Most case prep is a casebank and a lot of hope. lowercase is a training programme built on what learning science says about how skills stick, applied to the one skill a case interview tests: breaking a messy problem into clear pieces.
Every week on lowercase follows the same four steps. Each one exists because of a specific finding about how people learn.
A short explainer with a worked example. One concept, small steps, no wall of theory.
A quick check. You pass to unlock what depends on it, or get targeted practice where you slipped.
A live case with a friend or the AI interviewer. Facts arrive only when you ask for them.
Spaced, mixed review pulls older ideas into newer cases so they stay usable, not just familiar.
Our approach follows the research summarised in Math Academy's pedagogy notes, one of the clearest write-ups of how to teach a hard subject well. Here is how each principle shows up in case prep.
Students learn far more when they must show they've mastered a topic before building on it. Gaps left early compound into confusion later.
Further reading: Bloom, Learning for Mastery (1968). Kulik, Kulik & Bangert-Drowns, meta-analysis of mastery learning (1990).
Concepts sit in a map of prerequisites. You can't start market-entry cases until you can size a market, and Level 2 stays locked until your Level 1 results say you're ready.
One-to-one tutoring combined with mastery learning produces dramatically better results than classroom teaching. The difficulty has always been doing it at scale.
Further reading: Bloom, "The 2 Sigma Problem" (1984).
Peer practice gives everyone a live interviewer. The AI interviewer adds a patient, always-available tutor who asks rather than tells, so feedback comes during the case and not weeks later.
People learn more by doing problems with quick feedback than by watching or reading, even when passive learning feels more productive. Expertise comes from focused practice on specific weaknesses.
Further reading: Freeman et al., PNAS (2014). Deslauriers et al., PNAS (2019). Ericsson, Krampe & Tesch-Römer (1993).
Explanations are short and followed straight away by questions. In cases, you lead. The rubric names the weakest skill, so your next session works on that skill, not just on another case.
Working memory is small. When the basics run on autopilot, there's room left for the hard part: reasoning about a problem you've never seen.
Further reading: Miller (1956). Cowan (2001). Sweller, cognitive load theory.
The profit equation, case-type recognition and quick maths are drilled until they're automatic. In the interview, your attention goes to the client's problem rather than to remembering the framework.
Memory fades on a predictable curve. Reviewing just as you're about to forget, at growing intervals, makes knowledge last far longer than cramming.
Further reading: Ebbinghaus (1885). Bahrick et al. (1993). Cepeda et al. (2006).
Concepts you've mastered come back in review at widening intervals, and your daily streak makes the short review sessions a habit rather than a chore.
Practising one problem type in a block feels efficient but teaches little about choosing the right method. Mixing types forces you to work out what kind of problem you're facing.
Further reading: Rohrer & Taylor (2007). Kornell & Bjork (2008).
Real interviews don't announce the case type. Review sets mix profitability, growth, entry, pricing and M&A, so you learn to recognise the case before you choose a structure.
Knowledge sticks when new ideas build on old ones and add connections. Very similar ideas taught back to back blur together, so they're best spaced apart.
Further reading: research on retroactive facilitation and on proactive and retroactive interference.
Each level reuses the last: the profit tree reappears inside M&A synergies, and market sizing inside entry cases. Look-alike frameworks, such as 3Cs and five forces, are deliberately kept apart in the sequence.
Learners handle new material best in small steps, starting from a worked example, with steps grouped into meaningful chunks and ideas shown visually as well as in words.
Further reading: Sweller & Cooper on the worked-example effect (1985). Catrambone on subgoal labels (1998). Paivio, dual coding theory.
Every concept opens with a worked case snippet. The issue tree is the picture: blue for the branches you've opened, grey for the ones still to explore. Hints break a stuck moment into the next small step.
| Typical case prep | lowercase | |
|---|---|---|
| What you start with | Hundreds of cases | The concepts those cases depend on |
| How you move up | Whenever you feel ready | When your results show mastery |
| How facts reach you | All at once, in the case text | Only when you ask the right question |
| Review | Rarely, by rereading | Spaced and mixed, built into the week |
| Feedback | "That was good" | A six-skill rubric and a transcript |
| The goal | Solve more cases | Think like a consultant, then leave |
Spaced repetition only works if you show up. XP, streaks and a weekly league exist for one reason: to make the short daily session easy to keep. They reward finishing reviews and cases, never mindless clicking, so the game and the learning point the same way.
Your first lessons take ten minutes. Your first case with a friend takes thirty.
Start practising