Replacing random ad production with a hook-first creative factory.
Siam Sports Society is a curated subscription community for Bangkok's founders and creatives. Arctis Labs installed the creative testing discipline and scaling architecture it needs to grow paid acquisition without turning CAC into a guess.
- Active paying members
- ~56
- MRR
- ~AUD 2K
- Target LTV : CAC ratio
- 6 : 1
- Graduation-gate CPP benchmark
- ~AUD 45
A niche audience that punishes lazy creative.
SSS sells a curated membership — matched into small sports groups with people who share your energy, values, and ambition. The ICP is narrow on purpose: Bangkok-based founders, operators, and creatives, aged 25–45, willing to pay AUD 25/month for the coordination to be handled.
Narrow ICPs have a cost profile that rewards precision and punishes volume. Broad Meta scaling burns budget against the wrong audiences long before it reaches the right ones, and every weak ad drags down the account's ability to find signal. The creative work is the growth lever — and random creative production is the quiet way the unit economics fail.
Scaling acquisition in a category where most ads don't deserve spend.
The audit surfaced the structural problems that stop most subscription communities from profitably scaling paid — before the creative work even starts.
-
No separation between testing and scaling.
New ads launched directly into the main ad set. Winners and losers competed for the same budget. Fresh concepts got starved by proven ads while proven ads got diluted by fresh ones — the structure itself was working against both jobs.
-
Ads were products, not experiments.
Each ad was treated as a bespoke creative artifact. No shared body, no locked variables, no isolation of what was actually being tested. Which meant "this ad worked" was a statement about a specific creative — not a reusable lesson.
-
Hook decisions were taste-based.
In a feed where the first 1–3 seconds do almost all of the work, hooks were being written on instinct. No systematic coverage across identity, pain, status, curiosity, or outcome angles — which meant obvious winning angles were going unexplored while marginal ones were being recycled.
-
No graduation logic.
Without a written CPP benchmark and a promotion rule, every "this is working" call was political, not mechanical. Ads survived based on recency and preference, not on whether they had crossed a pre-committed performance line.
Two campaigns. One graduation rule. Modular creative.
The strategic bet: treat creative as a manufacturing process, not a craft. Every decision above — structure, template, hook coverage, promotion — was turned into a written rule so the account stops relying on judgment calls to find its winners.
Split testing and scaling campaigns
Scaling campaign runs the graduated winners at AUD 50/day on a broad audience, sales objective, purchase optimisation. Testing campaign runs fresh concepts at AUD 13/day against the same audience. Ads move one direction: up, through the graduation gate.
Write the graduation rule
An ad graduates to the scaling campaign when it hits a CPP of ~AUD 45. The rule is mechanical, not political. Nothing moves early on vibes; nothing survives late on loyalty.
Install the modular ad architecture
Five-block template — Hook / Pattern Confirmation / Core Mechanism / Proof Stack / CTA. Only the hook varies between tests. The body stays identical, so hook performance is isolated from every other variable.
Run a hook coverage matrix
Hooks batched across five angles — identity, pain, status, curiosity, outcome — with 5–10 variants per bucket. Pattern-break rules enforced: first frame must shock, micro-cut in the first second, subtitles always burned in.
Anchor the content engine to a value framework
Organic content planned against the four Hormozi levers — increase dream outcome, increase perceived likelihood, reduce time delay, reduce effort — with perceived likelihood identified as the strongest multiplier for this category.
Cadence the ad formats (ABACD)
Static / graphic explainer / testimonial / video. Every week the testing campaign refreshes the 3–5 concepts running against the same audience — so creative fatigue and audience fatigue never get confused with each other.
A process that compounds, not an ad that worked once.
The point of installing a creative testing system is that the deliverable isn't a winning ad — it's a repeatable mechanism for finding the next one, and the one after that. Early signal:
Baseline figures as of current quarter. The creative testing architecture is the deliverable; MRR is the score it's judged against over the 90-day read.
The 90-day read.
Creative testing is a system or it's an opinion.
Most brands producing paid ads are producing artifacts. Each ad is written, shot, and judged in isolation. When something works, nobody can say why. When something fails, everybody has a theory. The account drifts.
SSS is the case study for the Arctis creative-testing-as-a-system offer: a written structure, a written graduation rule, a locked template, and a matrix of hooks — so what we learn from one ad compounds into the next hundred. That's how niche-audience subscription businesses scale paid without their unit economics going with them.
Producing ads like artifacts instead of experiments?
If your Meta account is a pile of one-off creatives with no shared structure, no written graduation rule, and no way to tell why the winners won, that's the engagement we're built for.
Get in touch