From Growth at Any Cost to Profitability
Meta is no longer scaling reach for you based on targeting rules. The algorithm now leans heavily on diverse creative inputs. When your assets look too similar, Reach suffers — which drives CAC up and puts sustainable growth at risk.
That’s what happened at one luxury lifestyle brand: performance was faltering, not because the media buying was wrong, but because creative diversity had stalled.
The Old Creative Rhythm
The existing process was familiar:
Every month, performance and creative teams reviewed results.
They flagged “winning ads” and “losing ads.”
They drew sweeping takeaways: “ASMR is good.” “Design-process stories don’t work.”
The flaw? These judgments were based only on that month’s surface-level results, not on structured learning across assets.
The Limits of Basic Tagging
Creative naming conventions (via UTM fields) made deeper analysis impossible:
Overlapping categories: two totally different assets carried identical naming values.
Non-MECE fields: an ad had to be tagged “design” or “quality,” when in reality it was both.
Underscore-based parsing: “the fifth underscore = ad format” fell apart when optional values (like motion type: stop motion vs. live video) had to be added.
This system forced narrow analysis and hid what was really driving performance.
Building a Creative Ontology
We designed a robust creative asset ontology — essentially a structured taxonomy that captured every relevant dimension of an ad, such as:
Strategy (prospecting vs. retargeting, branded vs. non-branded)
Format (UGC, animation, product demo, static)
Story (process narrative, lifestyle moment, testimonial)
Visual strategy (models, environments, product focus)
Claims & value props (design, durability, sustainability — with multi-select allowed)
Production details (studio, UGC, influencer)
Delivery details (motion type, placement, CTA)
To operationalize it, we built a UTM generator that enforced structure. Instead of underscores and free text, it created standardized fields while still allowing optional values.
From Asset Chaos to Learning Loops
Once creative assets were properly structured:
We could see that despite the team’s heavy output, true diversity in spend was lacking.
When we invested into diverse creative types (UGC, relatable lifestyle shots with models, authentic human moments), performance improved.
A new learning loop was born: instead of monthly “winner/loser” meetings, teams tracked which categories of creative drove reach and CAC shifts.
The Results
Reach reversed its decline after two quarters.
CAC pressure eased, stabilizing growth.
The company was able to make its next leap toward 10-digit revenue ($1B+).
References That Informed the Framework
While the academic field on creative strategy and measurement is limited, we drew from:
Creative Strategy: A Management Perspective (Springer)
FTC guidelines on advertising substantiation (16 CFR §14.15)
Case research on creative strategy and campaign effectiveness (Scispace, Springer, Kantar, Binet & Field).
The Takeaway
Meta no longer rewards targeting tricks — it rewards structured creative diversity. But you can’t optimize what you can’t define.
A creative ontology transforms asset chaos into structured learning, letting you see not just which ad won this week, but which creative strategies compound Reach and lower CAC over time.
