The 90-day sprint planner
A date on the calendar changes everything about how you should study. Work backwards from yours: what to learn each week, how the review load actually behaves, and when to stop collecting words and start saying them.
"Learn Portuguese" is a wish. "Hold my own at dinner with Marina's parents in São Paulo on October 4th" is a project, and projects can be planned backwards.
Ninety days is enough to matter. With 2 lessons a week and 10 new words a day, you'll put roughly 900 words through spaced review by your date, on top of whatever you have now, and (this is the part most plans skip) you'll have rehearsed the actual conversations your date will throw at you.
Step 1 is on this page. Your email gets the full 12-week arc, the weekly template, the review-load math, and the miss-a-week contingency rules.
Step 1: define the finish line (5 minutes, tonight)
Write down three things:
- The date. A real one. If your goal has no date, give it one anyway; "by March 1st" beats "eventually" in every study you'll ever read about goal pursuit.
- Three scenes you must survive. Concrete ones: ordering for the table, the taxi negotiation, the toast at dinner, section 3 of the exam. These 3 scenes become your speaking curriculum in weeks 9 to 11.
- What "good enough" sounds like. One sentence. "I can keep up at dinner without switching to English for a full course." You'll test against it on the dry-run day (week 12, in the full plan).
Everything else in this planner hangs off these three lines. Vague finish lines make every study decision arbitrary; concrete ones make most decisions for you.
Get the full 12-week plan
The rest appears right here: the 4-phase arc, the weekly template with real numbers, how the review load builds (and when it peaks), the contingency rules for missed weeks, and what to do after the date so it doesn't all evaporate. Plus the 5-day email course on the science underneath.
No spam, unsubscribe anytime.
The 12-week arc
Weeks 1–2, foundation. Lock the routine before you chase volume: 2 tutor lessons a week booked as recurring slots, 10 new words a day from the highest-frequency material (the 650 essentials exist for exactly this), reviews every single day, even 5 minutes. Set up the post-lesson debrief and the lesson ground rules now, while enthusiasm is doing the heavy lifting.
Weeks 3–8, build. The engine phase. Lessons feed the debrief, the debrief feeds new cards, reviews compound in the background. Keep 10 new words a day, and start steering lesson topics toward your three scenes around week 6 (vocabulary for them, no rehearsal yet). Week 3 deserves a warning circled in red: it's where the first review wave matures and every self-built system wants to die. Survive week 3 and the habit usually holds.
Weeks 9–11, shift to speaking. Cut new words to 5 a day (the sprint's counterintuitive move, and its most important). Words you learn in week 10 barely get enough review cycles to stick by the date, so late-sprint effort pays better in production practice: rehearse one of your 3 scenes per week in lessons, full roleplay, tutor in character. Record them. The reached-for words from these rehearsals are the last vocabulary worth adding.
Week 12, taper and dry run. No new words at all. Reviews keep going (they're cheap now). One full dry run of all three scenes in a single lesson, then one lesson on whatever the dry run exposed. Test yourself against your "good enough" sentence from step 1. Then go do the thing.
The weekly template (build phase)
- Daily: reviews first, before new material. 10 new words. Total 20 to 30 minutes.
- Lesson days (×2): the lesson, plus the 10-minute debrief immediately after.
- One admin block (30 to 60 min): build the week's cards from the debrief harvest, prune what's not working. (Verbamor users: this block is what the app deletes.)
- One self-recording (10 min, weekly): talk to your phone about your week in the target language. By week 8 you'll hear the difference, which matters on the days motivation dips.
The review-load math, honestly
Nobody tells you this and everybody hits it: review load doesn't grow linearly. At 10 new words a day, your daily reviews build from ~20 a day in week 1 toward 80 to 120 a day by weeks 5 or 6, then plateau as old cards mature and space out. That's 15 to 25 minutes daily at the peak. It's fine. It's supposed to feel like a lot right there. Plan for the peak (don't schedule your sprint's hardest work weeks against it if you can help it), and don't respond by capping reviews, which just moves the debt.
Contingency rules
- Miss a day of reviews: nothing special, do today's pile today. The scheduler absorbs it.
- Miss a week (travel, flu, life): halve new words until the due pile is back under control, never skip the backlog. The backlog IS the curriculum now.
- Falling behind the plan at week 6: cut scope, not practice. Two scenes instead of three beats three scenes at half-strength.
- Ahead of plan: add listening hours, don't add words. Input is the cheapest upgrade at every stage.
After the date
You'll come home from the trip (or walk out of the exam) with the language at its lifetime peak. The forgetting curve starts that afternoon. The maintenance dose is small: reviews a few minutes a day, one lesson a week if you can. Decide on maintenance before the date, because afterwards the deadline that powered everything is gone, and unplanned maintenance mostly doesn't happen.
If you run the sprint in Verbamor: set your goal and target date in onboarding and the app runs a per-language countdown, seeds the 650, and its forgetting forecast shows you exactly what stopping after the trip would cost you. The plan above works with any tools. It just goes faster when the admin is automated.
Sprinting toward a date?
Tell Verbamor your goal and target date, and it runs the countdown, seeds the 650, and turns every lesson into scheduled practice for the day that matters.
Download the app