A modernization of how a global finance organization governs its master data.
Human-centered design, made load-bearing inside a Fortune 50 finance transformation.
Johnson & Johnson's global finance organization runs on dozens of master data domains. GL accounts, cost centers, profit centers, products, vendors, trading partners. They flow through every close, plan, forecast, and reporting cycle across four operating sectors: Pharmaceutical, Consumer, Medical Devices, and Vision. Over time, the request-and-approval processes behind those domains had grown fragmented, slow, and error-prone. The friction landed in the most sensitive moments of the fiscal calendar.
Zenda was engaged to lead the human-centered design and future-state operating-model design strands of J&J's Finance Systems & Technology modernization. As Senior Design Consultant, I led the HCD research and operating-model design for the Master Data Management workstream. A six-month engagement that became the foundation for J&J's MDM modernization roadmap.
Goals
What the engagement was set up to deliver, framed as outcomes rather than activities.- Surface the actual operational reality of the master-data request process across the global finance footprint.
- Quantify the human and operational cost of the current state in terms the executive committee could act on.
- Design a future-state operating model that solved for immediate operational relief and long-term modernization.
- Make the human-centered design layer load-bearing inside the transformation roadmap, not post-hoc change management.
- Synthesize input across four operating sectors into a single Conceptual Design ready for CFO-level review.
- Stand up a quick-wins set that delivered relief before any platform decision.
Approach
Contextual inquiry across five international sites. Enterprise and Sector workshops reaching 200+ finance stakeholders. The findings reframed the problem.The HCD research was structured as contextual-inquiry field work across five international J&J sites, paired with Enterprise and Sector workshops that reached more than 200 finance stakeholders. Three sites were visited on-site. Two were researched virtually.
Jacksonville, Florida. Global Vision Care MDM team, on-site.
Manila, Philippines. Southeast Asia and North America Pharmaceutical, Medical and Consumer MDM team, on-site.
Titusville, New Jersey. Pharmaceutical MDM, on-site.
Prague, Czech Republic. EMEA Medical and Pharmaceutical MDM team, virtual.
United Kingdom. EMEA Systems and Reporting team, virtual.
Field work was paired with Enterprise- and Sector-level workshops, interviews, and document review across all four operating sectors.
The research surfaced a consistent pattern across every site, every domain, and every team. The numbers came out of contextual inquiry, not survey work. They were what the teams were already living.
Five or more parallel submission channels coexisted. Formal form. Email. Attached spreadsheet. Photocopy. Screenshot. No shared workflow. No single source of truth for what a request was, where it lived, or who owned it next.
And in Manila, a single local mapper had already developed an informal IDoc Report workaround that was achieving a 70% error reduction inside her own team. A fix that the surrounding system was actively preventing from generalizing.
"The problem the client had framed as a technology problem was, on the evidence, an operating-model problem. The form was the receipt at the end of a process no one owned end-to-end."Reframing the brief — MDM workstream, month two
The convergence
What changed structurally between current state and future state, rendered as a single shape.Solution
Two interlocking layers. A smart-form UX that absorbed error at source, and a distributed operating model that named ownership end-to-end.The future-state design worked across two interlocking layers. A smart-form UX and a distributed operating model. Both were synthesized into a single Conceptual Design Document and a quick-wins-to-future-state roadmap.
The request form was redesigned to do work the previous form had never done. Auto-populate fields from upstream systems. Validate at field-level rather than at submission. Surface ambiguity before it reached an approver. The form, as a designed artifact, could absorb a significant share of the request-error rate by preventing bad data from leaving the field.
The five parallel submission channels were consolidated into one. Global and local responsibilities were named explicitly. The Manila mapper's informal IDoc workaround was promoted from a local fix to a global pattern. The black box of the approval queue was opened. Every requester had visibility into where their request sat and what was happening to it next.
The future-state model was anchored in input from more than 200 finance colleagues across J&J's Enterprise and Sector tracks. Captured in workshops, interviews, and contextual-inquiry sessions. The Conceptual Design Document framed the modernization approach for the broader Finance Systems & Technology transformation and went into CFO-level designee review.
Alongside the future state, the work produced a set of immediately actionable changes the organization could deploy without waiting for the platform decision. Operational pressure relieved inside the first quarter, before any technology procurement.
"The Manila mapper had built a 70% error fix inside her own team. The surrounding system was preventing it from generalizing. The future-state model promoted her work from local fix to global pattern."On the distributed RACI · Solution layer two
Outcomes
What shipped, what stuck, and what the engagement set up for the platform work that followed.Additional work
Two more engagements from the same period.A 66-slide future-state vision delivered at CFO level
HCD research for Long-Term Planning and Monthly Business Review. The Four Pillars framework embedded structurally. A target GTN close cycle from three days to half a day, drawn from a peer-function baseline already operating inside J&J.
One operating-model surface for sequencing transformation investments
Four-layer integrated artifact. Fifteen-plus roles, eleven lifecycle phases, three account tiers, team-emotion overlay. One surface for leadership to argue from, sequence from, and invest against.
Talk to Ben about your work
If the surface isn't holding, the operating model usually isn't either.
That's the work. Naming what the org chart, the queue, and the form are quietly doing to the experience, and designing the thing that actually fixes it.