We introduced the format Pilates Hunter never tried — and captured the audience they never reached.

Intro

We worked with Pilates Hunter— a multi-branch clinical and fitness pilates studio with four Jakarta locations and certified instructors — on a two-week Phase I performance-creative takeover on Meta. The credibility was real: real instructors, real results, a real brand. What the account had never done was build creative toconvert. It had run video-only ads to cold audiences, for every branch, every week, for as long as it existed. No static had ever been tested, and the most-aware prospects — someone already living with scoliosis or lower-back pain, already comparison-shopping — had nothing built for them.

We introduced a single static creative for that most-aware segment, aimed at all four branches at once. Result: it delivered the lowest CPL in the account — 38.3% below the best branch video at Week VI launch, produced +30 leads from one creative (more than any single-branch video), and proved two things the account had never tested — that static can beat video when the audience is primed, and thatone cross-branch creative can out-produce four branch-specific ones.

The three cuts

This page is the overview. The engagement comes apart into three deep-dives — each one discipline of the same two weeks, readable in any order: Performance Marketing, where “one video per branch, always cold” became an awareness-mapped media engine; Performance Creative, where a brand-inconsistent video library was rebuilt around the audience’s exact condition; and Data & Analytics, the measurement and diagnosis that turned a stuck account into one that can finally read itself.

Meet Pilates Hunter.

Pilates Hunter runs four pilates studios across Jakarta — Kelapa Gading, Menteng, BSD, and Kemang — specialising in condition-specific pilates for scoliosis, lower-back pain, HNP, muscle imbalance, knee and neck pain, alongside performance pilates for athletes and professionals. The intro offer is a low-cost 5-class trial, and every instructor is certified under Master Trainer Charles Xu, licensed by Pilates Academy International (PAI), New York. The credibility was never the problem. The problem was that none of the advertising was built to convert — the brand barely showed up in its own creative, every campaign spoke to cold audiences, and a video-only library meant that when performance hit a wall, the data to diagnose it had never been created.

The discipline we installed

Phase I wasn’t a pile of new ads. It was three tracks running in parallel, each with a hypothesis, each feeding the next. The static wasn’t added on instinct — it was the output of a structured thesis about who the existing library wasn’t serving. This discipline is the spine of the engagement; each cut below is one part of it applied in depth.

1. Thesis — map the awareness gap, build for the most-aware. — Existing creatives were audited by awareness stage. Cold-audience TOFU video was running fine for reach — but product-aware prospects who already knew they needed pilates, or had a specific condition, were unserved. A static was briefed for exactly that segment: clear offer, clear condition hook, direct conversion path. No storytelling overhead.
2. Test — introduce static against the video baseline. — The account had never run a static, so introducing one created the comparison the account had always lacked: same offer, different format. If static CPL beat video CPL, the lesson isn’t “video is wrong” — it’s that format is a variable to manage, not default. The static also changed distribution: one creative covering four branches at once, not four videos running in silos.
3. Kill or keep — cut what doesn’t compound, scale what does. — Every creative ran with explicit kill criteria. Branch videos that crossed CPL thresholds without qualified leads were flagged for creative overhaul, not just budget reallocation. The static survived both weeks. The branch videos that worked (Kemang, Kelapa Gading) were kept; the one that didn’t (Menteng · Cara Duduk) entered the overhaul queue. The library gets stronger each cycle.

The three tracks are designed to feed each other: the thesis tells you which audience to test next, the format test makes performance differences attributable rather than guessed, and the kill-or-keep discipline clears budget toward what compounds. Over time the library only carries hypotheses that have been validated — so every future test starts from a higher floor.

Where to go next — the three cuts.