ケース面接bot-case interview strategy engine

Example

You input “Craft beer gift set.” The bot outputs ≥6 high-probability conditions in order: ④ (gender skew), ⑤ (age skew), ㉓ (non-routine—gifts), ⑧ (long-history leisure good if the brand/category qualifies), ⑱ (hobby/habit), ㉖ (experience good). It purposely avoids store/footprint flags unless the input is actually a retail operation, and it doesn’t apply EC-only logic unless e-commerce is explicitly the business model.

Scenario

In a case interview where time is tight, the bot’s first response acts like a ‘hypothesis grid’—it tells you where the upside probably is (e.g., acquiring new segments, life-stage timing, gifting occasions) and where not to waste time (e.g., brick-and-mortar levers when you’re just selling a product, not running a store).

Example

You say “はい” to proceed on condition ③ (Digital service). The bot (1) states why ③ fits (e.g., app-based product, older users struggle), (2) surfaces the exact Direction/Target phrasing tied to the condition, and (3) enumerates every prescribed measure verbatim (e.g., simplify UI, literacy training with retail partners at first-use timing, parent–child invite, senior-specific needs such as health or nostalgia) and adds situation-specific ideas right beside each measure.

Scenario

In a PMM/GTM case about a language-learning app, you immediately get a full, interview-ready answer: simplified onboarding flows, retail-assisted setup at the moment of phone purchase, a family-invite discount to get adult children to tutor parents, and a nostalgia content lane for seniors—each item grounded in a named condition rather than hand-wavy brainstorming.

Example

For a tangible good (e.g., headphones), the bot will not list ⑭ (physical stores), ⑮ (face-to-face retail), or ㉑ (location/footprint-driven businesses). For entertainment goods, it always places ④ and ⑤ first. For health supplements/cosmetics/beauty, it places ㉕ at the very top. It only applies ㉒ (EC) when the business is actually e-commerce, and treats ⑲ (durable) and ⑳ (health food) narrowly.

Scenario

This policing prevents common case pitfalls (like recommending store-layout tactics when you’re not operating stores). The workflow also mirrors real interviews: the bot lists conditions → stops → asks if you want to continue → then deep-dives one condition at a time, top-down, until you say stop.