PayU's

Partners Platform

Partner ecosystem from a KAM-dependent onboarding process into a scalable self-service platform—reducing onboarding drop-off by 40%, accelerating merchant activation, and improving operational efficiency.

Industry

Jio ecosystem · web · mobile · WebView · POS

Platform

B2B Web App

Duration

8-10 Weeks

Problem & Context

PayU’s partner programme — developers, resellers, and enterprise platforms who refer merchants and earn incentives — was a strong growth channel on paper: 300 active partners, 350,000 referred merchants, 40 billion in transaction volume.

In reality, the platform was failing everyone involved.

The business problem:

40% of partners dropped off during merchant onboarding — direct revenue loss
Every step that should have been self-serve required a KAM (Key Account Manager) call
Business was absorbing the cost in support time and lost conversions
60+ pain points across 7 journey stages, documented but unaddressed
No transparent timelines — KYC verification showed “TBD” instead of actual days Incentive calculations were a black box — partners couldn’t verify their own earnings

40% of partners dropped off during merchant onboarding — direct revenue loss
Every step that should have been self-serve required a KAM (Key Account Manager) call
Business was absorbing the cost in support time and lost conversions
60+ pain points across 7 journey stages, documented but unaddressed
No transparent timelines — KYC verification showed “TBD” instead of actual days Incentive calculations were a black box — partners couldn’t verify their own earnings

Research & Discovery

This is the part that reframed the whole project for me. The brief wasn't "fix the chatbot." From the kickoff:

The brief said “redesign the platform.” I resisted the brief — a redesign brief almost always hides a more specific problem. Before opening Figma, I needed to know what redesign actually meant here.

The business problem:

14 in-depth user interviews across 5 partner groups — deliberately including partners who quit
Groups: active resellers, new resellers, stuck resellers (abandoned midway), enterprise platform partners, internal sales stakeholders
Gap analysis — mapping user needs at each journey stage vs what the platform offered
Support ticket + KAM call log analysis — what were partners actually calling about
Journey mapping across all 7 stages from signup to incentive payout

14 in-depth user interviews across 5 partner groups — deliberately including partners who quit
Groups: active resellers, new resellers, stuck resellers (abandoned midway), enterprise platform partners, internal sales stakeholders
Gap analysis — mapping user needs at each journey stage vs what the platform offered
Support ticket + KAM call log analysis — what were partners actually calling about
Journey mapping across all 7 stages from signup to incentive payout

In Their Own Words

10 end users and 4 internal stakeholders interviewed across developer, reseller and enterprise personas. These moments defined what the redesign had to solve.

Key Insights

The stuck resellers gave the most useful data: Recruiting people who quit — not just people succeeding — revealed the real failure points. Their abandonment reasons were structural, not cosmetic.

It was never about the UI

No transparent timelines, black-box incentives, forced KAM dependency at every step. The platform structurally didn’t support self-serve at any stage.

The navigation reflected PayU’s org chart, not partner jobs:

11 nav items ordered by PayU’s internal product structure. A reseller thinks: add a merchant, check if they’re live, see what I’ve earned. The IA fought them at every step.

KAM calls were a symptom, not a preference:

Partners didn’t want to call KAMs. They called because the dashboard gave them no other path. Every call was a self-serve failure logged as a support interaction.

Design Constraints

The challenge extended far beyond improving usability.

The platform had to support multiple partner types, reduce operational dependency, and integrate with existing business processes without disrupting ongoing merchant onboarding.

The challenge extended far beyond improving usability.

The platform had to support multiple partner types, reduce operational dependency, and integrate with existing business processes without disrupting ongoing merchant onboarding.

We designed within five key constraints::

Multiple partner personas with different workflows
Complex merchant verification and compliance requirements
Legacy information architecture centred around internal teams
High dependency on Key Account Managers
Incremental rollout without affecting existing business operations

Multiple partner personas with different workflows
Complex merchant verification and compliance requirements
Legacy information architecture centred around internal teams
High dependency on Key Account Managers
Incremental rollout without affecting existing business operations

Design Approach & Principles

The strategic shift: restructure the platform around partner jobs-to-be-done, not PayU’s product taxonomy — and make every journey stage self-serve by default, with KAM escalation as the exception.

We designed within five key constraints::

Task-based IA — navigation follows what partners do, not what PayU sells
Radical transparency — every process shows real timelines, every incentive shows its math
Self-serve first — KAM contact is an escalation path, never a requirement
Progressive onboarding — partners see value before completing full KYC
Status visibility — partners always know where every merchant is in the pipeline

Task-based IA — navigation follows what partners do, not what PayU sells
Radical transparency — every process shows real timelines, every incentive shows its math
Self-serve first — KAM contact is an escalation path, never a requirement
Progressive onboarding — partners see value before completing full KYC
Status visibility — partners always know where every merchant is in the pipeline

Ideation & Exploration

Before wireframing a single screen, I restructured the information architecture. Five primary sectionsreplaced eleven — organized around the partner’s job-to-be-done.

Create a single trusted AI identity across all Jio products — so users always know they're talking to the same assistant, regardless of the surface?

Design conversations that feel intelligent and contextual — reducing dead ends and eliminating the need to repeat yourself across sessions?

Build a scalable design system for conversational AI — so product teams across Jio can ship new chat features without fragmenting the experience again?

Build one trusted AI identity across all Jio products

One voice, one component library, one behaviour spec. Whether a user is in MyJio or JioLearn, the AI should feel like the same assistant.

Reduce conversation turns to task completion

A well-designed AI shouldn't need 5 back-and-forths for a simple recharge. We set a target: resolve 70% of intents in under 3 turns.

Ship a system teams can build on independently

Not just Figma frames — a documented, handoff-ready component library that product and engineering teams across Jio could extend without breaking consistency.

Design Decisions

Restructured IA from 11 product-based items to 5 task-based sections

What: Navigation reorganized around partner jobs: Add Merchants, Track Status, Earnings, Resources, Account.Why: Research showed partners think in tasks, not PayU’s product categories. This single structural change reduced time-to-first-action across every flow measured.

Architecture -

Single AI pattern library

One component set shared across MyJio, JioLearn, JioCinema and JioChat.

Conversation -

Context persistence

AI retains intent within session. Conversation state travels across escalations.

Escalation -

Intelligent handover

3 escalation points per flow. Agent receives full context summary upfront.

Personalisation -Proactive suggestions

AI surfaces relevant options based on user history — recharge due, last issue, open ticket.

The complete Partner Journey

From discovery to first incentive — 4 key stages, each redesigned to remove human dependency from the critical path.

Partner Dashboard

Your activity, referral stats, and next steps — all above the fold. Partners know exactly where they stand without calling anyone.

Referral Status Tracker

Live status for every referred merchant — pending, active, or stuck. Partners track progress and earnings in real time, not over email.

Refer Now

One inline form to submit a new referral. What was a multi-step offline process is now a 60-second action from the dashboard.

User & Permissions Management

Team accounts, roles, and access — all self-managed. Enterprise partners can onboard their team without raising a support request.

KYC Activation Flow

Merchants activate PayU in WhatsApp Business with a 5-step KYC process. It includes live verification and transparency at each step.

Completion & Cross-sell

The "All set" state provides clear next steps and auto-redirects to WhatsApp, turning success into discovery by showcasing relevant products with one-tap interest capture.

Your incentive, always in sight

The biggest driver of churn was mistrust. Partners had no idea what they'd earned. The redesign surfaces daily incentive amounts, merchant-level breakdowns and a quarterly claim trigger — all without contacting anyone.

Outcomes & Impact

On the numbers

CSAT (from 2.6)

4.2/5

Call volume drop

55–60%

Avg resolution time

<90s

Users reached

400M+

Where it goes next

Onboarding drop-off reduced by 40% — directly recovering lost activations
Onboarding completion 3x faster with self-serve flows and inline validation
KAM call volume significantly reduced — calls shifted from routine status checks to genuine escalations
Task-based IA adopted as the pattern for other PayU B2B products
Incentive transparency eliminated the #2 support call category entirely

Onboarding drop-off reduced by 40% — directly recovering lost activations
Onboarding completion 3x faster with self-serve flows and inline validation
KAM call volume significantly reduced — calls shifted from routine status checks to genuine escalations
Task-based IA adopted as the pattern for other PayU B2B products
Incentive transparency eliminated the #2 support call category entirely

Reflections

What worked

Resisting the “redesign” brief and finding the structural problem first — the IA restructure did more than any visual changeRecruiting stuck/churned partners for research — their data was the most actionableFixing trust (timelines, incentive math) before fixing aesthetics

What I'd do differently

Involve KAMs earlier — they initially saw self-serve as a threat to their role; earlier alignment would have smoothed rolloutInstrument the funnel before redesign started — baseline data took weeks to reconstructPhase the IA change more gradually — existing partners needed more migration support than anticipated