UX Crime Scene · 15 fictional cases
Diagnose the failure
before you redesign it.
Every case is a broken product experience. Inspect the scene, read the evidence, name the cause, choose the fix. About 10 to 15 minutes each.
Price: ₹799. Runs in your browser. Progress saves on this device.
Fifteen broken products.
One question each: why?
Each case starts with an incident brief, then six steps from the scene to the fix, then a debrief that compares your call with the case record.
15
fictional incident cases
7
product categories
4
levels, from Foundation to Ambiguous
6
steps per case, plus a debrief
- E-commerce 4
- Travel and booking 3
- SaaS 3
- Mobile apps 2
- Financial products 1
- Accessibility 1
- AI products 1
Counts come from the tool's own case data. Each case has its own exhibits (9 to 13 per case) drawn from 12 evidence types, such as product analytics, session observations, support tickets and error logs.
Six steps.
From the scene to the fix.
The step headings below are the tool's own. You move forward when a step has enough work in it, and your progress saves after every choice.
- 1Scene
Walk the scene. Pin what looks wrong.
Simulated product screens with pins to mark. You do not have to mark them all, and some pins are decoys.
- 2Evidence
Read the exhibits. Tag what each one proves.
Open an exhibit, then say which failure it points to and how strongly. Overstating a weak exhibit costs points.
- 3Suspects
Name your primary suspect.
Pick the failure that explains the reported behaviour, then add up to two contributing causes and rank them.
- 4Chain
Rebuild what happened, in order.
Place the events in order. Some cards do not belong to the case, and leaving them out is part of the answer.
- 5Verdict
File your verdict. Five questions.
Root cause, affected users, user and business consequences, and how confident you are.
- 6Fix
Decide what ships first, and how you will know.
Choose what ships first and what to deprioritise, then a primary metric and a guardrail.
Scored on the reasoning, not the guess.
The debrief scores each closed case out of 100 across 8 dimensions and compares your call with the case record.
Scores are exercise scores for fictional cases. They are not a certification and not proof of client work.
Score bands shown with your result
- 90+Excellent case practice
- 75–89Strong case practice
- 60–74Developing case practice
- 40–59Foundations in progress
- 0–39More practice needed
How you are scored · 100 points
- Evidence discovery15 pts
- Evidence interpretation20 pts
- Root-cause diagnosis20 pts
- Journey reconstruction15 pts
- Prioritisation10 pts
- Intervention quality10 pts
- Metric alignment5 pts
- Confidence calibration5 pts
- Total100 pts
All 15 cases.
In case order, easier cases first. Every product named here is invented.
| No. | Case | Product | Level | Time |
|---|---|---|---|---|
| 01 | The Size Selection Trap | Loomfold · Apparel store E-commerce | Foundation | 10 min |
| 02 | The Missing Delivery Date | Petalcraft · Gifting and flowers store E-commerce | Foundation | 11 min |
| 03 | The Unavailable Cart | Northgate Supply · Home and hardware store E-commerce | Investigator | 13 min |
| 04 | The Duplicate Payment Incident | Tapline · Event ticketing app E-commerce | Investigator | 12 min |
| 05 | The Hidden Total | Wanderpost · Accommodation booking site Travel and booking | Investigator | 13 min |
| 06 | The Date Picker Confusion | Railmark · Rail booking app Travel and booking | Foundation | 11 min |
| 07 | The Cancellation Maze | Staylight · Hotel booking platform Travel and booking | Advanced | 14 min |
| 08 | The Empty Dashboard | Cadenza · Team analytics SaaS SaaS | Foundation | 11 min |
| 09 | The Setup Spiral | Fielding · Field service scheduling SaaS SaaS | Advanced | 14 min |
| 10 | The Pricing Tier Mismatch | Quarterlight · Document collaboration SaaS SaaS | Investigator | 12 min |
| 11 | The Permission Ambush | Rambleway · Walking and route-tracking app Mobile apps | Investigator | 12 min |
| 12 | The Filter That Forgot | Palletown · Furniture marketplace app Mobile apps | Foundation | 10 min |
| 13 | The KYC Dead End | Vellum · Digital current account Financial products | Advanced | 14 min |
| 14 | The Keyboard Trap | Cordwell · Internal expense approval tool Accessibility | Investigator | 12 min |
| 15 | The Confident Wrong Answer | Halforth Assist · AI reply assistant for support agents AI products | Ambiguous | 15 min |
On a laptop
or a phone.
The same cases with a phone layout. Progress lives in the browser you use, on that device, and does not sync between devices.
A report for every closed case
Copy it as plain text, download it as HTML or print it. My record lists the cases you have closed.
Nothing leaves your browser
Opens in your browser through your GLE account. No AI and no network calls inside the tool. Clearing site data removes your progress.
A pilot release
The cases are authored training material. The case content still needs practitioner review.
Inspect. Diagnose.
Then fix it.
UX Crime Scene. Fictional UX incident cases for product designers.
15 cases across 7 product categories and 4 levels
Six steps per case: Scene, Evidence, Suspects, Chain, Verdict, Fix
A debrief scored out of 100 across 8 dimensions
Case reports to copy, download as HTML or print
Progress saved in this browser, on this device
UX Crime Scene
₹799
- RefundsRead our refund policy
Questions before you open a case.
Are these real companies or real data?
No. Every case, product and number is invented for training. The tool labels each case as fictional.
Is the score a certification?
No. Scores are exercise scores for fictional cases. Use them to see where your reasoning was strong or thin, not as proof of skill or client work.
Is there one right answer?
Not always. Every case has more than one defensible diagnosis, and the debrief compares your call with the case record, dimension by dimension.
Is my progress saved?
Yes, in this browser on this device, after every choice. It does not sync to another browser or device, and clearing site data removes it.
Does it use AI or send my answers anywhere?
No. Scoring runs in your browser from the authored case record. The tool makes no network calls.
How long does a case take?
About 10 to 15 minutes, as listed for each case above.
Find the cause.
Then redesign.
15 fictional cases, six steps each, and a debrief that shows your reasoning back to you.