Creative Generative AI Tools in Meta Ads Manager
Meta's smaller advertisers were disadvantaged by the cost of producing varied ad creative — I crafted the narrative that won leadership approval for Meta's generative AI creative vision in Ads Manager, then led content design for the first three tools that shipped: text generation, image expansion, and background generation.
Producing varied ad creative is expensive, putting smaller advertisers at a structural disadvantage. Closing that gap with generative AI meant convincing leadership to fund a multi-quarter bet — then getting advertisers, many skeptical of AI-generated content, to actually trust and use it.
Approach
I built the narrative that won VP approval and unlocked the MVP workstream, then led content design end-to-end across the three shipped features — naming, in-product education, trust-building UI, and the external narrative VPs used in press.
Winning the mandate to build was its own design problem — the same clarity and plain language that make in-product content trustworthy to a skeptical advertiser are what made the strategic narrative persuasive to leadership deciding whether to fund it.
Key decisions
- Named the features in plain language over jargon-forward alternatives some on the team preferred — critical for localization, since roughly half of Meta's global community is outside English.
- Treated advertiser distrust of AI output as a trust problem, not just an interface one: preview-before-approve tooltips, plus legal-partnered disclaimers in-flow.
- Delivered under a hard external deadline (Ad Week) while staying consistent with the existing Advantage+ design system.
Who I influenced
Outcome
- Casetify: 13% ROAS lift using Background Generation (TechCrunch, May 2024)
- Half of surveyed advertisers: 5+ hrs/week saved
- Leadership approved a dedicated MVP workstream for H1 2024
- Established a plain-language naming standard company-wide for the feature set