Closer/seed/questions/QUESTION_CONTENT_GUIDE.md

# Closer Question Writing Guide v8

## Mission

Closer is not a questionnaire.

Closer is not therapy homework.

Closer is not a personality quiz.

Closer should feel like a conversation game couples voluntarily keep playing because they are having fun.

Every question should move the couple toward at least one of these:

* laughing
* flirting
* learning something new
* remembering something
* planning something together
* feeling understood
* feeling appreciated
* creating a future memory

If a question does none of those things, delete it.

## Consumer First

Every question must pass this test:

```text
Would a real couple willingly answer this on a Friday night?
```

If not, rewrite it.

Never write for psychologists.

Write for normal couples.

## Product Standard

Closer should feel premium.

Every category should feel handwritten by an experienced relationship coach and game designer, not generated by AI.

If a user can predict the next question, the category fails.

## Standard Question Mix

Standard 250 question packs should use this mix:

| Type | Count |
|---|---:|
| multi_choice | 140 |
| single_choice | 50 |
| scale | 35 |
| this_or_that | 15 |
| written | 10 |

At least 76 percent must be choice based.

Typing should feel rare and meaningful.

Special packs may override this mix only when the override is documented in the pack metadata and in `QUESTION_REWRITE_PLAN.md`.

## Emotional Mix

Each standard category should roughly contain:

* 35 percent playful
* 25 percent everyday relationship
* 20 percent meaningful
* 10 percent future focused
* 10 percent deeper vulnerability

Never stack several heavy questions together.

Daily packs should lean more playful, fun, sweet, and low pressure than standard category packs.

## Conversation Goals

Every question should create at least one of these reactions:

* "I did not know that."
* "That is adorable."
* "We should actually do that."
* "I cannot believe you picked that."
* "I never thought about that."
* "That is a really good point."

If none apply, rewrite it.

## Daily Single Choice Voice Standard

Daily questions are different from normal category packs.

They should feel like small couples game moments.

They should sound like two people deciding what to do, eat, watch, laugh about, try, flirt with, or enjoy together today.

Daily questions should be:

* quick
* warm
* concrete
* playful
* fun before merely useful
* game-like
* easy to answer in under 10 seconds
* relationship focused
* doable in normal life

Daily questions should not sound like:

* therapy homework
* self-help worksheets
* HR wellness surveys
* abstract emotional processing
* generic AI relationship advice
* household admin
* bedtime logistics
* chore planning

If the question sounds like a couples counselor wrote it, rewrite it.

If the question sounds like roommates managing a house, rewrite it.

## Daily Fun Gate

A daily question passes only if it creates at least one of these:

* a smile
* a laugh
* a flirt
* a cute choice
* a tiny date idea
* a playful debate
* a small memory
* a small thing the couple might actually try tonight

Warm is not enough. Concrete is not enough. Useful is not enough.

Ask this before approving any daily question:

```text
Would this feel fun, sweet, or playful inside the app tonight?
```

If the honest answer is "useful, but not fun", rewrite it.


## Research-Informed Fun Rules

Recent relationship-question products and articles point to the same pattern: the strongest prompts feel like a game first, then create connection as a side effect. Paired markets quick, fun check-ins and says the relationship should not feel like work. Conversation-card coverage emphasizes that cards give people permission to skip small talk and reveal something interesting. Date-night and couples-question lists work best when they use preferences, memories, flirty choices, funny hypotheticals, tiny adventures, and low-pressure honesty.

Use this as the daily pack standard:

* Game first, insight second.
* A question should create a choice, mini-mission, playful debate, flirt, memory, or laugh.
* Avoid questions that only ask users to manage the relationship better.
* Avoid questions that sound like advice, coaching, therapy, or household planning.
* Ask about wants, favorites, memories, tiny plans, silly preferences, and sweet attention.
* For flirty or spicy prompts, start light, keep it consent-based, and never make the user feel cornered.
* For deeper prompts, make them feel like a game card, not a therapy intake form.

A daily question should answer at least one of these:

* What would be fun to choose right now?
* What would be cute to try tonight?
* What would make us laugh?
* What would make us flirt a little?
* What would help us learn a tiny new thing about each other?
* What would create a small shared memory?

If the answer is only "this would be responsible", reject it. Adults already have bills. The app does not need to become another one.


## Daily Game Mechanics

Every daily question should use at least one clear game mechanic.

Preferred mechanics:

* Would-you-rather style choice
* Fake award
* Tiny mission
* Mini date pick
* Snack draft
* Silly bet
* Guilty-pleasure pick
* Memory pick
* Flirty pick
* Compliment choice
* Photo challenge
* One-song challenge
* Mystery treat
* Cute dare, kept low pressure
* "Choose our vibe" for the next hour

Do not overuse one mechanic. A full weekday should not feel like 70 versions of the same snack question wearing different socks.

### Option Quality Standard

Daily options should be:

* vivid enough to picture
* short enough to tap quickly
* similar in effort
* similar in emotional weight
* tied to the prompt
* fun, sweet, flirty, silly, or date-like

Avoid vague options like:

* Something sweet
* A cute moment
* More connection
* Better communication
* Quality time

Avoid weird domestic options like:

* The good blanket saved
* A clean counter
* A bedtime plan
* Dishes handled
* The laundry moved

Better option style:

* A two-song kitchen dance
* A dessert walk
* A ridiculous fake award
* A flirty text from across the room
* A corner-store snack hunt
* A couch movie trailer voice

## Daily Option Direction

Prefer daily options like:

* snacks
* tiny dates
* silly bets
* inside jokes
* couch games
* music picks
* mini adventures
* playful compliments
* low-pressure flirting
* cozy but fun moments
* small surprises
* dramatic fake awards
* guilty-pleasure shows
* cute photos
* dessert runs
* short walks with a purpose

Avoid daily options centered on:

* bills
* laundry
* dishes
* clean counters
* errands
* appointments
* calendars
* bedtime planning
* household maintenance
* saved blankets
* clinical reassurance phrasing

These can exist in real relationships. They should not dominate the daily fun pack.

## Banned Daily Tone Words

Avoid these in daily questions:

* reset
* process
* mental load
* emotional load
* autopilot
* pressure
* soft landing
* relationship dynamic
* name the mood
* emotional processing
* communication style
* conflict framework

These are not banned from every serious category forever, but they should almost never appear in daily questions.

## Daily Good vs Bad

Good daily question:

```text
What tiny date move sounds best tonight?
```

Good options:

* Dessert on the couch
* A two-song kitchen dance
* A short walk with snacks
* A ridiculous movie pick

Bad daily question:

```text
Before phones win, what would make the night nicer?
```

Bad options:

* A clean counter
* A quick shoulder rub
* A simple bedtime plan
* The good blanket saved

Why it fails:

* too much household admin
* not playful enough
* one option is weirdly phrased
* it does not feel like a couples game

## Daily Single Choice Weekday Lineup

This is a special daily pack, not a standard 250 question category pack.

Target counts:

* 500 total
* 75 free
* 425 premium
* 500 single_choice

Weekday themes:

* Monday: Mood Check Monday
* Tuesday: Tiny Win Tuesday
* Wednesday: Real One Wednesday
* Thursday: Laugh It Off Thursday
* Friday: Flirty Friday
* Saturday: Side Quest Saturday
* Sunday: Slow Burn Sunday

See `DAILY_SINGLE_CHOICE_WEEKDAY_SYSTEM.md` for full theme rules.

## Daily Voice Enforcement

The daily weekday pack must use a review loop, not a one-pass bulk rewrite.

Daily questions must be written in small batches, reviewed, marked, fixed, and reviewed again before scaling.

A daily question is not approved until:

* the prompt sounds natural out loud
* the options cleanly answer the prompt
* the wording is concrete, not abstract
* the question feels like a game moment
* the question feels fun, sweet, playful, flirty, or date-like
* the theme is clear without sounding forced

If the question passes JSON validation but sounds weird when read out loud, it fails.

If the question is useful but not fun, it fails.

## Daily Patch Discipline

Daily pack updates must not become blind rewrites.

Review the full pack first. Mark only the questions that fail. Fix only those marked IDs.

Do not rewrite passing questions because the pack is being updated.

Passing questions are frozen unless they later fail one of the daily rules.

Normal content fixes should preserve metadata and change only the prompt and options.

A mass rewrite is allowed only if more than 60 percent of a weekday or pack fails for the same root cause, and the review report explains why patching is worse.

## Fun But Grounded

Daily questions should be fun, but not random.

Reject questions that become silly noise instead of a usable couple moment.

Watch for overuse of:

* snack drafts
* fake awards
* couch games
* mascot jokes
* random object choices
* dramatic bits
* cute phrases that do not match the prompt

A daily pack should feel like a relationship game for adults.

It should not feel like chore bingo, therapy homework, or a children's party game wearing a couples-app badge.

## Option Answer Test

Every option must cleanly answer the prompt.

Read the prompt and then each option as a sentence fragment. If the option sounds grammatically or logically wrong, fix it.

If one option fails, fix the option. If two or more fail, rewrite the whole answer set.

Examples:

Bad:

```text
Which tiny date move fits after dinner?
Choosing the fun mug
```

Better:

```text
Which tiny date move fits after dinner?
A two-song kitchen dance
```

Bad:

```text
What should we do for fun before phones win?
A bedtime plan
```

Better:

```text
What should we do for fun before phones win?
Trade dramatic compliments
```

## General Tone Rules

Avoid robotic openings:

* Describe
* Reflect on
* Discuss
* Evaluate
* In what ways
* How satisfied are you
* Explore your feelings
* Identify the ways

Use normal human wording.

Better:

```text
What tiny date move sounds best tonight?
```

Worse:

```text
In what ways could we improve our relational connection this evening?
```

## Option Rules

For choice questions:

* options must answer the exact prompt
* options must be similar in weight
* options must not overlap too much
* options must not shame either partner
* options must sound like real choices
* options should be short

Reject answer sets where one option is obviously the "correct" healthy answer and the rest are fake.

## Written Question Rules

Written questions must earn the keyboard.

Use written questions only for:

* memories
* appreciation
* future plans
* short stories
* meaningful personal answers

Do not use written questions for basic preferences.

## Final Content Test

Before shipping a pack, read random questions out loud.

Reject anything that sounds like:

* AI filler
* therapy homework
* a survey
* a worksheet
* household admin
* a chore list
* an app trying too hard to be deep

Closer questions should make couples want to keep tapping.