From complaint data to a prioritized roadmap

The pipeline

User complaints→ Problem matrix→ 4 clusters→ How-Might-We goals→ Crazy 8s→ ~30-idea feature pool→ RICE scoring→ PRDs + wireframes→ User testing (44)→ Roadmap

Step 1 — Plot the pain, don’t rank opinions

Real complaints were plotted on problem intensity × problem frequency, then grouped into four clusters. High-intensity/high-frequency items became the shortlist.

Cluster 1

UI / UX issues

Poor UI/UX · app not working or vague errors · OTP not received · crashes and freezes · in-app navigation · interface language.

Cluster 2

Operational inefficiencies

Long confirmation and wait times · driver ETA inaccuracy · “driver was not moving towards pickup” · driver not responding.

Cluster 3

Trust & perception

“Found another ride” · fare discrepancy after the ride · rude behaviour and extra demands · poor customer support with no real-time help.

Cluster 4

Feature gaps vs competitors

App reliability · advanced scheduling · female-focused safety features · general feature parity with other ride-hailing apps.

Each cluster was converted into How-Might-We goals — e.g. “HMW optimize auto allocation to lower wait times?” and “HMW address post-ride fare conflicts to avoid dissatisfaction?”

Step 2 — Score the pool

Crazy 8 sketching produced roughly 30 candidate features. Every one was put through RICE.

RICE = (Reach × Impact × Confidence) ÷ Effort. Reach /10, Impact /5, Confidence /1, Effort /5. Top five were shortlisted for PRDs.
Feature	R	I	C	E	RICE
One-Tap Payment	9.0	3.0	0.90	2.0	12.15
Ride Modification & Scheduling	8.0	3.0	0.85	3.0	6.80
Real-Time Traffic Intelligence	7.0	2.5	0.85	2.0	5.95
Secure Routes	7.0	2.5	0.85	3.0	4.96
AI Voice-Assisted Search	6.0	3.0	0.80	3.0	4.80
Visual Meter for Fare	6.0	1.5	0.80	2.0	3.60
Pre-ride Comfort Checks	5.0	2.0	0.75	2.0	3.55
Live Vehicle Density	8.0	2.0	0.65	3.0	3.47
Priority Pass	6.0	2.0	0.70	4.0	2.10
Request Any with Customisation	6.0	1.0	0.70	2.0	2.10

The North Star

Conversion rate = Completed Trips ÷ Searches

One number that only moves if the whole loop works

Step 3 — Specify, then test

Each shortlisted feature got a mini-PRD: problems solved, key metrics, user stories with acceptance criteria, and a wireframe. Example — One-Tap Pay:

Problems solved	Key metrics
Long confirmation / wait times · fare discrepancy after ride · driver demanding extra	Payment success rate · average payment time · dispute rate

Acceptance criteria were written at the screen level — e.g. “when the ride ends, an in-app notification shows the exact fare with a visible Pay with UPI CTA”; “the fare amount and driver’s UPI ID are pre-filled”; “trip status changes to Payment Completed.” The backlog was then scheduled in Jira across August–October with dependencies.

Step 4 — 44 users, and the scores that changed the plan

Mockups were tested with 44 users on a 1–5 Likert scale. This is the part that earned its keep: two features moved on the evidence.

Mean Likert scores (1 = least, 5 = most), n = 44, collected via Qualtrics on mockups.
Feature	Value add	Ease of use	Likely to use	Overall
One-Tap Pay	4.68	4.75	4.61	4.68
Secure Routes	3.91	4.80	3.94	4.22
AI Voice Assistant	3.55	4.50	2.91	3.65
Real-Time Ride Modification	4.05	2.61	4.05	3.57
Real-Time Traffic Intelligence	4.02	2.73	3.73	3.49
Visual Meter for Fare	1.57	3.73	3.27	2.86

Visual Meter for Fare was demoted. It scored 1.57 on value addition — easy to use, pleasant, and unwanted. Postponed or reworked rather than shipped.
Ride Modification and Traffic Intelligence scored low on ease of use (2.61 / 2.73) despite high value — flagged for UX rework and beta testing, not cancellation.
AI Voice scored high on value and ease, low on likelihood — kept, but re-justified as an accessibility feature rather than a mainstream convenience.
Secure Routes validated the safety thesis — the highest ease-of-use score in the set, addressing late-night travel anxiety.

The roadmap that came out

Short · 1–2 months

Quick wins

One-Tap Payment (direct UPI checkout) · traffic-based auto rematch.

Mid · 3–5 months

Trust & personalization

Secure Routes · real-time ride modification (add stops, change destination).

Long · 6+ months

Innovation

AI voice-assistant booking.