LegalCollaborator

Advance Template Customization

Advance Template Customization

Role

Product Designer

Timeline

May – November 2025

Company

Wolters Kluwer

Product category

Enterprise B2B SaaS, LegalTech Platform

Core team

DR
Lead Designer (myself)
UX Researcher
Engineer
Product Manager

Role

Product Designer

Timeline

May – November 2025

Company

Wolters Kluwer

Product category

Enterprise B2B SaaS, LegalTech Platform

Core team

DR
Lead Designer (myself)
UX Researcher
Engineer
Product Manager

The Problem

  • 20% improvement in collaboration efficiency (platform-level, this work contributed)

  • 15% increase in user satisfaction

  • Custom question components built from the design system to support template types the existing library couldn't

  • Shipped as part of the broader Legal Collaborator platform improvements

The Problem

One-size-fits-all questionnaires in a niche-driven field

  • Legal Collaborator runs on questionnaires. Every engagement starts with one, and the MVP shipped a fixed set of default templates. That standardization is exactly what makes side-by-side firm comparison possible, so it wasn't an accident. It was the point.

  • But clients in niche practice areas (insurance claims, class action, jurisdiction-specific work) needed to ask different questions than the defaults allowed. LC let them customize a questionnaire for a single engagement, but those changes vanished the moment the engagement closed. Every new matter meant rebuilding the same custom questions from scratch.

  • I led UX/UI design on Custom Templates: a system that lets clients save their modified questionnaires as reusable, company-owned templates. Turning a one-time workaround into a scalable self-service tool.

Why It Mattered Beyond UI

Cost avoidance for clients

  • Without self-service templates, clients either lived with generic questionnaires or asked Wolters Kluwer to custom-build one for them, a paid services engagement.

Competitive table stakes

  • Our closest competitor already allowed unlimited custom questions per client. We were behind on a feature enterprise legal buyers expected.

Platform Expansion

  • The same template engine could extend beyond RFP engagements into lightweight, everyday vendor-engagement workflows, turning a point feature into platform infrastructure.

It makes every other signal usable

  • Our own research showed clients select outside counsel on lowest price only about half the time. The rest is content, staffing, and expertise signals captured in the questionnaire. A usable template system means clients can interrogate those signals consistently across every engagement.

The Real Problem (Design Tension)

The MVP's entire comparison advantage depends on standardized questionnaires. Standardization is what lets the platform compare firms side by side and, later, lets Expert AI score them against consistent criteria. The moment clients save and reuse their own custom templates, that standardized data model is at risk of fragmenting. If everyone customizes freely, nothing is comparable, and the platform's core value quietly erodes.

So the design problem was bounded flexibility: how do you let a client build a template that feels genuinely theirs, while preserving enough structure that comparison and AI scoring don't break? Total freedom breaks the platform. Zero freedom doesn't solve the pain. The whole design lives in the boundary between those two.

Research Approach

I tested where the friction actually was, before building further

Before investing more engineering time, we ran internal concept testing to pressure-test two prototypes: Customizing & Saving Templates, and Managing Templates. Six internal SMEs across Product Strategy, Training, and Product Management. These were chosen specifically because they sit closest to client friction.

Each session was moderated. Participants narrated where they'd click and what they expected, before I revealed the actual interaction. I tested three things specifically: how easily users could tailor a questionnaire, how intuitive saving it as a reusable template felt, and how easy ongoing template management would be.

Key Findings

Template Selection

  • What worked: Defaulting to the last-used tab reduced repetitive navigation. Clear template count and usage limit built trust in the system.

  • What didn't: "Default" vs. "Custom" terminology read as unclear and concerning. No way to preview a template before committing. Users assumed there'd be permission controls. There weren't.

  • What users want: Access controls and roles so not "anyone" can spin up a template. Terminology that signals where a template came from, not just that it's "custom."

Building Custom Components from the Design System

The existing design system covered standard form inputs, but the question types this feature needed went beyond what the component library offered. To build the questionnaire builder, I designed custom question-type components that extended the system rather than breaking from it: [name the specific components you built — e.g. multi-select question type, conditional/branching fields, file-attachment question, protected/locked section pattern].

The constraint that made this a design-system problem, not just a UI problem: every new component had to stay consistent with the existing library's patterns and tokens so the builder felt native to LC, and it had to preserve the underlying data structure that keeps questionnaires comparable. Extending the system was how I kept flexibility and standardization from fighting each other at the component level.

Questionnaire Customization

What worked

The lock icon was an immediately understood signal for protected sections. "Reset Template" gave users a clear way back to the base version. The existing input-type library covered most expected use cases (the custom components above covered the rest).

Friction Points & Iteration

What didn't work

  • The slider icon for section actions wasn't read as interactive. Hover-to-reveal actions were missed entirely by several participants. Drag handles for reordering were in an unexpected position. Deleting a section gave no real sense that it was irreversible. The "Required" checkbox on questions was easy to miss. The save flow (radio buttons disappearing into a new dialog) broke users' mental model. "Done" as a CTA implied the task was finished when it actually opened the editor.

What users want

  • Access controls and roles for template creation. A more obvious icon for section-level actions, visible by default (not hidden behind hover). Drag handles repositioned to the left. Stronger visual cues for irreversible actions. Template preview before saving. A true multi-select question type and support for standard Microsoft file formats. Default answers for multiple-choice questions.

Resolved in the iterated concept

  • Renamed "Default / Custom" terminology to remove negative connotation. Replaced the slider icon with an always-visible action pattern. Repositioned drag handles. Added stronger visual treatment around irreversible actions. Clarified field labels. Expanded file-attachment support. Restructured the save flow. Changed "Done" to "Next."

Scoped for future release

  • Role-based access controls. Full template preview before saving. Default answers and multi-select question types. Template usage analytics (last used, frequency). Intuitive discard/reset path while editing.

Impact

Measurable outcomes

Custom Templates shipped as part of the broader Legal Collaborator platform improvements I led design on, driving a 20% improvement in collaboration efficiency and a 15% increase in user satisfaction.

Structural shift

Clients no longer needed Wolters Kluwer to build a template for them when the standard library didn't fit. That's direct cost avoidance for the client and reduced services load on our side. And it laid the groundwork for templates to extend beyond RFP engagements into everyday vendor-engagement workflows.

Reflection

The biggest lesson from this round wasn't a UI pattern. It was about where ambiguity hides. "Custom" sounds like a neutral, even positive word in a design system. In a room of legal and compliance-minded users, it read as a governance risk. I didn't catch that until I put it in front of the people who'd field the client escalation, not just the people who'd use the feature themselves.

That shaped how I scope concept testing since: I deliberately include the internal teams adjacent to client trust and risk, not just the primary end user, because they surface concerns the end user won't think to raise until it's already live. Looking ahead, the access-control and preview gaps aren't nice-to-haves. They're the next blocker to broader adoption.

Deliverables

Template Selection & Questionnaire Builder

Adding Questions, Saving Templates & Repository

New Template Flow