All work
Design OperationsEnterpriseSherwin-Williams · 2024

Building Design Operations at Scale

Design operations for one of America's largest paint brands. I rebuilt the shared infrastructure that a research and design team of more than 20 relied on, across six digital products.

Role Design Operations Lead
Client Sherwin-Williams (via TCS)
Industry Retail / Consumer Products
Timeline Dec 2023-Aug 2024
Design Operations · Sherwin-Williams · 2024
01

File Architecture

Scattered files across OneDrive and Teams, rebuilt into one numbered, navigable system the whole team could use.

3 1 file versions
02

Research Templates

Shared templates for interviews, insight decks, and stakeholder presentations, no more starting from scratch.

×3 formats standardised
03

Handoff Process

A five-step handoff so research findings reached design as a starting point, not an afterthought.

5 step process
30-40% less rework
20+ team members
6 digital products
NDA

All client work is under NDA. The file-architecture and process diagrams here are reconstructed from memory. The structure and the thinking are accurate; specific file names and content are illustrative.

Context

One of America's largest paint brands, and a complex digital portfolio

Sherwin-Williams runs one of the largest digital portfolios in the paint industry: consumer colour tools, contractor apps, e-commerce, and internal coatings systems. The research and design team supporting all of it had grown faster than its infrastructure.

5-7
UX researchers
10-15
Product designers
6+
Digital products supported

Products the team supported:

ColorSnap Visualizer

Web and mobile colour tool for homeowners

Color Expert App

AI-powered paint recommendation tool

PRO+ App

Contractor toolkit for professional painters

Paint Stores e-commerce

Main consumer shopping platform

Contractor & dealer portals

B2B account management tools

Specification & coatings tools

Internal, for the Performance Coatings Group

The problem

A team of 20+ with almost no shared infrastructure

Files existed, but nobody could find them

Research and design assets lived in personal OneDrive folders and scattered Teams channels. Folders existed, but were inconsistent, unlabelled, and constantly duplicated.

Duplication was the default

Designers and researchers regularly redid work that already existed, only to find out afterwards that the asset was there all along. The rework was the symptom. The broken structure was the cause.

Decks were made and abandoned

Research presentations had no consistent format and no connected home. Once delivered to a stakeholder, they disappeared into personal folders and were never referenced again.

Research rarely reached design

There was no handoff process. Insights sat in researcher files. Designers started work without access to what had already been learned.

Approach

Before changing anything, understand the real pain points

I interviewed researchers and designers separately before proposing any solution. The order mattered: diagnosis first, then structure, then iteration with the team in the loop.

01

Interview both sides

Separate sessions with researchers and designers to surface real friction without cross-contamination or bias. As someone who had worked both sides, I knew the questions to ask, and I made sure to speak to more than one person in each group so no single perspective skewed the findings.

02

Map the pain, not just the symptoms

The duplication and rework were symptoms. The real problem was inaccessible structure and no shared language between disciplines. Naming that distinction changed what we chose to fix.

03

Pressure-test with the Design Ops Manager

Brought findings back to the manager, challenged assumptions, and agreed on what to prioritise: research infrastructure first, then the designer side. Sequencing the work mattered as much as the work itself.

04

Fix iteratively, get feedback fast

Each improvement was tested with the team immediately. If it did not land, it was revised. The 30-40% efficiency gain came from accumulated small wins, each validated before moving on.

What I built · File architecture

From scattered to structured

The first fix was the one everything else depended on: a single, navigable file architecture in Teams and OneDrive. Numbered folders, consistent naming, a read-only archive for legacy work, and cross-linked documents so a file could be found from more than one path.

Before
▸ Sarah_Research_FINAL▸ Sarah_Research_FINAL_v2▸ James_Insights_March▸ Team Files (old)▸ Team Files BACKUP▸ ColorSnap_designs▸ ColorSnap_designs_NEW▸ Misc▸ Deck (no date, no owner)▸ Interview Guide copy▸ Interview Guide REVISED
After
▸ 00_Templates [linked] └ Interview Guide v1.0 └ Insight Deck Template▸ 01_Research [by product] └ ColorSnap / PRO+ / Portals▸ 02_Synthesis [by quarter] └ Q1 2024 Findings▸ 03_Stakeholder Decks └ YYYY-MM_Product_Topic▸ 04_Archive [read-only] └ Pre-2024 files preserved
What I built · Research templates

Standardised templates across three formats

Interview Guide

  • Standardised opening and consent framing
  • Modular question blocks by research goal
  • Probe suggestions for each section
  • Consistent time allocation per block
  • Debrief and follow-up prompts

Insight Deck

  • SW-branded slide template
  • Executive summary on slide 1
  • One insight per slide, evidence-backed
  • Severity and frequency scoring
  • Recommended next actions at close

Stakeholder Presentation

  • 3-slide maximum for leadership reviews
  • Business impact framing first
  • Data visualised, not listed
  • A clear ask or decision required
  • Appendix for anyone wanting detail
Before

Each researcher had their own format. Senior stakeholders received decks of 30+ slides with no summary, no structure, and no clear ask.

What I built · The handoff

Connecting research to design

The most valuable infrastructure was not a folder or a template. It was a process: a five-step handoff that made research insights part of the design workflow rather than a parallel activity.

01

Research conducted

Interviews, usability tests, and surveys. The raw material, run by the research team.

02

Synthesis & insight deck

Findings written up in the standardised insight format, linked to source files.

03

Handoff session

The researcher walks designers through the findings directly, not a document tossed over a wall.

04

Design takes over

Insights are referenced in Figma annotations, so the rationale travels with the design.

05

Design critique

A structured crit checks design decisions back against the research that informed them.

Before

Research findings were delivered to stakeholders and then effectively lost. Designers had no consistent access to what had been learned, so they started from scratch, sometimes contradicting decisions research had already informed.

Impact

The team noticed the difference immediately

30-40%
Reduction in rework and duplicated effort, the team's own estimate
3 → 1
File versions to manage, collapsed from duplicates to a single source of truth
20+
Researchers and designers working from one shared system

Each improvement was validated with the team before moving to the next. The feedback was immediate and specific, because people could see what changed and why it mattered.

Reflection

What I finished, what remained, and what I learned

Completed

  • Full file architecture rebuilt in Teams and OneDrive
  • Research templates standardised: interview, insight, stakeholder
  • Researcher-to-designer handoff process established
  • Design critique framework introduced
  • Cross-linking of documents for easy navigation
  • Research infrastructure working consistently across the team

In pipeline · next phase

  • Designer-side file infrastructure: Figma library governance
  • Component ownership and version-control process
  • Cross-product design consistency review
  • Onboarding documentation for new team members

Design ops problems are rarely about tools or templates. They are about trust and habit. The file architecture was broken partly because nobody had taken responsibility for it, and creating structure also meant creating ownership. That took as much facilitation as it did design.

What this engagement taught me