GenAI Prompt for
UX Writing
Industry
ECommerce
Company
Lycle
Year
2025
about.
As Lycle scaled, product language drifted across screens, reducing clarity and trust in critical flows like checkout, cancellations, and error states. I led the UX Writing Guidelines initiative and packaged it into a generative AI prompt, enabling the team to produce consistent, guideline compliant copy faster with less review overhead.
why.
While working on funnel improvements, I noticed the same language issues repeating across screens. The intent stayed the same, but wording changed depending on who last touched the UI. I shared a few examples and asked whether we were comfortable letting the product voice drift as we scale.
This started with a conversation in our Design chapter Slack
how.
Full copy visibility through a product-wide audit
We couldn’t standardize what we couldn’t see. Copy lived across shared components, hard-coded flows, dynamic states, and edge cases. I partnered with engineering to extract every user-facing string into a reviewable dataset, grouped by user journey with variants and conditions included when meaning changed.
This started with a conversation in our Design chapter Slack
Pattern analysis and taxonomy
I reviewed the dataset like UI, focusing on friction patterns rather than grammar. I created a tagging system to capture recurring issues such as tone inconsistency, missing next steps in errors, terminology drift, inconsistent button patterns, and trust risks in policy-sensitive flows.
This started with a conversation in our Design chapter Slack
Alignment and decision making
I facilitated a working session with Design, PM, and Engineering to align on what Lycle should sound like, which terms must be standardized, and where strict templates were required. This shifted the work from screen-level edits to building a shared language system.
This started with a conversation in our Design chapter Slack
ux writing guideline system.
I built a compact, operational guideline that teams could apply quickly. It included a voice framework (values, persona, tone), three principles (Kind, Clear, Trustworthy), scenario patterns for high-frequency copy, strict templates for errors and policy messaging, formatting rules, and a terminology glossary.
This started with a conversation in our Design chapter Slack
This started with a conversation in our Design chapter Slack
This started with a conversation in our Design chapter Slack
principle.
I built a compact, operational guideline that teams could apply quickly. It included a voice framework (values, persona, tone), three principles (Kind, Clear, Trustworthy), scenario patterns for high-frequency copy, strict templates for errors and policy messaging, formatting rules, and a terminology glossary.
This started with a conversation in our Design chapter Slack
final look.
I built a compact, operational guideline that teams could apply quickly. It included a voice framework (values, persona, tone), three principles (Kind, Clear, Trustworthy), scenario patterns for high-frequency copy, strict templates for errors and policy messaging, formatting rules, and a terminology glossary.
This started with a conversation in our Design chapter Slack
This started with a conversation in our Design chapter Slack
This started with a conversation in our Design chapter Slack
Key learning.
UX Writing is product infrastructure. Pairing standards with a prompt based workflow made adoption easier than avoidance, turning the guideline from a document into a living system.
