← Back

Research Operations

Research Ops — process, documentation, tooling, and socialisation

Role: Senior Designer

product-designux-researchdesign-leadership

Problem

Research at CloudBees was project-by-project — no shared process, no tooling, no taxonomy. Insights were siloed, duplicate work was common, and there was no institutional memory across a growing design team.

Solution

Built the first Research Operations practice: process documentation, Dovetail as the shared repository with a coordinated tagging taxonomy, and a Double Diamond-aligned process mapped into Jira and Confluence. The programme earned executive buy-in — a dedicated researcher was hired and a Data Insights team merged into design.

Double Diamond alignment

01 Process

The starting point was alignment — mapping the research lifecycle to the Design Council's Double Diamond, then replicating the model in both Jira and Confluence: linked, parallel, and accessible to any designer or PM in the organisation. Research sits in the first diamond: 'Research' is about building the right thing; 'Operations' is about doing it right.
Research Ops — Double Diamond process model
001

Confluence and Jira

02 Documentation

I broke the first diamond into detailed Confluence sections — best practices, method templates, and guidelines for each research phase. A parallel Jira Research Epic template gave teams four subtasks, one per phase, each linking to the documentation behind it. A colleague extended the structure into a detailed diagram that became the canonical internal reference.
Research Ops — Confluence structure, research practices and templates
002
Research Ops — documentation diagram
003
Research Ops — Jira Research Epic template
004
Research Ops — Jira task breakdown, four phases
005

Dovetail research repository

03 Tooling

After evaluating Confluence as a research repository and finding it insufficient, I identified Dovetail and made the case for it. Before rolling it out to the team, I designed a data taxonomy — a controlled tagging vocabulary so that research from multiple contributors could be sorted, filtered, and compared coherently. Without a shared taxonomy, tags spiral into noise. Three capabilities made the investment worthwhile: charting tag distributions, searching across all projects at once, and drilling from any insight back to the source interview transcript.
Research Ops — Dovetail research templates
006
Research Ops — tagging example in Dovetail
007
Research Ops — data taxonomy draft
008
Research Ops — global tag set, all users
009
Research Ops — Dovetail charts, tag visualisation
010
Research Ops — search, sort, and filter across all projects
011
Research Ops — pulling insights from tagged research
012

Socialisation and exec buy-in

04 Outcome

The rollout went to the design team first, then to product. The Dovetail demo — tagging and insights in particular — was an immediate sell to PMs. It eventually expanded from the design team to the wider product organisation. Executive buy-in followed: a dedicated researcher was hired, and the Data Insights team was imported into the Design group and merged with research.

Part of the CloudBees selected work