Outside of formal presentations, data scientists spend plenty of time writing. From emails, slack pings, comments in documents, and more, your day will likely involve a fair amount of (non-code) typing. Here are tips to make that written communication effective.
The In-Person Test
In general, you should never write a message that you would not be comfortable saying aloud to the recipient in-person. Now, of course, such dictation might be embarrassing or uncomfortable - for example, if you are admitting to a mistake, or delivering constructive feedback. But remembering this rule of thumb will help to moderate what you write.
Consider also that in this imaginary monologue, anyone else included on the slack thread or email chain will also be present in the room. Because of this, when writing a message that could be difficult for your coworker to receive - perhaps you found a bug in their code, or a flaw in their methodology - consider sending it to them directly versus sharing in a broader project channel. The recipient will likely appreciate the chance to make things right without the rest of the team immediately knowing.
If you aren’t sure about a message, wait a while. Sleep on it, and see how it reads tomorrow. Even when you’re upset or frustrated, try to avoid escalation - you can’t easily take back what you write, and it always looks good to be the cooler head in the (virtual) room.
The All-World Test
A crucial generalization of the in-person test. Never write anything via work software that you would not be comfortable getting revealed to the entire world. Internal emails leak all the time, and even if you are not quite on the level of a Steve Jobs/Mark Zuckerberg, it’s worth considering that everything you write is recorded for posterity.
One practical piece of advice here - as your work friends become real-life friends, consider moving your conversations out of your company’s messaging software - to texts, WhatsApp, your own private Slack workspace. It’s nice to gossip without your employer being able to read it.
Choosing The Channel
Your author finds that Slack is now the default communication channel at most jobs. Email is still available, but generally saved for formal communication - for example, to message a client, or to talk to HR (basically, for items where you want a very obvious document trail).
Plenty of work communication also takes place in the comments of Google docs, or whichever software your company uses. This can be convenient, but keep in mind such threads are easy to miss or forget. If you need an answer ASAP, it may be best to message the person directly.
The Right Time
As any marketer knows, you must send the right message to the right person at the right time. For work communication, consider when your send is likely to yield the desired response from a colleague. You’ll almost certainly want to avoid weekend pings (unless it’s an absolute emergency). Consider a delayed send instead, whereby the message you wrote Sunday afternoon is not delivered until Monday morning. Slack and most messaging clients offer this functionality now.
Even during the work week, it’s often helpful to check coworker’s calendars and optimize the best time to send. Aim to pings folks when they have a break in meetings, but avoid sending messages during lunch. Regardless of the timing, avoid @here-ing or @channel-ing on Slack unless absolutely necessary - your colleagues will thank you.
The Right Tone
Without access to the non-verbal cues we use in everyday conversation, written communication can be tough. What you intended as a joke could easily be taken otherwise by your stressed out manager.
For this reason, assume negative intent when writing. What is the worst way a recipient could interpret your message? Avoid missteps by making your writing clear, concise, and unambiguous. Later on, as you build rapport, your writing can return to echoing your everyday speech, but initially it is best to focus on being understood.
The Right Audience
Especially early in your career, it can be intimidating to write to large groups of coworkers, particularly senior folks. You’ll proofread your messages again and again, and generally be tempted to ping only those who absolutely need to know.
Try to push back on this instinct. It’s best to be as open in communication as possible. It shows maturity to message a large project channel with the latest update, and colleagues will appreciate your candor. Ask a peer to review an important message before sending, but don’t be worried about writing to larger groups. Of course, be reasonable - the whole company likely doesn’t need to know about your daily schedule - but in general it’s best to include too many folks than too few.
What Do You Want?
Make it clear what you expect from message recipients. In practice, this often means articulating clear action items with deadlines. If you’d like someone to review a presentation, for instance, give them a specific date by which you’d like feedback.
Particularly when messaging a larger group, avoid uncertainty by explicitly stating who should be doing what. Avoid coming across as a drill sergeant by phrasing such requests politely and appropriately, but at the same time make sure folks know what you expect from them.
Hacking Data Science provides practical career advice for data scientists, aspiring and practicing. You’ve just read a free weekly article. Paid subscribers get access to two more paid-only posts each week, along with our full archive.