Weekly creative cycle on a luxury Meta account. One CD, no agency in between.
Aera makes a luxury home-fragrance system designed with master perfumers, in a category full of cheap candles and $30 diffusers. The in-house CD cared about brand. Two campaign moments a year wasn't enough to feed a paid program, and the existing ad library had been built for awareness, not for the weekly variant cycle a buyer needs to scale spend.
Ongoing creative direction retainer. All five hats on one account: designer, filmmaker, photographer, analyst, and the production system that ties them together. Weekly cycle against the ad data, no agency in between, every variant touched by the CD before it shipped to the buyer.
Performance creative is a search problem. You map the territory, narrow with data, and double down on what's moving the account. Aera ran the standard five-step loop with one constraint that shaped everything: every variant has to survive the brand book.
Audit. Read the existing ad account against the brand library. Most assets were campaign-built: beautiful, slow, designed for a launch moment. Almost nothing was built for the weekly test cycle. The in-house team had perfumer interviews, customer transcripts, and a year of unaired footage sitting unused. Loaded all of it into the system as raw material.
Theme map. Pulled eight candidate themes from the interview library and the category audit: fragrance education, what bad fragrance does to a home, why a luxury system is different, the science behind the swappable capsule, gifting, scent-as-ritual, sleep, and seasonal moments. Each became its own batch of variants against the brand book so the buyer could read angle against angle in the account.
Narrowing. Within four cycles, three themes earned the spend share: fragrance education, the swappable-capsule science, and the why-a-luxury-system frame. Sleep and gifting got cut from the always-on rotation and held for seasonal moments. Scent-as-ritual got folded into the wellness campaign as a brand-led companion to the performance lane. The Palo Santo concept inside the education theme ran away with spend. Built iterations around it: same scent, new video styles. ASMR, voiceover, unboxing. Same brand voice, new opener every cycle.
Asset gap-fill. Used AI on existing footage to refresh the library between shoots. New backgrounds against the same product truth, capsule reveals in scenes the original shoot didn't cover. The brand approved every generation. Saved a quarterly shoot day for the moments AI couldn't fake: real customer scent reactions, perfumer commentary, capsule-in-the-hand product truth.
Ship, read, iterate. Weekly cycle against the buyer's read of the account. One test from the early phase: 30% Off offer vs. a $50 Gift Card framed as the same dollar value. The 30% Off pulled 100%+ more purchases at the same spend. Both versions kept running against different audiences once the read came through. That cadence, run for a year, is what 500+ ads at +38% top-hook lift looks like under the hood.
Hook tests, re-cuts, and shoot-driven launches from the last twelve months.
A four-cut wellness set. Two voice-led sentence ads paired with two headline-support variants.
The retainer keeps renewing because the work keeps performing. 500+ ads shipped. The habit of reading the account weekly is what drives the numbers.
500+
Ads shipped
in 12 months
+38%
Top hook lift
vs. scripted control
Weekly
Cycle against
the ad account
1:1
Creative director
every cycle
I never had to check in week to week. I trusted you would create high-quality work that was on brand, with a lot of intention behind it. I didn't feel like I had to hold your hand.
Creative DirectorAera