Case study: Kestrel Microcopy Style Guide
Scaling content consistency across 70+ product designers after UX writing team layoff
Overview
70+ product designers creating inconsistent microcopy across unified data platform after 23 UX writers were laid off. Solo content designer tasked with making microcopy decisions for new UI, documenting content patterns, and training designers to B+ competency while maintaining consistency. During this process
Role: Senior Content Designer
Timeline: 9 months
Tools: Miro, Figma, Confluence
Core team: Myself
Project Context
After layoffs eliminated a 23-person UX writing team, 70+ product designers were creating inconsistent microcopy across CrowdStrike's new unified data platform without content expertise, causing cross-functional confusion and operational chaos
Research
The Vice President (VP) of Design noted inconsistencies in Figma prototypes. Instead of manually auditing all Figma files (too time-consuming), I crowdsourced audits by asking designers to provide examples in a Miro template when working on specific patterns or components. I built a MIRO template for consistent review with use cases.
Insight 1
Different microcopy patterns and formatting inconsistencies (past and present tense verbs, noun + verb or verb + noun)
Insight 2
Terminology conflicts (Edit vs Update)
Content Design Process
Discovery & Collaboration
Work sessions with Security Strategists and Practice Advisors informed my content redesign ideas. They walked me through the steps where they saw their customers struggling. After these sessions, I went through the wizard to create each goal and SLA a few times to familiarize myself with the existing flow.
During this exercise, I realized that asking users to pick a goal type at the start added unnecessary complexity, since they needed to know what they wanted to make in the first place.
Wireframing & Iteration
I started wireframing in Miro to wireframe and share some new content design ideas with the UX designer, product manager, advisors, and strategists. I made several rounds of iterations after getting some feedback:
Attempt 1:
Added a new goal to Step 1 and rewrote the querying copy in question format
Attempt 3:
Rewrote copy based on the type of data the user chose
Progressive Disclosure
Break tasks into smaller chunks to avoid overwhelming users
Attempt 2:
Asked Engineering if we could display content based on user selections—they said yes
Attempt 4:
Combined Steps 2 and 3, since content was relevant to the selected content type
After getting approval from other disciplines, I shared my final recommendation with the PM. We didn't have time to test it, but given the feature's dismal performance, anything was an improvement.
Content Design Strategy
Based on research insights, I proposed redesigning content around three key principles:
Solution: Step-by-Step Content Redesign
Step 1: Goal Selection
Changes made:
Added preset templates of popular use cases like Remediate All Critical Vulnerabilities and Ensure Assets Have Ownership Tag for easy customization
Included a New Goal option that didn't require users to know the goal type upfront
Removed the barrier of having to understand technical goal types before starting
Background
Company/Product: CrowdStrike/Kestrel, a new unified data platform
Timeline: ASAP (urgent business need)
Team: Solo content designer who made microcopy decisions and trained 70+ product designers after a team of 23 UX writers were laid offs
Scope: Platform-wide microcopy standardization across all product areas
Challenge
Insight 3
Ambiguous or multiple terms that describe the same concept
Designer Pain Points
Couldn't distinguish when to use component variations
No clear approach for solving content problems
Concerns about maintaining consistent terminology across product UI
Solo Content Designer Pain Points
Product designers constantly Slacking with questions]
Needed systematic way to prioritize work - everything seemed important
Required context first before helping designers effectively
Match Mental Models
Format tasks in questions to help user think about tasks at hand
Contextual Guidance
Provide help when and where users need it
Before: The original interface assumed users knew which goal type they wanted to build and used jargon
After: The new design offers a "NEW GOAL" option, plus customizable preset templates based on common use cases
Step 2: Configuration
Changes made:
Allowed users to select either Remediate Vulnerabilities or Configure Assets, which refreshes the wizard to display the related vulnerabilities or assets flow
Decoupled query building into separate steps
Rewrote content as questions to help users think about the task at hand
Added contextual explanations ("You'll load or build a query to narrow down your assets") and links to help documentation
Before, Step 2: Original flow used vague terminology like "Load Query" and "Add Criteria" without context
Before, Step 3: Initially, users needed to fill in the percentage field before building a query, without any guidance on what to do
After, Step 2: After users chose vulnerabilities or assets, the relevant content and query prompts were combined into a single step
Step 3: Review
Changes made:
Told users the type of goal they created and defined it
Required goal naming to help with future goal management
Before: The fourth step had basic instructions and optional form fields that users could chose to fill out
After: The redesign clearly explained what type of goal users had created and provided a definition. Users also had to give their goal a name.
Results
Quantitative Impact
95%
completion rate (up from 65%)
Qualitative Impact
30%
reduction in support tickets
“THANK YOU, THANK YOU, THANK YOU! This is so much easier to use than the first version.”
“This is a huge improvement. My customers won’t stop raving about it!”
75%
faster creation time
800+
goals created in Week 1