Aerathe Rajeev Nair

Orchestrating a 7-app enterprise rebuild

Turned seven siloed app projects into one KPI-driven transformation.

JK Cement, with BCG, Salesforce and Stetig Engineering2024–2552 weeksProgram governance
JK Cement app suite
01

Outcomes

3 follow-on engagementsThe client came back for more work after the program
5 organisations, 1 roadmapShared success criteria across client, consultants, engineering and design
Rework avoidedData contracts validated before design freeze
UX in UATEdge cases, empty and error states made acceptance criteria for the first time
Architecture settled in 2 weeksA feasibility spike replaced a design-vs-engineering stand-off
Scope grew by evidenceA legacy loyalty app brought into the program, with client budget
02

Context

Seven interconnected apps (Sales, Supply, Dealers, Retailers, Influencers, CTS, SFA) with multiple owners. Requirements documents disagreed, goals shifted mid-stream, and a Salesforce constraint forced an architectural pivot.

There were no success criteria, no MVP, and no clarity on API or backend readiness. Design rework risk was high.

03

Two decisions that shaped the program

Custom or Salesforce Lightning? Design planned custom builds for the CTS and SFA apps; the engineering vendor planned to use Lightning's pre-built components. Instead of letting the teams argue it out, I ran a two-week spike: designers produced representative screens, engineers tried to build them. With client IT, business, BCG, engineering and design in one room, we agreed a hybrid: Lightning for internal staff apps, custom for dealer and partner apps.

A loyalty app nobody had scoped. The dealer loyalty app, unchanged since 2013, moved the same KPIs we were trying to lift. Usability tests and a prototype of a better redemption flow showed the client how it was leaking value from their investment. They approved the budget, and the redesign joined the main program.

More on how I approach problems like these →

04

What I did

  • Defined success criteria from one shared set of KPIs
  • Set up change control and a decision log
  • Ran workshops to separate MVP from enhancements, and to lock data contracts
  • Added two feasibility gates, backed by time-boxed technical spikes
05

The phase-gate model the program ran on

  1. IntakeRequirements, success metrics, RACI, SOW
  2. DiscoveryStakeholder map, ambiguity log, KPI mapping
  3. Gate 1WireframesBackend feasibility, API checks at 60%
  4. Gate 2UILatency, edge and error states before freeze
  5. HandoffAnnotated specs, UX acceptance criteria
  6. QA → UAT → LaunchUX and functional QA, retros

It ran as one transformation, not seven design projects.