🎯 Skill Acquisition Sprint — The 30-Day Ramp

Skill Sprint: SQL for PM self-service analytics, 6 weeks, shippable = run my own product metrics queries

A skill acquisition sprint is a 20-focused-hour learning plan that reaches 'competent enough to use' in a narrowly-scoped skill, based on research from Josh Kaufman's 'First 20 Hours' and Scott Young's 'Ultralearning' methodology. Research shows that 80% of usable skill in most domains comes from the first 20% of learning effort (Pareto applied to skill acquisition). Your scope — SQL for PM product analysis, NOT SQL for database administration — is perfectly bounded. 6 weeks × 5-7 hrs = 30-42 hours available; more than enough. You're not over-scoped.

Scope Check

Well-scoped. Your goal translates to ~15-20 hours of focused learning:

• SELECT, WHERE, ORDER BY, LIMIT (basic querying)
• JOINs (INNER, LEFT — you won't need OUTER/CROSS)
• GROUP BY + aggregations (COUNT, SUM, AVG)
• Date functions + windowing (for time series)

You are NOT trying to learn: stored procedures, indexing optimization, database design, triggers, or any DBA-level topics. Skip them all.

The Minimum Effective Set

3 sub-skills for product analytics SQL:

1. Basic SELECT with filtering — pulling specific rows with WHERE clauses. (Week 1)

2. JOINs across 2-3 tables — 90% of product analytics is joining user table + event table + plan table. (Week 2-3)

3. Aggregations + time series — COUNT(users), SUM(revenue), GROUP BY week/month. (Week 3-4)

Everything else is bonus.

What to IGNORE

• Subqueries and CTEs (initially). You CAN learn these — but 80% of your queries won't need them. Add week 5-6 if time.
• Window functions. Advanced. Unless you specifically need RANK() or ROW_NUMBER(), skip.
• Database design / normalization. Not your job. You're a consumer of the DB, not architect.
• Performance optimization. Your queries will be slow. The data team will tell you if it's a problem.

The 30-Day Sprint Plan (adjusted to your 6-week window)

Week 1 (5 hrs): Fundamentals

• Day 1-2: Interactive course — SQLBolt.com (free, ~2 hrs, covers SELECT + WHERE + ORDER BY)
• Day 3-4: Write 20 practice queries against a public dataset (SQLBolt has these built in)
• Day 5: First real query against your company DB — just pull 100 rows from users table. Success = it ran.

Week 2 (6 hrs): JOINs

• Day 1-2: SQLBolt's JOIN lessons + Mode's SQL tutorial (free, excellent for PMs)
• Day 3-5: Write 15 queries that join 2 tables from your company DB. Target: user_events table + users table.
• Exit: you can answer 'how many users from signup cohort X did action Y?'

Week 3 (6 hrs): Aggregations + Time Series

• Day 1-2: GROUP BY, COUNT, SUM, AVG drills
• Day 3-4: DATE_TRUNC, EXTRACT — weekly/monthly rollups
• Day 5-6: Build your FIRST PRODUCT METRIC QUERY. E.g., 'weekly active users by cohort.'

Week 4 (5 hrs): Real Product Questions

• Pick 3 product questions your team actually has. Write queries for each.
• Debug the ones that break. This is the hardest week — most learning happens here.
• Shippable: one dashboard-ready query your team can consume.

Week 5 (5 hrs): The Feature Launch Metrics

• Write the 5-7 queries you'll need for your new feature launch.
• Prepare a pre-launch baseline query set.

Week 6 (5 hrs): Deploy + Iterate

• Launch day: run your queries. Expect 2-3 to break. Debug.
• End of week: you've self-served on your own feature metrics without asking data team.

Weekly Practice Drills

• Week 1: 20 queries (SELECT + WHERE practice)
• Week 2: 15 JOIN queries, increasing complexity
• Week 3: 10 aggregation queries, mix of GROUP BY / date functions
• Week 4: 5 'real product question' queries (must produce usable answer)
• Week 5: The 5-7 launch-specific queries
• Week 6: Debugging-real-data drill — take broken queries, fix them

Resources (Curated, Not Exhaustive)

1. SQLBolt.com — free, interactive, covers 80% of what you need in Week 1-3. Start here.

2. Mode's SQL Tutorial (mode.com/sql-tutorial) — excellent for analytics-specific SQL. Free.

3. Your company's actual Postgres — YOUR real learning environment. Use as much as possible.

4. 'Practical SQL' by Anthony DeBarros (book) — only if you prefer books. Skippable if you do 1+2+3.

5. Slack your data team: ask for 3 examples of queries they run. Reverse-engineer them.

Skip: DataCamp SQL tracks (too slow for your goal), Stratascratch (too interview-prep focused), most YouTube 'learn SQL in 4 hours' videos.

The First Shippable Output

By end of Week 4: a dashboard-ready query that measures one product metric for your team.

E.g., 'Weekly active users by signup cohort, broken down by plan tier.'

Share it with your data team: 'I wrote this — can you sanity check the logic?' This gets you feedback + builds credibility.

Exit Criteria

You've hit competence when:

• You can write a correct JOIN + GROUP BY query without looking up syntax
• You can debug a broken query by reading the error message
• You ship product metrics without asking the data team for basic pulls
• Your queries answer 'good enough' for product decisions (not perfect — good enough)

At that point: STOP learning more SQL. You can always level up later if needed. Right now you're done.

Common Plateaus + Fixes

• Week 2 plateau (JOINs): JOINs are the conceptual hump. If stuck: draw tables on paper, trace the JOIN manually. Mental model beats more tutorials.
• Week 3 plateau (aggregations + GROUP BY): GROUP BY is weird. Pattern: every column in SELECT must be either IN the GROUP BY or an aggregate function. If you violate this, your brain is still thinking imperatively. Drill 20 queries, it clicks.
• Week 4 plateau (real data): Your company's data is messier than tutorial data. Expect queries to break in weird ways. This is the real learning.
• Motivation dip at week 3-4: mid-sprint depression is normal. Push through. It gets better.

Beyond 30 Days

Keep going if:

• You find yourself writing SQL every week for real questions
• You start to want advanced features (CTEs, window functions)
• You're enjoying it

Stop / pause if:

• You hit your competence bar and don't need more
• You realize your actual need was Excel pivot tables, not SQL
• You're burning out

Key Takeaways

• Your goal is well-scoped: 15-20 hrs to 'PM self-service analytics SQL.' You have 30+ hrs available. Surplus.
• 3 sub-skills only: SELECT+WHERE, JOINs, aggregations+dates. Ignore everything else tutorials teach.
• Week 4 real-data drill is the hardest part. Messy company data breaks queries — that's where real learning happens.