Adobe
Generative AI arrived in the creative industry as both the biggest product story in a generation and the most confused conversation in it. At Adobe I owned the Adobe Express Learn blog — its content strategy, SEO architecture, editorial calendar, and contributor roster — through the period when Firefly went from announcement to shipped product.
That meant writing and editing at volume, but the harder work was structural: figuring out what a small business owner, a solopreneur, a classroom teacher, and a marketing lead each needed to understand before any of them would trust a generative tool with real work. Audience research, segmentation, and persona development drove the editorial plan. The plan fed the product sprints, and the sprints fed back into the content.
Principal duties + roles :
Content Strategy • SEO Strategy + SEO Audits • Editorial Development + Management
Editorial Calendar + Contributor Management • Copywriting + Copy Editing
Audience Research + Segmentation • Persona Development
Applied Generative AI • Product + Market Positioning • Product Development Sprints
CMS Management • Creative Strategy • Project Management
The Adobe Express Learn Blog
Building the on-ramp to Generation AI.
When Firefly launched, the gap wasn’t interest — it was comprehension. People wanted to use generative AI and didn’t know what it was, what it was trained on, whether they could legally sell what it made, or how to write a prompt that returned anything usable. Content had to close that gap fast, at scale, across audiences with very different starting points.
I built the AI cluster around a central hub, The Ultimate Guide to AI Art, which organized the whole body of work and routed readers to the piece that matched their need: prompt-writing for the curious, Generative Fill walkthroughs for brand owners, a buyer’s guide for small businesses comparing tools.
Underneath it sat a 59-term generative AI glossary covering not just the mechanics — prompts, diffusion, GANs, inference — but the questions that actually governed adoption: training data provenance, commercial indemnification, bias, the Content Authenticity Initiative and C2PA. Trust was the real conversion barrier, so the glossary treated it as content strategy rather than legal boilerplate.
-
Search was the distribution model, so SEO was an architecture decision rather than an editing pass. Every piece had to answer a question someone was actually typing, and sit in a structure that made sense to a crawler and a reader at once. Keyword targeting, heading hierarchy and internal linking were set at the brief stage, before anyone started writing. SEO audits ran against the existing library in parallel: finding what had decayed, what was competing with itself for the same term, and what deserved a rewrite rather than another new post.
-
A blog publishing at this pace fails one of two ways: the schedule slips, or the work gets thin. Structure is the defense. The editorial calendar mapped what shipped when and why, so no slot got filled simply because it was empty. A contributor roster carried the volume, working from briefs that fixed the audience, the search intent, the structure and the voice before a draft existed — which is most of what separates a usable piece from a rewrite. I copy-edited to one standard across every contributor, so a reader moving between articles couldn’t tell how many hands the library had passed through.
-
The subject range ran wider than any one person’s specialism: generative AI one week, restaurant menu design the next, then business letters, data visualization, Pinterest strategy. Authority on a how-to piece has to be earned per subject, so each one started by finding what the credible sources actually said — institutional research where it existed, platform documentation, practitioner consensus — and the writing’s job was to make that usable for someone running a small business with no time to read the source material. The 59-term generative AI glossary is the clearest case: it covered the mechanics, but also provenance, commercial indemnification, bias and C2PA, because trust was the real barrier to adoption and almost nobody was explaining it plainly.
Adobe Express Learn Blog Content
Click the images below to view full blog posts.