Stated goal (verbatim)

"run discovery brain to understand the app Ladder - it's our competitor. you would need to find screenshots of Ladder, understand it's core mechanics and how it works in general as business and as product."

Refined goal

Produce a decision-grade teardown of LADDER Strength Training Plans (Ladder Technologies, Inc.) covering (a) what the product actually is screen-by-screen, (b) the mechanics that make it work, and (c) the business behind it.

This is a teardown, not a strategy doc. It describes Ladder. The "so what" lives in the Consolidated Brief's implications section, bounded by what the Findings prove.

Consumer of this Inquiry — REVISED 2026-08-16T23:05

The original framing assumed Ilia was building his own fitness app and needed competitive intelligence. That is wrong. Ilia clarified:

"not planning on my app now, it's gonna be a separate step." "it's gonna be a case study for hiring project. i want to give them a prototype of an app how i would if i join them bult ladder competitor"

So the real deliverable is a hiring case study: Ilia is a candidate, and he intends to arrive with a working prototype of a Ladder competitor demonstrating how he would approach the problem if he joined that company.

What this changes

Consequences for this Inquiry

1. The teardown became more important, not less. For a product case study, the quality of the competitive analysis is a large part of what is being graded. The Findings are now a primary deliverable, not background research.
2. A6 business-market gains weight. A candidate who can size the market, read the funding history, and reason about unit economics is demonstrating exactly the judgement being hired for. Do not trim it.
3. A4 content-coach-supply becomes the strategic crux. The honest finding — that Ladder's moat is an expensive human content treadmill — is the most interesting thing a candidate can put in front of a hiring panel, because it frames the build-vs-partner decision any competitor must make.
4. Synthesis must produce a defensible point of view, not just a description. A teardown with no thesis reads as homework; a teardown with a thesis reads as a product leader.
5. Prototype fidelity is a design decision, not an engineering one. The Consolidated Brief should recommend the cheapest artifact that demonstrates the judgement — which is almost certainly not a native iOS app.

Still open

Which company is hiring, and what they already have, is unknown as of this revision. It does not block any Scout (Ladder is Ladder regardless) but it materially shapes the handoff. Tracked in STATE.md.

Subject identification (settled, primary source)

Ilia's own field notes (verbatim, RU — the seed hypotheses)

These are first-party observations from actually using the app. Each Scout must treat the relevant note as a hypothesis to confirm, refine, or refute against evidence.

1. удобная навигация для трекинга (можно однако лучше голосом) → A2
2. Голосовое сопровождение → A2
3. У них триал - это прям триал, они подогревают к оплате, у них типа тестовая неделя со своей отдельной программой → A3
4. Сама подборка упражнений, нет базовых (точнее мало, они смиксованы) → A2 / A4
5. Тренера классно онбордят в приложении → A3 / A4
6. Есть кастомизация в рамках онбординга → A3
7. Есть геймификация → A5
8. Новые тренировки. Они на контенте еще едут → A4

In scope

• Screen-by-screen product surface, IA, navigation model, design system.
• Core training loop: programming, workout anatomy, in-workout UX, tracking, progression.
• Audio/voice coaching model.
• Onboarding, customisation, trial design, paywall, pricing, conversion mechanics.
• Content and coach supply model; release cadence; the 2026 nutrition expansion.
• Retention machinery: Teams, Chat, Volts, Rewards, streaks, social proof.
• Business: company, funding, scale, revenue estimates, competitive positioning.
• Apple Watch companion and third-party integrations (Spotify / Apple Music / HealthKit).

Out of scope (explicit cuts)

• Ladder's technical stack. Reverse-engineering their implementation is a separate question and does not change what Ilia builds. Cut.
• Android parity. Ilia is building iPhone-first. Note Android existence only.
• Full competitive landscape survey. Ladder is the subject; peers appear only as positioning reference points, not as their own teardowns.
• Ilia's own product strategy. This Inquiry describes Ladder. Strategy is the handoff.
• Legal/IP analysis of what can be copied. Flag if something looks patented; do not opine.

Success criteria

The Inquiry succeeds if Ilia can, from the Consolidated Brief alone: 1. Describe Ladder's core loop and monetisation to someone else without opening the app. 2. Name the 3–5 mechanics that actually drive its 4.95★ / 180k-rating retention. 3. Point at the specific things that are expensive to copy (content supply, coach roster) versus cheap to copy (UI patterns, trial structure). 4. State where Ladder is weak or unserved — the opening.

Areas

Source legend (Inquiry-wide)

• [AS] Apple App Store listing + metadata (assets/appstore-us.json), pulled 2026-08-16
• [SS] Official App Store screenshots (assets/screenshots/appstore-NN.png)
• [IB] Ilia's first-hand field notes (verbatim above)
• [WEB] joinladder.com and Ladder-owned web properties
• [REV] User reviews (App Store, Google Play, Reddit, Trustpilot)
• [PRESS] Press, funding databases, interviews, podcasts
• [VID] Video walkthroughs / reviews (YouTube, TikTok)
• [3P] Third-party analytics estimates (Sensor Tower, Appfigures, data.ai) — mark as estimates, never as fact

Inquiry: Ladder (fitness app) competitor teardown — core mechanics, product model, business model. Location: personal (Ilia's own fitness-app venture, not Fortis).

Log

2026-08-16T22:45 · — → Scoped · Goal captured verbatim. Subject identified from primary
                     source: App Store id 1502936453, bundle com.ladder.bootcamp, Ladder
                     Technologies Inc. Pulled appstore-us.json + 10 official screenshots at
                     1242x2208. SCOPE.md written with 6 Areas and an 8-source legend.
2026-08-16T22:47 · Scoped → Triaged · Classified FULL. Six Areas, real dependency (A1
                     surface grounds A2/A3/A5; A4 content feeds A6 economics). Output must
                     be a Consolidated Brief, not a single Finding.
2026-08-16T22:52 · Triaged → Grilled · Scope Grill run as 2 rounds of orchestrator
                     self-adversarial probing (see below), per Ilia's standing preference
                     for <=1 grill round with the human. One human question deferred to
                     post-dispatch so Scouts are not blocked on it.
2026-08-16T22:55 · Grilled → Authorized · 6 Area Briefs written, 6 Scouts dispatched in
                     parallel (general-purpose, not Explore — Scouts must Write their
                     findings file, which Explore cannot do).

Scope Grill record

Framework floor is >=2 rounds before authorizing any Area Brief. Ilia's standing preference (memory: feedback_question_volume) is that framework configuration is the Orchestrator's call and the human gets at most one round. Both honoured: two full adversarial rounds run against my own scope, one question reserved for Ilia.

Round 1 — goal clarity and scope cuts

Probe: "Teardown" is a proxy. What decision does this actually gate? It gates what Ilia builds and what he refuses to build. A pure feature inventory would fail — Ladder has 20 coach-led programs and a chat community that a solo developer cannot replicate. The Brief must therefore separate cheap to copy from expensive to copy. → Added success criterion 3 and the cost axis to A4.

Probe: is the competitive landscape in or out? Out. Ilia named one competitor. Surveying Fitbod/Future/Peloton/Centr would triple the Inquiry and answer a question he did not ask. Peers appear only as positioning anchors inside A6. → Recorded as an explicit cut.

Probe: is Ladder's tech stack interesting? No. It does not change what he builds, and it duplicates the separate stack workflow already running. Cut.

Round 2 — information sufficiency and sequencing

Probe: can these Areas actually be answered from outside the paywall? Partially, and this is the Inquiry's main risk. A1/A4/A6 are well served by public sources (App Store, joinladder.com, press, funding data). A2/A3/A5 depend on post-signup and post-paywall surfaces that no Scout can reach. Mitigations: Ilia's own field notes are first-party evidence for exactly those Areas; video walkthroughs show paid surfaces; the 10 official screenshots already expose in-workout, Journal, Chat and analytics screens. → Every Scout instructed to mark paywall-blocked claims as Parked, never inferred. This is the "Proven rule" and it is the single most likely failure mode here.

Probe: do the Areas need sequencing? A1 nominally grounds the rest, but the Orchestrator has already read all 10 screenshots and folded the screen inventory into every Brief as shared context. That removes the dependency, so all six run in parallel.

Probe: what does "done" look like, and who carries it forward? Done = Consolidated Brief + plain-language SUMMARY. Carried forward either into an AtelierBrain Mission Seed (if he builds) or a local task list. Decided at handoff.

Probe: what is genuinely unanswerable without Ilia? One thing: his own product's angle. A teardown without a thesis is a feature list. This does not block any Scout — it only shapes synthesis — so it is asked after dispatch.

Parked / open

• [x] ~~Ilia's differentiation thesis~~ — deferred by Ilia to "a separate step". The case study still needs a stated point of view; synthesis must propose one rather than wait for it.
• [x] ~~Whether "our competitor" implies a team venture or a solo build~~ — answered, and the answer invalidated the premise. See below.

Premise correction — 2026-08-16T23:05

2026-08-16T23:05 · Authorized (unchanged) · SCOPE.md consumer section REVISED. The Inquiry
                     was scoped on the assumption that Ilia is building his own fitness
                     app. He is not. Actual purpose: a HIRING CASE STUDY — he will bring a
                     prototype of a Ladder competitor to a company he is interviewing with,
                     showing how he would approach it if he joined.
                     Scouts NOT recalled: the teardown is unchanged by this and in fact
                     becomes a primary deliverable rather than background. Re-weighting
                     applied at synthesis instead: A6 economics and A4 content-supply gain
                     weight; shipping-infrastructure depth drops to near zero.

This is the Scope Grill's Round-2 probe ("what decision does this actually gate?") coming back to bite — the honest answer is that the Orchestrator accepted "our competitor" without establishing who "our" was. Recorded so the framework's scope.md step gets a mandatory "who consumes this and what do they do with it" question before Areas are drafted.

Log (continued)

2026-08-17T00:05 · Authorized → Findings-Complete · 6/6 Findings on disk. A5, A2, A1, A6,
                     A3 returned cleanly. A4 was TERMINATED by an account session limit —
                     but had already written its Finding (802 lines, the largest of the
                     six). Its dying message flagged one measurement it wanted to correct
                     ("the public workout page serves a preview, not the full session");
                     inspection shows the Finding ALREADY labels that figure as a
                     preview-based extrapolation with assumptions stated. No content lost.
                     A4 is therefore counted complete, not Stalled.
                     Degradation to note: the session-wide WebSearch budget (200) was
                     exhausted mid-run. Several Findings carry parked items that more
                     searching would have closed. Recorded in the Brief's confidence map.
2026-08-17T00:20 · Findings-Complete → Synthesized · consolidated-brief.md + SUMMARY.md
                     written in the main thread. Cross-Area confirmation step run by the
                     Orchestrator rather than via AskUserQuestion (Ilia's standing
                     preference for question volume); clusters are stated in the Brief and
                     he can redirect.

Cross-Area headline

The Inquiry's most valuable output is a correction of its own working assumption. The Orchestrator twice told Ilia that Ladder's moat was an expensive human content treadmill that a solo builder could not copy. A6 and A4 jointly refuted it:

• S&M ≈ 56% of revenue; coaches + studio ≈ 7% [A6]
• ~1,716 authored sessions/year serve ~6,864 delivered workout-days — 4:1 leverage [A4]
• Ladder's own Team Manager job req lists Claude as a required tool [A4]

The moat is paid acquisition and a coach's name, not content volume. Corrected in the Brief, and flagged to Ilia in-session.

Consumer profile — settled 2026-08-16T23:12

• Hiring company: generalist tech company. Fitness is the substrate, not the domain. → The panel does not know Ladder. Domain depth is not the grade; judgement is. → Any deliverable must teach enough Ladder to make a thesis land, fast, then land it. → A dense teardown works against him. The Findings are raw material, not the artifact.
• Brief: open-ended. Ilia sets the scope himself. → The scoping decision is itself being graded. A tight, well-argued slice with an explicit "here is what I deliberately did not build" reads senior. A sprawling feature-complete clone reads junior. → Optimise for impact per hour.

Synthesis directives (binding on the Consolidated Brief)

1. Lead with a thesis, not a description.
2. Assume zero fitness-domain knowledge in the reader.
3. Recommend ONE mechanic to prototype, and justify the cut list as loudly as the build list.
4. Prototype target: web, iPhone-styled, live URL. Not native. Not TestFlight.
5. Keep the A4 content-treadmill economics — a generalist panel will grade commercial reasoning even in a product exercise, and "their moat is also their ceiling" is the kind of argument that travels outside the domain.

Open

• [ ] Does Ilia want the wedge proposed for him? He deferred it as "a separate step" while the framing was still "my own app". Under the case-study framing there is no deliverable without a point of view, so synthesis will propose one and mark it as a proposal he can override.
• [ ] Deadline unknown. Affects prototype fidelity only.