BOOK 5 OF 12 · BrainStation Miami & Daily Squad Discovery

Problems first: Picking a hill worth climbing

The two-sided whiteboard in the loft: peeling back the 5 Whys below the waterline on Side A, and filtering scope on the Kano curve on Side B.

Someone walks past your desk, or drops a polite message in Slack, and asks: “Can you send me a daily email?”

In 2026, you can point a coding agent at a cron job and ship that daily email before lunch. Everybody checks a sprint box, velocity looks great on Friday, and nobody notices you just automated a bad guess instead of solving the problem.

Every ten weeks teaching product managers at BrainStation Miami, I draw that exact conversation on the board. It’s the same diagram sitting on the whiteboard inside Home Port:

  1. “Can you send me a daily email?” Why?
  2. “So I don’t have to check the status page ten times a day.” Why are you checking it ten times a day?
  3. “Because I need to know the moment availability changes.” Why?
  4. “So I don’t miss my window.”

Nobody ever wanted another 8:00 AM digest in their inbox. The daily email was just a tired person’s best guess at a solution, and one that shows up twenty-three hours too late when an inventory window opens at 9:05 AM.

Cheap code makes climbing the wrong hill dangerously fast

When building a first draft used to cost three engineering sprints, scarcity forced teams to argue about whether an idea deserved to exist. Now that coding agents have pushed first-draft code close to free, output volume stops being a flex and turns into a trap. Shipping ten unexamined requests in a week just gets you to the top of a tiny local hill faster, leaving users squinting at cluttered screens while your team inherits permanent maintenance debt.

I come back constantly to Chris Dixon’s essay on climbing the wrong hill: incremental polish is great once you’re on the right mountain, and a dead end if you started on a small bump across the valley.

Years ago on cruise search, we saw that play out in real dollars. Brand-new visitors landed on sailings sorted by Price: Low to High. Judged by clicks on the top row, the page looked healthy because everyone clicks a $199 weekend sailing out of curiosity. Families flying into South Florida for a seven-night spring break trip bounced, though, because the sailing that actually fit their week sat buried three scrolls down. Stepping back to ask what job a first-time visitor was actually trying to finish led us to cold-start geographic and session-intent ranking in Cruise AI, lifting booking conversion by 13% in controlled holdout A/B tests ($300M/yr in incremental bookings) without adding a single new filter checkbox to the page.

Look for the ugly workaround, then watch hands silently

Before anyone opens an editor, two quick checks tell you almost everything about whether a problem is real:

  1. Audit the ugly workaround: Show me the spreadsheet, camera-roll screenshots, sticky notes, or group-chat ritual people are enduring today to get the job done. No ugly workaround usually means low real-world urgency (with two deliberate exceptions I carve out in class: pure hospitality delighters people wouldn’t think to ask for, and accessibility fixes where people had quietly given up).
  2. Separate building to learn from building to earn: Instead of debating flat Figma slides where testers have to pretend, spin up a working prototype in a couple of days, skip the login walls, and sit on your hands while someone uses it. Watching silently where a real person pauses, squints, or opens a second tab beats ten stakeholder meetings.

Every discovery loop ends in one of three honest calls: Stop or simplify (fix the copy, change the default, or drop the feature), Prototype to learn (spike the riskiest assumption in days), or Build to earn (commit real production engineering once human value is proven).

Restraint on the Kano curve (and knowing when to walk away)

On Side B of the whiteboard sits the three-curve Kano model I pair with the 5 Whys. Customers punish you immediately when basic Must-Haves break, compare you consciously on one or two Performance bars like speed or price, and tell their friends about a single genuine Delighter.

Starting with the human situation instead of a favorite technology is what keeps each build honest:

Take the playbooks off the whiteboard tray

If you want to run this exact discovery filter with your own squad or drop it into Claude Code, Cursor, or Codex before scoping a feature, grab the two Problem-Space playbooks straight off the tray:

Next: Wedding Cam