The Partner Experience Paradox

The workplace nobody was designing anymore

Porter is one of India's largest intra-city logistics marketplaces, and the driver-partner app is where its supply side lives — the tool 2-wheeler riders and truck drivers use to go online, accept orders, run trips, collect cash, and track earnings. It's not a companion app; it is the partner's workplace.

That workplace had accumulated years of debt: flows bolted on per requirement, a hamburger menu hiding a sprawling surface, communication decayed into banner spam, and a live-order flow still running on a bottom-card pattern competitors had left behind. The app worked. It had just stopped being designed.

And the partner is a brutal user to design for — on the road, one-handed, glancing for split seconds, often low-literacy, across many Indian languages, under time and income pressure. Every decision in this story traces back to that reality.

First, prove the 4.7★ is lying

The app's 4.7 stars across 530k+ reviews looked like health — and quietly argued against spending anything on a revamp. So before designing, I built the strategic case. I audited public Play-Store reviews, complaint forums, and Reddit, and synthesized the signal into an analysis I called "The Partner Experience Paradox." Underneath the rating, dissatisfaction clustered into four systemic pillars:

1. Economic instability — opaque commissions, a "black-box" wallet (one documented case: a ₹1,800 trip paying out ₹1,200, unexplained), and the "honeymoon & cliff" volume pattern.
2. Technical fragility — most cited: the "Start Trip" bug, where tapping before reaching pickup wipes the customer's details. A single review describing it drew 313+ helpful votes.
3. Support collapse — the recurring phrase across every platform was "pathetic support": callback black holes, no in-trip line, unempowered agents.
4. Trust deficit — forced acceptance and punitive suspensions that left partners feeling, in their words, "treated like slaves."

(These figures are public research evidence I compiled to justify the revamp — the "why now" — not outcomes I'm claiming.)

The brief this produced: not a reskin — rebuild the relationship through transparency, a working core, and real communication.

Three research strands

I lived it. In January 2025 I onboarded as a real 2-wheeler partner and ran actual deliveries. That immersion surfaced ~40 concrete UX issues — and I ran sessions to transfer the learnings to the whole design team, so the empathy became shared rather than anecdotal.

I benchmarked it first-hand. I installed 12+ competitor partner apps — Rapido, Uber, Swiggy, Zomato, Grab, Zepto, Ola, Shadowfax, Amazon Flex, Flipkart, Uncle — ran their real flows, and annotated every screen myself. Seven cross-cutting patterns fell out, mapping directly onto the four problem pillars: live-map order flows (not bottom cards), earnings-first wallet unification, suspension-avoidance controls, a communications inbox instead of banner spam, demand heatmaps and slot booking, onboarding and welfare as differentiators, and delight micro-interactions at key moments.

And it stays live. A fresh partner FGD round fed the strategy this year, and a card-sort study with partners validates the new navigation before it's built.

Nine principles and a five-tab spine

The audit and benchmarking converged into nine design principles, deliberately split into how the product behaves (proactive transparency · muscle-memory consistency · efficiency balanced with comprehension · context-driven personalization · psychological safety) and how it looks (show-don't-tell · Porter's in-transit visual language · Material-3-expressive fundamentals · fluid interactions, deliberately deprioritized). Split that way, they're testable — not poster decoration.

The architecture they produced: a mental-model-driven 5-tab bottom nav (Home · Orders · Earnings · Updates · More) replacing the hamburger — with a context-driven Home that changes by partner state: Offline (planning: heatmaps, deep tools) → Online-idle (execution: stripped to what matters) → Online-active (the live order flow).

The alignment machine: 152 screens, three colours

A revamp on a surface this contested doesn't die from bad pixels; it dies from misalignment. The questions were real: what's a cheap facelift vs. an expensive deep revamp? Who owns what? What will engineering actually commit to?

My core move was building a shared language for the argument: I broke the entire app into a 152-screen triage and colour-coded every screen — done (no change) · facelift · deep revamp. Vague debates became a countable plan: ~48 done · ~62 facelift · ~42 deep revamp — proof that most of the app ships cheap, and the expensive work is concentrated and ownable.

The hard calls that board let us broker: scoping the homepage into three tiers of increasing backend cost (another designer owns the facelift tiers; the full UX-revamp tier is mine); arguing polish like "liquid-glass" interactions down the priority list because it buys no partner-critical comprehension; and sequencing the order-life-cycle work first because an OLC backend refresh was already underway — design riding engineering that's already in motion.

From there I sequenced seven research-gated phases — OLC → communications → homepage + IA migration → suspension → earnings → wallet → the long-tail facelift sweep — each gated on partner validation, not just build order. I authored that phasing plan and presented it to the CTO to commit engineering capacity. Presenting a delivery strategy at that altitude — not designs — is the most senior work in this story.

The flows I own end-to-end

OLC — the live order flow. The deepest redesign. I designed its full architecture (accept → trip phases → cash collection → rating) plus a six-surface taxonomy — map, bottom card, expanded sheet, toast, bottom sheet, full-screen blocking banner — so every OLC feature maps to exactly one rendering surface. Concrete decisions, each answering a researched pain: geofence the "Start Trip" button (killing the 313-vote bug outright), a genuinely live map, a transparent fare breakdown (gross − commission − payout, cash vs. digital explicit), ratings that capture why, an end-of-order stats screen, and a "stop new requests" control so partners stop getting chained into orders they can't fulfil.

Earnings. Realizes the earnings-first hierarchy from the IA — track · strategize · consolidate — with the wallet folded in and the ledger demoted, straight from the benchmark insight that every strong competitor unifies the two.

Homepage, tier three. The full-revamp tier: the bottom-nav IA migration plus complete suspension-state coverage — the inflection point where facelift ends and backend-heavy revamp begins.

Where it stands — honestly

This is a live program, not a shipped retrospective, and I'd rather show real adoption than invent a launch metric:

• The 9 principles and 5-tab IA are adopted as the team's target architecture — every concept since is an increasing approximation of it.
• OLC and Earnings concepts are built and queued for partner user-testing, riding the OLC backend refresh already underway.
• The phasing plan has been in front of the CTO to commit engineering capacity phase by phase.
• Validation is designed in, not bolted on: moderated old-vs-new partner task tests on the OLC prototype, plus the card-sort study gating the IA migration.

What I carry forward

On a contested surface, the design leader's job is to build the system everyone argues inside of. The 152-screen triage, the nine principles, the phased plan — each converted a turf war into a countable, shared decision. Owning OLC end-to-end kept me honest about depth while doing it.