Stage 0
Observe
Data ingest, baselining, market contextWhat it does
Collects data, surfaces insights.
How it advances
Becomes Recommend once 2–4 weeks of clean data lands.
Autonomous Growth Engine
We optimise bookings, not open rates.
Loop's AI story is specific and defensible: outcome optimisation, root-cause reasoning, and a per-studio learning loop measured against a 7-day attribution window — never a vanity metric.
From detection to outcome — every week.
One closed loop. The engine doesn't just send things; it knows whether what it sent moved the business.
- 1Detect
Spots drop in attendance, rising at-risk cohort, slow class, lapsed regulars.
- 2Explain
Identifies which cohort is leaking and why — class mix, instructor change, weather, package expiry.
- 3Recommend / Execute
Picks the right lever — email, reward, journey, challenge — for the right member at the right time.
- 4Measure
Attributes outcome over a 7-day window. Counts bookings, not clicks.
- 5Learn
Promotes winning variants. Retires losers. Gets sharper for your studio every week.
Honest about what AI should do on day one.
Loop earns autonomy. Every studio sees a stage indicator and exactly what unlocks next.
Stage 1
Recommend
Root cause + suggestion + confidenceWhat it does
Suggests actions. You approve, edit, or skip.
How it advances
Earns Auto-optimise on classes of action that beat baseline three weeks running.
Stage 2
Auto-optimise
Safe, reversible, narrow scopeWhat it does
Runs send-time, frequency, variant and segment tests on its own. Escalates the rest.
How it advances
Earns Operate on plays where attribution is consistently positive.
Stage 3
Operate
End-to-end growth playsWhat it does
Promotes winners, retires losers, reallocates effort. Strategic calls escalate.
How it advances
Stays here. You can revert any change, any time.
Not another "AI-powered" marketing tool.
The difference — line by line.
| Generic "AI marketing" | Loop | |
|---|---|---|
| What it optimises | Open rate, click rate | Bookings, attendance, retention, referrals, revenue |
| How it decides | Static rules · 'if X then send Y' | Root cause + confidence + 7-day outcome attribution |
| What it learns | Generic playbook | Your studio · your members · your market |
| What you see | Dashboards. Figure it out yourself. | Two screens, two questions |
| When it's wrong | You find out next quarter | One click to revert. Logged. Explained. |
Safety & control
Safe. Reversible. Explainable.
Every autonomous action carries a confidence score and a plain-English root cause. A live "Recently Optimised" feed shows exactly what changed, why, and what it produced — and one click rewinds it.
- Confidence score on every actionSo you know how sure Loop is before it acts.
- Root cause in plain EnglishNo "the model said so." Always the why.
- One-click revertOn any autonomous change. Original state preserved.
- Hard guardrailsOn send frequency, suppression and consent.
- Audit trailEvery change is logged, attributed and queryable.
- Stage gatingHigher autonomy only on plays Loop has earned for your studio.
Your studio's growth operator. Never sleeps. Never guesses.
See exactly what Loop's engine would have done for your studio last week — root cause, recommendation, expected outcome.