Embedding content strategy
Approach & early-stage strategy
As the sole content designer on the project, I embedded content strategy from kickoff through launch. I shaped product vision, naming, onboarding, instructional copy, and tone frameworks while collaborating daily with PMs, engineers, designers, researchers, legal, and data scientists.
Approach
1. Content at the table from day one
-
Defined PRD and co-led a 2-week AI design sprint.
-
Collaborated with research, data science, and engineering to align on goals and user problems (JTBD, HMWs).
-
Drove early decision: AI should live inside Stories, Reels, and Composer — not a separate hub — saving weeks of rework and ensuring adoption.
2. Training the LLM
-
Partnered with data science to author thousands of prompts and evaluate AI responses.
-
Established rules and patterns that shaped how Meta AI interacted with users — essentially teaching the model how to “sound” in product.
3. Designing tone and voice
-
Initiated a cross-app tonal framework for generative AI, uniting content designers across Meta.
-
Ran sentiment testing with research to validate tone directions.
-
Created a tonal matrix that became a foundation for all Meta AI creation tools.
4. Building usable, inspiring interactions
-
Named features with descriptive labels like “Help Me Write” (instead of “AI Writer”) to reduce friction.
-
Designed prompt pills strategy (Inspirational, Funny, Professional, Fix grammar, etc.) and recommended dynamic ranking by user data.
-
Shaped onboarding with “show, don’t tell” microcopy (e.g. Imagine a corgi as a balloon) and progressive disclosure.
-
Partnered with legal early, preventing costly content rework.
5. Driving alignment and buy-in
-
Framed concepts around user problems and business goals.
-
Presented to design leads and stakeholders to secure alignment.
-
Collaborated daily with product designers in Figma, iterating flows, error states, empty states, and tooltips.
