I’m a creative strategist working at the intersection of culture, audience, and data. I take fragmented inputs - social listening, search, trend signals, syndicated and first-party data - and turn them into a point of view that marketing, creative, product, and executive teams can act on.
Gap’s seasonal brand planning ran on inputs that each told a partial story: social listening, search, audience profiling, market intelligence. I synthesized them into Need States, four plain statements of what consumer messaging had to address that season. Marketers began pulling the materials for their own teams and agencies without being asked, and it rolled out to Gap’s international markets, with a second-brand test underway.
Seasonal strategy depended on many signal sources that never resolved into one point of view. Each report described what the audience was doing. None of them said what the work had to do.
Brand marketers and their creative and media agencies didn’t need another data deck. They needed a brief: a shared, decision-ready read on what messaging had to address for the customer that season.
Not “what is the audience doing?” but “what does our messaging need to address?”
Each signal set is partial on its own. Stacked and read against each other, they converge on a small number of underlying needs: the states a season’s messaging has to speak to.
Also used to advise the CMO and marketing leadership on seasonal brand strategy: one point of view built from social listening, EMV, search, audience profiling, and market intelligence.
I saw it early: publishers would need their own earned-media story to keep pace with influencer marketing, and Condé Nast had no credible, repeatable way to quantify the earned media its brands and events generate. I ran the build end to end: vendor RFP and evaluation, methodology and best practices, the executive recommendation, productization, and integration into go-to-market. The capability went on to influence $90M+ in tentpole and event revenue.
Earned Media Value existed as one-off reports, not a standard. No shared methodology, no repeatable output, no agreed read on what the number meant.
To support revenue conversations it had to be consistent, credible across brands and events, and legible to people who don’t work in analytics: sales, editorial, the C-suite.
How do you build a measurement standard when the market doesn’t hand you one?
Embedded in go-to-market for major moments including Vogue World and the Met Gala, and in RFP responses to prospective advertisers.
YouTube ad-exclusion lists (the channels a campaign refuses to run on) were built by hand, one channel at a time, in reviews that took up to 60 hours. I prototyped a workflow that queried Mobian’s enriched database directly and generated the list in about three minutes. I built the first version solo, then hardened it with an engineer once it needed to scale.
The manual process: an analyst reviewed channels and videos one by one, then compiled the exclusion list. Up to 60 hours per campaign, and the output was a flat list with no read on why a placement was wrong. It was already AI-supported, but off best practice, with no pipeline behind it.
“Brand-safe” wasn’t the same as “brand-fit.” In one campaign analysis, most spend landed in low-risk, positive content that still missed the target audience: a large share went to children’s content, and almost none to product-relevant environments.
Safe inventory can still be the wrong inventory.
The workflow made it easy to move budget off misaligned placements and toward environments that matched the brand and audience.
Exclusion-list entries, after the prototype was hardened with an engineer and scaled as a product.
Built solo as a prototype, then refined for production with a dedicated engineer once it had internal buy-in.
The GTM team kicked off reporting and deck creation by email, then waited about five days while it was built by hand. Working with Convey, I translated that manual process into an autonomous agent: it takes the email trigger, runs the analysis, and returns a client-ready deck in roughly 15 minutes. The work was defining what the agent should do, in what order, and to what standard.
Reporting was manual and email-triggered. Pulling the data, building the analysis, assembling the deck and reviewing it ran to about five days each cycle.
Strategic account work couldn’t scale while people spent days hand-assembling mechanical outputs. The constraint wasn’t the thinking, it was the production.
The bottleneck wasn’t the analysis. It was the assembly.
Automating the workflow meant decomposing it into stages an agent could run reliably, and deciding what “good” looked like at each one.
Built with Convey. Zoe’s role was translating the business process into an automatable workflow and defining how the agent should execute it.
Anthropologie Weddings and Vogue polled more than 2,000 engaged and recently married brides to understand how style fits the wedding-planning journey. I led the analysis and advised editorial and client stakeholders on which findings mattered, the ones that could carry content. The survey shaped a three-part Vogue.com series, and the wardrobe story became one of Vogue Weddings’ top-performing articles of the period.
A rich custom dataset (2,000+ respondents drawn from Vogue insiders, Anthropologie and fellow URBN customers) and a brand partner, but no content strategy yet.
The survey spanned the full bridal journey. Editorial needed a point of view: which findings were interesting enough to build stories around, and what was the through-line that held them together?
A dataset isn’t a story until someone decides what it’s about.
The published article followed the sequence the analysis surfaced: each event on the bridal calendar as its own styling moment, in order.
Research to insight to content strategy to a brand-partnership story. Zoe led the analysis and advised editorial and client stakeholders on what mattered most.