The spreadsheet we use to decide which features to cut
The NestBoard team · 2026-07-22
We keep a shared spreadsheet called "Feature requests that make us sweat." It's not the ideas that are obviously bad—those are easy. It's the ones that sound good in isolation but would slowly turn NestBoard into something we wouldn't want to use ourselves.
The spreadsheet has six columns. Every request that makes it past our first gut check gets scored on each one.
Does it add a new concept?
This is the first filter. If a feature requires us to introduce a new mental model—a new kind of object, a new layer in the navigation, a new distinction the user has to remember—the bar is very high.
When someone asked for separate grocery lists per store, it sounded reasonable. But it meant teaching everyone what a "list scope" was. We said no. The pantry already handles "what do we have," and the meal planner generates a single unified shopping list. Adding list management would be a third concept for the same problem space.
Does it make the happy path longer?
We measure this literally. If the most common use case now takes an extra tap, an extra decision, or an extra conditional ("only if you turned on X"), that's a point against it.
Custom color-coding for calendar events is a good example. Plenty of people asked for it. But it would mean every event creation flow needs a color picker, and everyone looking at the calendar needs to learn your personal color system. The cost is constant; the benefit is occasional.
Does it compete with calm?
Some features are good ideas that create anxiety by design. Read receipts for chores. Streaks that punish you for missing a day. Notifications that escalate if ignored.
We've built a few things that flirt with this line—the chore leaderboard, for instance—but they're opt-in and reversible. When a feature's core value depends on pressure or guilt, it doesn't make the cut.
Could Robin do it instead?
This column has saved us a dozen times. If the request is something you'd ask a human assistant to handle, we consider whether Robin could handle it with conversation instead of adding a UI for it.
Forwarded emails to Robin started this way. People wanted a "capture inbox" for confirmation emails and PDFs. We realized that wasn't a feature—it was a job for the assistant.
Does it assume everyone's household works the same way?
We're suspicious of features that encode a specific family structure or routine. "Assign a primary parent to each kid." "Mark weekday vs. weekend chores." They sound helpful until you meet a family they don't fit.
Why kid accounts work differently was the result of rejecting a dozen "standard" parental control patterns that assumed too much.
Does it survive the kiosk-mode test?
If the feature wouldn't make sense on a shared kitchen tablet where anyone in the house might glance at it, it's probably too personal or too complex. This test has killed more good ideas than any other.
The rule isn't that everything must work in kiosk mode—allowance tracking, for instance, lives in the parent's phone. But if the feature is meant to coordinate the household and it can't survive on the shared screen, that's a signal we're designing for individuals instead of the family.
The spreadsheet doesn't make decisions for us. But it makes our reasoning visible. When we say no to something, we try to point to the column where it failed. And when we say yes—like we did with shared medication bottles—it's because it scored well on all six.
Not every app should work this way. But for NestBoard, staying small and calm is a feature in itself. The spreadsheet helps us remember that.