Jessica Nguyen

30-day plan to learn a skill, with weekly pass/fail checkpoints

Builds a 30-day learning plan for any skill from your starting point and daily time, weighted toward practice, with a weekly test and a final challenge.

At a glance

Best for: Busy adults learning a practical skill (SQL, a language, an instrument, a design tool) who want a plan that fits a daily slot and shows real progress each week.
Tested on: Claude · Opus 5.5
You fill in: skillstarting_pointtime_per_daygoal
You get: Time: 45 minutes a day, about 5 hours a week. Week Focus and sub-skills Learn / practice Mini-project Checkpoint (pass mark) 1 (days 1-7) One table:… (full result below)

Prompt

Build me a 30-day plan to learn {{skill}}, starting from {{starting_point}}. I have {{time_per_day}} a day, and by day 30 I want to be able to {{goal}}.

Requirements:

• Four weekly blocks, plus days 29-30 for a final test. Each week gets a focus, the 2-3 sub-skills it builds, and a mini-project I finish by the weekend.

• Most of the time goes to doing, not watching: at least 60% of each week is practice. Show the split.

• A checkpoint at the end of each week: a concrete test with a pass mark (e.g. "do X in under Y minutes without looking anything up"). If I fail, tell me what to repeat, not to push on.

• Name the kind of resource for each week (official docs, a type of exercise, a community) rather than specific courses you can't vouch for.

• Flag the week where most people quit this skill, why, and one tactic for getting through it.

• End with the day-30 test that proves I've reached my goal.

Keep it scannable: a table for the weeks, bullets for the rest.

Fill in

{{skill}}: The skill to learn, as specifically as possible.e.g. SQL for data analysis
{{starting_point}}: What you already know that's related.e.g. comfortable in Excel, never written code
{{time_per_day}}: Realistic daily time.e.g. 45 minutes
{{goal}}: What you want to be able to do at the end, stated as a real task.e.g. answer my team's weekly sales questions from our Postgres database

Example input

skill: SQL for data analysis
starting_point: comfortable in Excel (VLOOKUP, pivot tables), never written any code
time_per_day: 45 minutes
goal: answer my team's weekly sales questions from our Postgres database myself, including joins and grouping by month

1.7k0

Generated Outputs (1)

2 hours ago

Claude

Opus 5.5

2 hours ago

Generated Output

Time: 45 minutes a day, about 5 hours a week.

| Week | Focus and sub-skills | Learn / practice | Mini-project | Checkpoint (pass mark) |
|---|---|---|---|---|
| 1 (days 1-7) | One table: SELECT, WHERE, ORDER BY, LIMIT | 30 / 70 | List all orders over $500 last month from a sample sales table | 5 single-table questions in 20 minutes, no lookups: 4 of 5 correct |
| 2 (days 8-14) | Summarizing: COUNT, SUM, AVG, GROUP BY, HAVING, grouping dates by month | 35 / 65 | Monthly revenue by region | Rebuild one of your own Excel pivot tables in SQL; totals match exactly |
| 3 (days 15-21) | Joins: INNER vs. LEFT JOIN, keys, duplicate rows | 30 / 70 | Revenue by product category across 3 tables | 3 join questions correct, and explain one row-count mismatch |
| 4 (days 22-28) | Real questions: WITH (CTEs), CASE WHEN, NULLs | 25 / 75 | Answer one of last week's actual team questions | Your number matches the team's report, or you can explain the gap |
| Days 29-30 | Final test | | | See below |

Resources by week

• Weeks 1-2: the tutorial section of the official PostgreSQL documentation, plus an interactive in-browser SQL exercise site. Use a local install or a free online sandbox with a sample sales dataset.

• Week 3: join-focused exercise sets. Sketch the tables on paper before each query.

• Week 4: your company's database. Ask for read-only access in week 1, because approvals take time. Book one 15-minute review with someone from the data team.

If you fail a checkpoint
Redo that week's exercises on a different dataset for 3 days, then retake it. Don't start joins until GROUP BY feels routine.

The week people quit: week 3.
Joins are the first thing that feels unlike Excel, and a wrong join doesn't throw an error. It silently duplicates rows, so totals look plausible but are wrong. Tactic: write each join as the VLOOKUP you would have used, and run COUNT(*) before and after every join.

Day-30 test
Pick 3 real questions your team asked in the past month: at least one needing a join and one grouped by month. In 90 minutes, with no help, write queries that answer them. Pass: all 3 match the numbers your team reported, or you can explain every difference.

Why it works

Checkpoints with pass marks turn "I think I'm learning" into a yes or no each weekend, and "tell me what to repeat, not to push on" stops the common mistake of stacking new material on a shaky base.

The 60% practice rule counters the tendency of learning plans to fill up with videos. Flagging the week people quit means the plan prepares you for the dip instead of pretending it won't happen. A final test written as a real task ties the whole plan to what you actually wanted, not to finishing a course.

When not to use it

Thirty days builds working basics, not mastery; don't expect interview-level depth. The plan assumes you can get practice material, so for skills that need equipment or a partner (an instrument, a sport, a spoken language), you'll need to add that yourself. If you miss more than a few days, restart the current week rather than skipping ahead.

Comments (0)

No comments yet. Be the first to share your thoughts!