Scaling a Health and Wellness Brand Through Systematic Creative Testing
How a disciplined creative testing engine kept EXALT's paid social growing through a year of disruption.
EXALT
Health & Wellness
12 Months
Paid Social, Performance Creative
EXALT
Health & Wellness
12 Months
Paid Social, Performance Creative
Challenge
EXALT operates in one of the most saturated corners of Meta advertising: health and wellness supplements, where rising CAC, fast creative fatigue and heavy competition make efficient new-customer acquisition genuinely hard.
Across the 12-month period the account also had to absorb two disruptions at once. Meta’s Andromeda algorithm update reshaped delivery and optimisation, and the account was migrated to a new ad environment mid-flight. On top of that, a soft patch in early autumn pushed in-platform ROAS below breakeven for two months.
The brief was clear: keep feeding Andromeda enough fresh, diverse, high-quality creative to sustain new-customer acquisition and revenue growth, without losing momentum through the disruption.
Solution
Our solution centred on a systematic creative engine built to keep pace with Andromeda while protecting efficiency through a year of change.
Systematic Creative Testing Framework
We built a disciplined testing engine that shipped a steady stream of new concepts every week, designed for a category with long consideration cycles. Rather than waiting on final purchase data, we read leading indicators (site visits, video watch time, first-frame retention, thumbstop rate and add-to-cart) to spot promising directions early and iterate fast. Testing spanned multiple content types, from UGC creator videos to statement graphics, collection ads and lifestyle imagery, each tagged and tracked through a custom naming convention and bespoke reporting for granular analysis across the funnel.
Creative That Won
The framework produced scaled winners across formats. The “Michael” UGC creator series became the account’s workhorse, driving £35.3k in revenue at 1.4x ROAS with 76% first-frame retention. Statement graphics such as “Pooping Every Day” (£17.5k revenue, 213 purchases) and the iterated “Weather App” series (up to 1.98x ROAS) proved that a strong, ownable angle could scale. After the migration, fresh concepts like the “Post-Its” and “GLP-1 Nutrition” graphics hit 2.1x to 2.3x ROAS out of the gate. Winning videos consistently posted 76% to 92% first-frame retention and 9% to 19% thumbstop, validating the early-signal approach.
Andromeda Adaptation
When Meta rolled out Andromeda, we restructured to work with the new machine-learning delivery rather than against it: simplified campaign architecture, consolidated learning and a prospecting-led budget, with roughly 95% of spend allocated to new-customer acquisition. Shopify attribution confirmed the role this played, 100% of Meta-driven orders came from new customers, while the returning base compounded through email and organic channels.
Navigating Disruption and Migration
We protected performance through the Andromeda transition and executed a clean migration to a new ad environment without losing acquisition momentum, carrying proven winners across and standing up fresh tests that scaled immediately.
Performance Monitoring and Optimisation
Continuous fatigue monitoring let us scale winners and sunset underperformers before decay hit results. After the autumn dip, this disciplined lifecycle management rebuilt efficiency from below breakeven to a 2.1x monthly peak.
Revenue
New Customer Acquisition
ROAS
More creative tested
Top Ads
Working with The Graygency transformed how we think about paid social. Their systematic approach to creative testing meant we were never guessing, we always knew what was working and why. When Meta rolled out Andromeda, they adapted our strategy almost immediately, and the results speak for themselves: 45% revenue growth and a 2.1x ROAS peak in what was honestly one of our most disruptive years.
