← Files Intuitive Software DesignARCHIVED FILE
skills/intuitive-software-design/references/examples.md
8.63 KB · Oct 5, 2026 · 18:35 UTC
# Intuitive Software Design Examples
These examples teach the standard; they are not universal component prescriptions. Always adapt them to the intended user, task, consequence, and evidence.
## 1. Orientation — configuration scope
### Before
```text
Settings
Notifications Permissions Defaults
[ Save ]
```
### After
```text
ACME HOSPITAL / ESTIMATE 1042
Pricing settings
Labor default Commercial crew ▼
Markup 18%
Unsaved changes [ Save pricing settings ]
```
### Why the second is more intuitive
- object and configuration scope are visible;
- the action names the consequence;
- changed state is visible;
- the user need not reconstruct the navigation path.
### Standard mapping
UI Orientation; Recognition; Interpretation and Memory Friction; Context preservation.
## 2. Prediction — vague completion action
### Before
```text
Proposal editor
[ Cancel ] [ Done ]
```
The control closes the editor, saves the draft, and sometimes sends it when an `Auto-send` preference is enabled elsewhere.
### After
```text
Proposal editor
Saved 2:41 PM
[ Close ] [ Save draft ] [ Review and send ]
```
### Why the second is more intuitive
- each action has a predictable consequence;
- saved state is visible;
- sending remains a distinct high-consequence step;
- the hidden preference no longer creates a large Prediction Gap.
### Standard mapping
Feel Predictability; Predict; Interpretation Friction; Critical Failure candidate when sending is binding.
## 3. Direct interaction — team role
### Before
```text
Project
⋮
Manage
Project configuration
Personnel
John Smith
Edit
Role
```
### After
```text
PROJECT TEAM
John Smith Foreman ▼
Maria Lopez Carpenter ▼
+ Add team member
```
### Why the second is more intuitive
- relevant objects and current state are visible;
- actions are located next to affected objects;
- fewer unrelated transitions are required;
- consequences are easier to predict.
### Design boundary
If role changes alter payroll, access, or legal responsibility, keep the direct entry point but add a proportionate explanation/confirmation or focused permission flow.
## 4. Feedback — proposal sending
### Before
```text
[ Send proposal ]
```
After activation the button remains unchanged for eight seconds. The user is returned to the list with no durable status.
### After
```text
[ Sending… ]
Proposal sent to bids@acme.example at 2:43 PM ✓
[ View proposal ] [ Copy link ]
```
Failure:
```text
Proposal was not sent. Your draft is preserved.
Connection timed out. [ Try again ] [ Save draft ]
```
### Why the second is more intuitive
- immediate acknowledgment prevents duplicate activation;
- pending, success, and failure are distinct;
- the result names what happened;
- recovery preserves work and makes continuation clear.
### Standard mapping
Feel Feedback; Confirm and Continue; Feedback and Recovery Friction; Critical Failure candidate for consequential sending.
## 5. Recognition over recall — estimate to scope
### Before
```text
Step 2: Scope
Type __________
Area __________
```
The user must remember the customer, estimate type, and building area selected on the previous screen.
### After
```text
ESTIMATE 1042 / ACME HOSPITAL
Commercial renovation · 42,000 sq ft
Step 2 of 4 — Define scope
Scope type ▼
Included areas ▼
```
### Why the second is more intuitive
- previous choices carry forward;
- current object and progress remain visible;
- options can be contextualized to the chosen estimate type;
- backtracking and memory work decrease.
## 6. Progressive disclosure and Decision Density
### Before
```text
NEW PROJECT
Name
Customer
Project type
Project subtype
Billing method
Workflow
Labor model
Markup
Tax configuration
Location profile
Crew type
Notification policy
[ Create ]
```
### After
```text
NEW PROJECT
Name __________
Customer __________
Type ▼
Defaults: Standard workflow · Customer tax profile
[ Review settings ]
[ Create project ]
```
After creation, labor, crew, and workflow choices appear when scope and team information make them meaningful.
### Why the second is more intuitive
- only necessary creation decisions are required now;
- consequential defaults are summarized, not hidden;
- additional settings remain discoverable;
- decisions appear with better information.
### Design boundary
Do not defer a choice if work becomes unsafe, invalid, unexpectedly billed, or costly before the user sees it.
## 7. Error recovery — preserved work
### Before
```text
Error: invalid request.
[ Back ]
```
All 18 entered fields are cleared.
### After
```text
We could not create the project yet. Your entries are preserved.
2 fields need attention:
- Start date must be on or after August 25. [ Go to Start date ]
- Customer billing profile is inactive. [ Choose another profile ]
[ Review fields ]
```
### Why the second is more intuitive
- cause and location are specific;
- valid work remains;
- recovery actions correspond to each cause;
- retry does not risk duplicate creation.
## 8. Navigation — contextual action
### Before
```text
Project detail -> Administration -> People -> Assignments
```
The user leaves the project and must search for it again in a global assignment screen.
### After
```text
Project detail / Team
+ Assign team member
```
The assignment picker is already scoped to the project and returns to the same Team view.
### Why the second is more intuitive
- the entry point appears at the moment of need;
- project context is preserved;
- system administration no longer dictates the job flow;
- return location is predictable.
## 9. Minimal but cognitively complex
### Before
```text
Project 1042
✦ ◇ ↗
```
The icons mean calculate, duplicate, and send. None has a visible label.
### After
```text
Project 1042
[ Recalculate ] [ Duplicate ] [ Send proposal ]
```
On narrow screens, the two lower-frequency actions may move into a clearly labeled `More actions` menu while `Send proposal` remains visible.
### Why the second is more intuitive
- visual simplicity no longer creates interpretation work;
- actions and consequences can be recognized;
- priority remains clear across viewport sizes;
- cognitive simplicity takes precedence over icon minimalism.
## 10. Dense expert interface
### Deliberately dense screen
```text
CREW SCHEDULE — WEEK 35
Filters: Region [North] Trade [All] Status [At risk]
Rows: 42 crews × 7 days, color + text status, utilization %, conflicts
Keyboard: J/K navigate · E edit · Shift+S assign · ? shortcuts
Saved view: Superintendent morning review
```
### Correct evaluation
Do not penalize the screen because it contains many rows, abbreviations, or expert shortcuts. Evaluate:
- whether dispatchers recognize the grid and domain terms;
- whether hierarchy separates conflicts from routine work;
- whether state is not color-only;
- whether shortcuts are discoverable and visible actions remain available;
- whether edits preserve week, filters, selection, and scroll position;
- whether assignment results and conflicts are confirmed.
### Possible legitimate finding
If selecting a crew silently resets the region filter and scroll position, that is Context/Memory Friction and a Feel inconsistency. “Reduce the number of rows” is not supported by that evidence.
## 11. Beginner-to-expert progression
### Before
```text
Save is available only as Ctrl/Cmd+S.
```
### After
```text
[ Save ]
Shortcut shown in tooltip/menu: Ctrl/Cmd+S
Saved view and bulk-save patterns available to frequent users.
```
### Why the second is more intuitive
- beginners can recognize the action;
- the shortcut becomes discoverable;
- expert efficiency grows without removing the visible path;
- the system supports mastery rather than choosing one audience.
## 12. Smallest complete improvement
### Diagnosis
The existing proposal screen is clear, but sending provides no pending, success, failure, or durable sent state.
### Incomplete “small” change
Show a green toast saying `Sent`.
### Smallest complete change
Keep the layout and button. On activation:
1. change the action to `Sending…` and prevent duplicate activation;
2. on success, show `Sent` with recipient and time in durable proposal state;
3. on failure, state that sending failed, preserve the draft, and expose safe retry;
4. ensure the result is semantically announced;
5. keep `View proposal` as continuation.
### Why this is the standard's preferred improvement
It fixes the complete feedback contract without redesigning the screen, introducing a new navigation layer, or changing behavior that already works.
SHA-256: 48caef4163e95fd0e92aef0b340dbf9af80040de6eae4717d539e4c02bc3f5c1