SQL for Product Managers
The Query Is the New Follow-Up Question
A dashboard answers the question someone already thought to ask. A query answers the question you just thought of.
Dashboards are built in advance, by someone guessing what questions will matter. That's genuinely useful, most of the time. But the interesting PM questions are almost never the ones anyone anticipated. "How many users hit the paywall last week" might be on a dashboard. "How many users hit the paywall last week, on mobile, who had already used the free tier for more than fourteen days, broken out by whether they came from the referral flow" is not on any dashboard, and it never will be, because dashboards can't anticipate every combination of filters a real product question needs.
That second question is a two-minute SQL query for someone who can write one, and a ticket that takes three days to come back for someone who can't. The gap between those two timelines is, functionally, the gap between a PM who can test an idea today and a PM who finds out next sprint whether their idea was worth having.
Dashboard | SQL query | |
|---|---|---|
Who decided what it shows | Someone else, in advance | You, right now |
Flexibility | Fixed filters, fixed breakdowns | Any filter, any breakdown, any combination |
Speed for a novel question | Requires a new dashboard or a ticket | Minutes |
Risk of a wrong number | Low, if built and tested carefully | Higher, because you're now responsible for the logic |
What it's good for | Recurring, known questions everyone needs | The specific, one-off question nobody anticipated |
Notice that last row on the risk side. This course is not an argument that PMs should replace analysts or stop trusting dashboards. It's an argument that a PM who can write and, just as important, sanity-check a query is a fundamentally faster and more independent thinker, and that independence is worth the modest effort of learning the syntax.
WORKED EXAMPLE
A PM at a subscription product wants to know whether a recent pricing page redesign changed the mix of plans people choose. No dashboard was built for this, because nobody knew to build it before the redesign shipped. Without SQL: file a ticket, wait, get a spreadsheet back three days later, by which point the launch decision window has closed. With SQL: write a query grouping signups by plan and by whether they signed up before or after the launch date, get an answer in ten minutes, and still have time to act on it this week.
YOUR MOVE THIS WEEK
Write down the last three data questions you asked someone else to answer for you. For each one, guess how long it took from asking to getting an answer. That total is the tax you're currently paying for not knowing SQL, and it's the number this course is trying to shrink.