← Back to work
Financial Client
Beyond translation: building a bilingual content system
How I reframed a three-month translation assignment to address the gaps behind a fragmented Spanish-language experience.
A financial coaching program had a mature English experience, but its Spanish experience lacked shared content standards. Translations were handled by an operations team member without a glossary or translation guidelines. Coaches adapted English scripts during sessions. Product and marketing teams made content decisions separately.
I was hired to improve the translations. After auditing the experience and interviewing bilingual coaches, I proposed a broader approach: create the guidance teams needed to make those decisions consistently.
My role: Content design and localization strategy
Timeline: Three months
Timeline: Three months
Finding the problem behind the translations
I audited content across the web app, booking flow, emails, SMS, and coaching materials. I documented issues by product area and impact, interviewed partners across Design, Product, Growth, Localization, and Coaching, and attended sessions as a user to understand how the experience worked in practice.
The problems extended beyond individual phrases:
- Literal translations preserved English wording but lost its meaning or warmth.
- Spanish copy was constrained by interfaces designed for English.
- Terminology and tone varied across touchpoints.
- Coaches translated or corrected materials while delivering sessions.
- The landing page had not been translated.
Three of seven bilingual coaches had stopped conducting sessions in Spanish. Their feedback made the operational consequences of the content gaps visible.
Reconstructed from the case study • Summary, not original research notes
Reconstructed from the case study • Summary, not original research notes
Reconstructed from the case study • Summary, not original research notes
Choosing what to do with three months
I could spend the engagement correcting as many translations as possible, or invest part of that time in standards that would support future work.
I recommended building the shared foundation while prioritizing immediate fixes by impact and effort. I presented the audit, coach feedback, and proposed roadmap to my manager and Product partners and secured support for the change in scope.
The tradeoff was time: creating the guidance meant completing fewer individual translations during the engagement. It also gave teams a reference for decisions that would otherwise keep returning to the same people.
Different content needed different treatment
I organized the content into three areas based on its purpose:
Reconstructed from the case study • Summary, not original research notes
This structure helped explain why one translation approach could not serve the whole experience.
I built a shared Confluence reference covering voice and tone, content patterns, terminology, translation guidance, ownership, and AI prompts and use cases. The guidance explained the reasoning behind decisions so designers and translators could apply it beyond the examples provided.
Validating with the people using the content
I worked with coaches on tone and clarity, localization partners on linguistic choices, Growth on brand voice, and Design and Product on interface constraints and patterns.
For example, “Let’s paint your picture” had become “Pintemos tu dinero,” which lost the original metaphor. “Pintemos la imagen de tu dinero” made the intended idea more explicit.
| Version | Wording |
|---|---|
| English original | Let’s paint your picture |
| Previous Spanish | Pintemos tu dinero |
| Revised Spanish | Pintemos la imagen de tu dinero |
These examples connected the framework to actual writing decisions.
What I delivered
In three months, I delivered a bilingual content framework reviewed with designers, coaches, and translators; a prioritized roadmap of content fixes; and a content wiki and Jira board to support ongoing work.
The project established shared guidance where teams had previously relied on individual judgment. Its longer-term effects on translation quality, coaching participation, and delivery time were not measured during the engagement.
The central lesson was that improving a bilingual experience requires attention to how content decisions are made, shared, and maintained—not only to the translated words.