Extract. Transform. Read.
What Makes Data Engineers Say “No”
How to reject a data engineering request without ever saying the word “no.”
The following short read is last week’s edition of my weekly newsletter, Extract. Transform. Read. sent to 2,000+ aspiring data professionals each Thursday. If you enjoy it, you can sign up and receive your free project ideation guide.
When I worked at Disney, the absolute worst thing you could say to a guest was “No.” It’s not so much that Disney guests would hear “yes” throughout their vacation; they just wouldn’t hear no. And that’s why, as a developer and individual contributor, it’s important to master what The Art of Being Indispensable At Work author Bruce Tulgan calls the “good no.”
A bad no is a no said to something that could be feasibly accomplished. A good no is a no uttered in an effort to establish or reiterate priorities, like telling a stakeholder “No that dimension can’t be added to the pipeline during this sprint because it will require a new request to a separate API endpoint.”
Good nos can also put requests into the context of larger team and organizational goals; for instance, saying “no” to production-izing an ML model that only generates 15 rows of data that isn’t relevant to larger initiatives is, without a doubt, a good no.
It’s not just the sentiment of a denied request that can irritate a stakeholder or colleague. The word “no” just doesn’t sound good. It’s sharp. Definitive.
If I can’t say “yes” to someone, I focus on fulfilling two needs that are almost as good:
- Providing context
- Offering an alternative
Providing context: “I can’t put your query into production because I’m currently working on x initiative which impacts our team’s OKR for this quarter.”
Offering an alternative: “Unfortunately, I don’t have the bandwidth to take on a backfill of your requested scope. Instead, I can backfill the data to the prior quarter so you can at least see data within the last 90 days, which is what your dashboard seems to focus on.”
Even if you’re not working in a data role currently, this lesson can be applied to the bane of most students’ existence: Group projects. Putting your efforts into context can help avoid scope creep and make sure you don’t end up with a disproportionate amount of work.
Until next time — thanks for ingesting.
-Zach
