Case studies: proof I don't just talk a good game. A few names and numbers below are wearing a disguise for client confidentiality, the strategy behind them isn't.
south Read the StudiesTurning Lead Data Into a Creative Source of Truth
THE CHALLENGE:
Across a portfolio of brands and programs, creative and media decisions (headlines, ad targeting, and audience segmentation) were largely being made on instinct and brand-wide assumptions. In reality, the audience wasn’t one audience: motivations, education background, and readiness to convert varied significantly by region and by program, but the creative treated everyone the same.
THE APPROACH:
I led an in-depth analysis of a large lead dataset spanning multiple brands and programs, looking at factors like education level, geography, and age/graduation cohort to understand who actually converts, not just who inquires.
- Segmented leads by demographic and regional patterns to identify which audience traits actually predicted conversion, versus traits that just predicted volume.
- Translated the analysis into practical creative tools: audience personas, regional “motivator maps,” and messaging guidance (what to emphasize, what to avoid, tone by segment).
- Used the findings to prioritize hypothesis-driven CRO tests instead of testing at random.
- Fed the same insights to the media team to adjust targeting parameters by region and program.
RESULTS:
- A single headline change, informed by the audience research, lifted conversion roughly 8% and lead quality roughly 4–5%, pushing gross profit percentage up more than 10% on that test.
- Media targeting adjustments based on regional demographic differences improved the efficiency of ad spend across programs.
- Creative ideation grounded in real audience data more than doubled the team’s rate of producing a “winning” test variant.
- Now extending this into a structured hypothesis-testing framework intended as a standing source of truth across all programs and campaigns.
THE HIGHLIGHT:
The shift wasn’t just “more testing”; it was testing with a hypothesis instead of a guess. Once creative and media were both working off the same audience data, the win rate on tests changed, and the org stopped re-litigating the same assumptions campaign after campaign.
Cutting Organic Post Production Time 60% Without Losing the Human Voice
THE CHALLENGE:
Producing organic social content across multiple brand pages was almost entirely manual: a social media manager would scan platforms for ideas, research what was trending, write the copy, and brief or build the creative; from scratch, per post, per page. A single post could take up to an hour, which capped how much the team could realistically produce and left little room for anything beyond the next post.
THE APPROACH:
I built an internal post-ideation and drafting dashboard using workflow automation tools (n8n), an AI model (Claude), and social platform APIs, designed to speed up ideation and drafting without handing creative judgment over to the tool.
- Configured each brand page’s established tone and content focus directly into the system, so suggestions came out on-brand rather than generic.
- Had it continuously scan competitor pages, a curated set of relevant websites, and even trend sources like Reddit for timely ideas and formats worth adapting.
- Built it to surface multiple post concepts per page, help draft the copy, and generate AI image-prompt starting points for the creative — assistive, not autonomous.
- Added a comment-analysis layer that scans audience comments on posts and ads and surfaces the underlying themes people are reacting to, so the team knows what’s resonating before the next post goes out, not after.
RESULTS:
- Cut time spent writing and conceptualizing organic posts by at least 60%, freeing up the social team’s time for higher-value work.
- Social engagement increased following the shift to comment-informed, trend-aware content.
THE HIGHLIGHT:
The goal wasn’t to have AI write the final post, it was to remove the blank-page problem and the hour of scanning/researching that came before any writing happened. The social media manager still writes and makes the final call; the system just hands them a much better starting point, backed by what the audience is actually reacting to.
Starting Every Ad Test From What's Already Winning
THE CHALLENGE:
Paid social reporting and creative testing were disconnected: pulling performance data and identifying what was actually working across campaigns was manual and time-consuming, and new creative tests were often built from a fresh idea rather than from evidence of what had already proven effective.
THE APPROACH:
I built a dashboard, again using workflow automation (n8n), platform APIs, and an AI model (Claude), that pulls in paid social campaign data for a given time range and identifies which ads and creative themes are winning — then suggests the next round of creative, copy, and targeting tests. It can optimize its recommendations around whichever metric matters most for the goal at hand: CTR, CPC, total spend, or ROAS.
- Applied a design-thinking approach to paid social: study what’s already resonating before deciding what to test next, instead of testing blind.
- Automated the reporting and analysis work that used to eat into the time available for actual creative strategy.
- Used it to brainstorm new creative and targeting directions grounded in evidence from live campaigns, not just intuition.
RESULTS:
- New creative tests now start from patterns that are already working, rather than an untested idea — improving the odds of each test and reducing wasted spend on cold concepts.
- Increased the rate of “winning” creative tests across the team.
- Gave both the creative and media teams a shared, faster way to understand campaign performance and agree on what to test next.
Rebuilding Creative for Scale
THE CHALLENGE:
EducationDynamics ran more than 15 owned-and-operated brand properties, each with its own audience, positioning, and marketing mix. Creative output had scaled with the business, but process hadn’t — the team was producing a lot of work without a shared system for prioritizing it, governing brand consistency, or connecting creative decisions back to performance. Turnaround times were long, and it wasn’t always clear which creative choices were actually moving the numbers.
THE APPROACH:
I took over creative leadership for a 10+ person multidisciplinary team spanning art direction, design, copywriting, development, SEO, and UX. The mandate was twofold: sharpen brand positioning across a complex portfolio of properties, and rebuild how the team actually worked so quality and speed didn’t have to trade off against each other.
- Redefined brand architecture across 15+ properties so each retained its own identity while rolling up to consistent governance standards.
- Replaced ad-hoc request intake with agile creative workflows and clearer accountability — from brief to review to launch.
- Built a testing rhythm across design, copy, UX, and marketing strategy, so decisions were informed by results, not opinion. • Took direct ownership of agency and vendor relationships, holding outside partners to the same creative and performance bar as the in-house team.
- Pushed accessibility work (WCAG AA) across our digital platforms as a standing part of the process, not a one-off audit.
RESULTS:
- Profit targets exceeded by 45%
- ROAS up 20%, engagement up 15%
- CTR up 12%, CPA down 16%
- Project turnaround time down 40%, operating costs down 30%