Back to all work

UI/UX Design · PM SalesForce

From process-first to intent-first: redesigning the Request Center

PMSF supports internal teams in managing product-related requests such as sampling, quotations, and approvals. Users consistently struggled with the Request Center, leading to high drop-off rates, incorrect submissions, and long onboarding times. I led user-centered design to simplify the workflow and align it with business goals.

Role
Senior UX/UI Designer
Timeline
13 weeks
Platform
Web (B2B)
Focus
Stakeholder AlignmentPrototypingAI-assisted UX
PMSF platform overview

Problem Overview

Users didn't know how to start

Users didn't know how to start or what information was required, especially for complex requests. The system was built for processes, not people.

01

High Drop-off Rates

Users abandoned submissions mid-process.

02

Incorrect Submissions

Operations team wasted time on clarification.

03

Long Onboarding

New users relied on colleagues instead of the system.

Goals & Success Metrics

Business goals meet user goals

Biz

Reduce back-and-forth

Improve request accuracy and increase completion rate.

User

Submit with confidence

Reduce cognitive load; understand requirements without reading long documentation.

Research & Discovery

Understanding how people actually think

Stakeholder interviews with Product and Ops, 1:1 interviews with 8 internal sales users, shadowing real request submissions, and analysis of failed or rejected requests all pointed to the same root cause.

01

Intent, Not Process

Users think in intent, not process: forms felt intimidating and "too technical."

02

Reused Old Requests

Starting fresh was painful, so users duplicated past requests instead.

03

Relied on Colleagues

New users leaned on peers rather than trusting the system.

The Insight

How might we help users submit requests accurately without forcing them to understand the system logic? This led to an AI-assisted Request Flow that understands user intent, guides users step by step, and auto-fills and validates data dynamically.

Discovery & Research

Designing inside someone else's platform

PM SalesForce is a third-party tool, and I don't own the roadmap. Research had to separate what reps genuinely needed from what the platform's defaults assumed they needed.

Note on confidentiality: these are internal enterprise platforms. Due to company data policy, full working files, raw research data, and complete dashboards can't be published here. The previews below use illustrative, non-production data to represent the structure, format, and thinking behind each deliverable.

Persona: Daniel Wong

Research
Daniel WongRegional Sales ManagerReviews pipeline dailyMobile access neededReporting frequency

Represents the regional manager who relies on PM SalesForce to track pipeline health and coach the team.

Persona: Lisa Chong

Research
Lisa ChongTerritory RepUpdates opportunities dailyLogs activity on the goComfort with tech

Updates deals between client visits and resents any field that makes her type the same thing twice.

Persona: Aznan Yusof

Research
Aznan YusofSales Operations AnalystBuilds custom reports weeklyAudits data qualityComfort with tech

Builds the reports leadership actually reads, and spends real time cleaning up what reps leave incomplete.

Persona: Farah Idris

Research
Farah IdrisVP of Sales (Executive)Reviews forecast monthlyNeeds summary dashboardsComfort with tech

Wants a trustworthy forecast number without digging through the platform herself.

Proto Persona: Territory Rep

Research
Territory Rep (proto)ASSUMPTION-BASED · LOW FIDELITYNeed: log activity withoutduplicate data entryFrustration: re-entering thesame account data across two

Assumption persona used to frame the rep side of the workflow ahead of field interviews.

Proto Persona: New Regional Manager

Research
New Regional Manager (proto)ASSUMPTION-BASED · LOW FIDELITYNeed: trust the forecastnumbers withoutFrustration: doesn't knowwhich fields reps tend to skip

Used to frame manager-level trust issues with the data, before manager interviews were scheduled.

Proto Persona: Sales Ops Analyst

Research
Sales Ops Analyst (proto)ASSUMPTION-BASED · LOW FIDELITYNeed: pull custom reportswithout filing a vendorFrustration: the reportbuilder is locked behind admin

Assumption persona framing the reporting workflow ahead of ops team interviews.

Proto Persona: Marketing Coordinator

Research
Marketing Coordinator (proto)ASSUMPTION-BASED · LOW FIDELITYNeed: see which leads salesactually followed up onFrustration: no visibilityinto the platform outside the

Assumption persona representing an adjacent team blocked by the platform's access model.

Empathy Map

Research
SAYS"Why am I typing thistwice?"THINKS"Is this report evenaccurate?"DOESUpdates opportunitiesfrom memory at the endof the weekFEELSDisengaged from dataentry

Synthesized from interviews with 8 territory reps and 3 regional managers.

Stakeholder Map

Research
ProductTerritory RepRegional ManagerSales OpsVendor (Platform Owner)FinanceMarketingDirectInfluencing

Distinguishes internal users of PM SalesForce from the third-party vendor and adjacent teams.

Affinity Diagram

Research
DATA ENTRY"Duplicate fields""Too many clicks"REPORTING"Reports hard tocustomize""Forecast feelsmanual"MOBILE"App feels slow""Missing offlinemode"

Raw research notes clustered into themes during a synthesis workshop, surfacing the patterns behind user pain points.

Competitive Benchmark

Research
UsabilitySpeedMobile UXSupportCustomizationUsCompetitor avg

Scored comparison against comparable enterprise platforms across usability, performance, and support dimensions.

Design Response

An AI-powered entry point that asks, not demands

01

Entry Point

Instead of a list of request types, users see: "What would you like to request today?", with examples, or the option to upload a relevant document.

02

Intent Recognition

AI classifies intent into request type, required modules, and mandatory data, removing the need for users to decide upfront.

03

Guided Questions

AI asks progressive questions only when needed: product category, quantity range, delivery urgency, reducing form fatigue.

04

Smart Form Generation

The system builds the correct form, pre-fills known data, highlights required fields, and explains why information is needed.

05

Review & Submit

Users review a clean, structured summary before submission, with the ability to re-edit, building confidence in their request.

Journey & Flow

Working within the vendor's architecture

Information architecture wasn't mine to redesign, so the flow work focused on the parts I could influence: how reps move through their day-to-day pipeline tasks.

Journey Map

Journey
Open appFind accountUpdate stageAdd notesSave3.9

End-to-end emotional curve across the pipeline update journey, from first touch to task completion, mapped against satisfaction (1–5).

Service Blueprint

Journey
FrontstageBackstageSupport Systems

Frontstage actions, backstage operations, and supporting systems mapped against each other to expose handoff gaps.

User Flow

Journey
OpenpipelineSelectaccountDatacompleteUpdatestageLogactivitySave

Primary task flow through the product, including key decision points and drop-off risk areas.

Outcome & Impact

Errors down, trust up

Fewer Errors

Submission errors reduced significantly.

Faster Onboarding

New users ramped up faster on the system.

Ops Workload Reduced

Higher trust in the system meant less manual clarification work.

Impact

Current Achievements in DKSH

Scale and ownership across the PM SalesForce (Salesforce CRM) platform, from strategy through to delivery.

  • Led UX improvements for a Salesforce CRM used across 20+ global markets.
  • Redesigned 30+ CRM workflows, including quotation, opportunity and pipeline management.
  • Improved sales-process efficiency by an estimated 20–30%.
  • Simplified quote-to-order journeys, reducing manual workflow complexity.
  • Collaborated with international stakeholders across Europe and APAC.
  • Built reusable UI patterns to improve consistency across Salesforce modules.

Strategy & Prioritization

Making the business case inside vendor constraints

Every recommendation had to be justified against what the vendor platform could actually support, so each metric below is reported as before → after against the pre-redesign baseline.

VOC Dashboard

Voice of Customer
BeforeAfterPipeline visibility12% (was 27%)Mobile access10% (was 24%)Report customization9% (was 21%)Data entry duplication7% (was 19%)Integration lag5% (was 14%)

Share of support tickets and feedback tied to each theme, before the redesign shipped versus today.

CSAT Dashboard

Voice of Customer
86%CSATwas 55%

Customer satisfaction climbed from 55% to 86% across 190 post-task survey responses.

CES Dashboard

Voice of Customer
"It was easy to complete my task"5.7/7was 3.2/7wasCUSTOMER EFFORT SCORE

Effort score improved from 3.2 to 5.7 out of 7 across 175 responses. Lower effort correlates with higher retention.

Power BI Dashboard

Behavioural Analytics
Active usersConversion %FebMarAprMayJunJul

Monthly active users against conversion rate, tracked to monitor adoption after each release.

Prioritization Matrix

Prioritization
Big betsQuick winsMoney pitFill-insAuto data syncSimplified entryCustom reportsMobile offlineForecast automationActivity remindersEffortImpact

Candidate initiatives scored on impact vs. effort to sequence the roadmap.

KPI Dashboard

Strategy
84%▲ was 68%Forecastaccuracy78%▲ was 44%Adoption91%▲ was 77%Datacompleteness5.4m▲ was 7.1mOpportunityupdate time

North-star and supporting metrics compared against their pre-launch baseline to show whether the redesign is actually working.

Business Impact

Strategy
BeforeAfterData entry time5.4m3.9mForecast accuracy68%84%Report build time22m8m

Operational metrics compared before and after launch to quantify the business case.

Executive Summary

Strategy
Simplified a third-party platform withoutowning the roadmapLed UX improvements within vendor platformReduced duplicate data entry throughAdoption and forecast accuracy both improved24%Faster updates+16ptForecast accuracy45NPS

One-page leadership readout summarizing problem, approach, and measurable outcome.

Key Learnings

Good AI UX doesn't feel like AI. It feels like clarity. The most successful part of this project wasn't showcasing AI capabilities, but removing friction from the user's mental model. Shifting from "process-first" to "intent-first" design fundamentally changed how users interacted with the system, reducing cognitive load and building confidence.