in your project or job search, and c) going through their own struggles in life.
So, how do I deal with this? I triage based on a set of internal heuristics, namely:
- Questions are well-written
- The question is respectful of the reader’s time constraints and easy to act on
Asking good questions is a critical skill, both online and at work, because asking a good question gets you a good answer.
There are two great resources for figuring out how to ask good questions:
- Julia’s excellent post on how to ask questions
- Stack Overflow’s guidelines on asking questions on their platform
But these are specific to in-person and the Stack Overflow platform, so I wanted to write something specifically about email/Twitter. Here are some of my (fast and loose) guidelines for asking good questions that have a much higher probability of being answered:
Good questions make it obvious that the person has already done some basic research on their own. If me Googling for half an hour gives the same result as you doing it, it’s not a good question.
Question: “What’s the best resource for learning Python?”
Better question: “I’ve seen that some people use Learn Python the Hard Way, but that there’s also some controversy about it online. Do you have an opinion on this book and would you recommend it?”
Good questions are very specific, narrow in scope, and tailored to the person’s area of expertise. Usually you can ask these questions by reading a person’s past blog posts/watching talks/etc.
Question: “Can you tell me which data processing library I should use?”
Better question: “I see that you’ve given talks on using both Spark on EMR, Hadoop, and even doing data science on your laptop. Do you have one that you prefer if I have 10 TB of data?”
Good questions are brief and to the point (respectful of the reader’s time), leaving a call to action as the last part. Often, I’ll get pretty long emails including every single detail that’s not relevant to the question at hand.
We’re all super busy and get hundreds of emails. The ones we’re more likely to answer are short and easily-skimmable.
Here’s an email template I’ve developed for asking good questions that allow the recipient to skim and more easily address what you need:
Third, make this your email template:— Vicki Boykis (@vboykis) August 16, 2019
1. First paragraph: Ask your question
2. Second Paragraph; Give the relevant background to the question
3. Third paragraph: Finish by restating the q + with what you'd like the recipient to do (email you, call you, set up time to chat, etc.)
Another caveat I want to add about “life advice”-type questions is that most people who give life advice are only going off of what made them successful, which might not work for you at all.
Asking good questions is a time investment. But, you’re expecting the person answering your question to put in time for you, too. So, it’s an investment that will pay dividends.
I hope this is helpful in your quest for gaining more knowledge online. Good luck! :)