Дата: 2026-05-07 Режим: deep/full Основа: чтение standalone-репозитория и сохранённых в нём исследований рынка, конкурентов и текущих workflow трейдеров.
Strong, but only after validation.
Текущий wedge trade upload + rule violation audit является правильным первым сужением, потому что он проверяет платную боль без брокерских API, live capital, SaaS, стратегии и Telegram scraping. Но продукт пока не доказал бизнес: текущий код уже закрывает Phase 5 concierge workflow, а главный риск теперь не engineering, а то, заплатят ли трейдеры за post-trade дисциплину повторно. Следующий шаг - не расширять продукт, а продать 3-5 ручных аудитов реальным трейдерам с реальными экспортами и письменными правилами. Если трейдеры говорят "классный отчет", но не платят, не дают экспорт и не возвращаются через 2-4 недели, проект надо остановить или split/pivot.
Что это реально, исходя из repo: локальный deterministic audit workflow. Трейдер дает trade export и risk policy, оператор нормализует CSV, система детерминированно проверяет правила, атрибутирует P&L, генерирует Markdown report, Telegram-ready packet и artifact manifest. В коде уже есть canonical trade schema, CSV importer, policy schema, review packet gate, deterministic rule evaluation, P&L attribution, report model, claim guard, manifest, retention/delete и pilot regression fixtures.
Что это не: не trading journal общего назначения, не backtesting platform, не signal analytics product, не live risk guard, не broker automation layer, не AI strategy generator, не SaaS dashboard и не investment advice.
Какие комментарии трейдера совпадают с текущим продуктом: "первый вариант может просто загружать сделки и проверять все сделки против правил"; Telegram-first packaging как delivery surface; идея контроля дисциплины после факта; возможность small teams / Russian trading teams как warm-intro ICP.
Какие комментарии ведут к отдельным продуктам: backtesting strategies, live API lockout "минус 2%, 24 часа не торговать", Telegram/X signal stats/equity curve, natural-language strategy-to-code, screenshot/chart learning. Это adjacent wedges, но они имеют другую data model, другой risk profile и другую buying trigger.
Что не строить сейчас: broker/exchange API, order blocking, public SaaS, full dashboard, mobile app, strategy generator, Telegram channel parser, paid X API dependency, private Telegram scraping, marketplace, live execution, institutional compliance tooling.
Самый сильный narrow wedge: "audit-as-evidence" для трейдеров, у которых уже есть письменные risk rules и внешний стимул соблюдать их: prop/funded-account traders, active discretionary crypto traders с self-imposed rules, small teams/coaches, которые хотят weekly accountability report.
Скоринг: 1 = плохо/слабый сигнал, 5 = сильно/хороший сигнал.
| Area | Score | Evidence |
|---|---|---|
| Pain intensity | 3/5 | Потеря денег из-за rule-breaking реальна, особенно у prop/funded traders: FTMO прямо определяет max daily loss/max loss как account-failing rules. Но еще не доказано, что pain достаточно сильна именно для post-trade paid audit. |
| Buyer clarity | 3/5 | Primary ICP можно сузить до prop/funded-account traders и small teams с explicit rules. Broad retail слишком размыт и часто не имеет формализованных правил. |
| Urgency | 2/5 | After-the-fact report может быть vitamin, если нет внешнего наказания, coach/team accountability или payout risk. Urgency вырастает, если трейдер уже терял funded account, payout или капитал из-за нарушения rules. |
| Differentiation | 3/5 | Journals уже умеют analytics, imports, risk metrics, AI/replay/backtesting. Дифференциация только в deterministic violation truth + source-row evidence + self-defined policy audit, а не в "красивой аналитике". |
| Speed to validate | 5/5 | Можно валидировать вручную за 14 дней: 10 calls, 5 real exports, 3 paid reports. Код уже достаточен для concierge pilot, но его не надо продавать как SaaS. |
| Founder advantage | 3/5 | Технический founder может быстро строить deterministic audit artifacts, но это же главный риск: уйти в infra вместо продаж. Advantage появится только через доступ к trader communities и теплым интро. |
| Trust/liability risk | 3/5 | Manageable в v1, потому что нет live control и advice claims. Становится 1/5 при broker API/order blocking: API trading несет trading-loss liability, regulatory scrutiny и security burden. |
| Engineering scope risk | 4/5 | Риск низкий, если держать local-first CLI, fixed schemas, rule taxonomy, claim guard. Риск резко растет при broker integrations, Telegram scraping, SaaS auth, AI strategy generation. |
| Distribution access | 2/5 | Пока нет доказанной access path. Russian teams могут быть reachable, но relationship-heavy sales нельзя считать scalable channel до теплых интро и paid pilots. |
| Retention potential | 3/5 | Weekly/monthly audit может повторяться, если отчет становится accountability loop. Если это одноразовая curiosity report, retention слабая. |
| Fit with current codebase | 5/5 | Repo уже идеально соответствует narrow wedge: deterministic local audit, artifacts, Telegram-ready summary, retention, no live API guardrails. |
| Wedge | User | Pain | Current workaround | WTP likelihood | Sales difficulty | Engineering complexity | Liability/compliance risk | Fit with workspace | Fastest validation test | Verdict |
|---|---|---|---|---|---|---|---|---|---|---|
| 1. Trade Upload + Rule Violation Audit | Prop/funded traders, active discretionary traders, small teams/coaches | "Я нарушил свои правила, потерял деньги, не вижу точную картину" | Excel/Sheets, broker CSV, TradeZella/TraderSync/Tradervue/Edgewonk, Notion, manual journal, ChatGPT | Medium: если есть external accountability, payout risk или coach | Medium: trust/data-sharing barrier | Low-medium: importer + policy schema + deterministic rules | Low if no advice/live control | Excellent | 10 prospects -> 5 exports -> 3 paid reports | Build only as manual/concierge pilot; validate now |
| 2. Live Broker/Exchange Risk Guard | Traders who want self-lockout and pre-trade blocking | "Я не могу остановиться после -2%" | Broker limits, exchange risk settings, prop dashboard rules, manual discipline, separate account, platform max loss tools | Potentially high, but only if trust is solved | High: requires credentials, trust, legal comfort | High: APIs, secrets, order state, latency, fail-safe, broker variance | Very high: capital-adjacent control, trading loss disputes, advice/control boundary | Poor for current v1 | Concierge interview: last live breach, willingness to use read-only monitor first, legal/security objection | Defer; decision gate only |
| 3. Hypothesis & Backtesting Workbench | Systematic retail traders, strategy testers | "У меня есть идея, хочу быстро проверить" | TradingView Strategy Tester, MT5 tester, Python notebooks, Excel, backtesting.py/vectorbt, TradeZella/TraderSync/TradesViz replay/backtest | Medium-high in broad market, but crowded | High: strong expectations and benchmarks | High: data, execution assumptions, OOS, slippage, strategy DSL | Medium-high: performance claims risk | Weak for this project; belongs in a separate hypothesis/backtesting product | Ask paid pilot users whether their next paid request is audit or hypothesis test | Defer/separate unless audit pilots demand it |
| 4. Telegram Signal Analytics Bot | Signal subscribers, crypto traders, market researchers | "Канал обещает win rate, хочу проверить" | Telegram groups, screenshots, Telemetrio/TGStat, SignalBlink/BotLedger-like tools, manual spreadsheet | Medium: viral curiosity; trust hard | Medium: public channels easier, private channels harder | High: parsing signals, market data, edits/deletes, private access, X API cost | Medium: not advice if framed as analytics, but claims sensitive | Separate sandbox already exists | Manually audit 3 public channels and sell one report to subscribers | Separate product; do not contaminate TRA |
| 5. Natural-Language Strategy Builder | Novice systematic traders | "Опишу стратегию, пусть сделает код/backtest" | ChatGPT/Claude, TradingView Pine, freelancers, strategy builders | Low-medium: many want magic, few pay for rigorous constraints | High: expectations unrealistic | Very high: codegen, backtesting, validation, support | High: performance/advice claims | Bad | Landing/interview test only; no build | Reject for now |
| 6. Chart/Screenshot-to-Rules Assistant | Visual discretionary traders/coaches | "Покажи на графике сетап и найди похожее" | Screenshot journals, chart markup, TradingView screenshots, coaching sessions | Unknown | High: users struggle to specify truth | Very high: OCR/multimodal, chart state reconstruction, labeling | Medium: false pattern claims | Bad for v1 | Manual labeling service with 3 users; no automation | Reject/defer hard |
The business works only if active traders with explicit rules will repeatedly pay for a deterministic post-trade audit that shows rule violations, source rows, and violation-attributed P&L better than their current journal/spreadsheet workflow.
| Assumption | Why it matters | Evidence currently available | Fastest test | Disconfirming signal | Decision if false |
|---|---|---|---|---|---|
| 1. Traders will share real exports and written rules | Without real data, no audit product exists | Trading tools already rely on imports; Tradervue documents import quotas and supported broker/platform workflows; TradingView and MT5 expose export/report paths | Ask 10 qualified prospects for last 30-90 days export + written rules before showing product | Fewer than 5/10 qualified prospects provide export/rules within 7 days | Stop upload product; pivot to education/coaching content or no-build |
| 2. Rule violations are painful enough to pay for | Curiosity reports do not build business | Prop firms enforce daily/max loss rules; funded traders have account-failing stakes | Sell $49-$149 one-time audit before building more | They accept free sample but refuse payment or delay indefinitely | Keep only as internal tool or pivot ICP |
| 3. Deterministic audit beats journals | Existing journals are strong and habit-forming | TradeZella, TraderSync, Edgewonk, TradesViz already offer imports, risk analytics, replay/backtesting/AI | Compare report against user's current journal in a call | User says their journal already shows everything needed | Narrow to teams/coaches or prop rules; do not compete with journals broadly |
| 4. Post-trade is enough before live blocking | If users only want blocking, audit is weak | Trader colleague asked for API lockout; exchanges/brokers already have some risk controls; live API adds high burden | Discovery: last rule breach, what they wanted at the moment, what they did after | Majority say "I only pay if it blocks trades live" | Defer live guard but evaluate read-only monitor as separate gate |
| 5. Reports can become recurring workflow | One-time autopsy is not a company | Current repo supports repeatable artifacts and retention, but no pilot evidence yet | 6-week pilot: weekly/monthly audit cadence | No repeat audit, no referral, no coach/team sharing | Stop productization; sell one-off service only or pivot |
| Risk | Severity | Why it matters | Evidence to seek | Fast test | Kill/pivot threshold |
|---|---|---|---|---|---|
| Traders may not pay for post-trade discipline audits | High | They may like the report but not change payment behavior | Prepaid audit orders | Offer $49-$149 manual audit to 10 qualified prospects | <3 paid audits from 10 qualified prospects |
| Existing journals may be good enough | High | TradeZella/TraderSync/Edgewonk/TradesViz already own journaling, analytics, imports, replay/backtesting | Side-by-side comparison with user's existing tool | Ask users to show last journal review workflow | >50% say current journal already identifies rule breaks clearly enough |
| Traders may want live blocking, not after-the-fact reports | High | If the acute pain is impulse control at execution time, audit is late | Past breach stories: what they wanted in the moment | Discovery around last max-loss/cooldown breach | >50% only show WTP for live lockout |
| Live blocking creates liability and API/security burden | Critical | Order blocking means secrets, availability, false blocks, missed blocks, trading loss disputes | Legal memo, API terms, security review, broker coverage | Read-only monitoring pilot before any trade permissions | Any pilot requires trade permission before trust/legal gates |
| Exports may be too messy to normalize manually at scale | Medium | Broker CSVs vary; commission/timestamps/open P&L often missing | 5-10 real exports from ICP | Run manual taxonomy across exports | >40% exports require custom manual cleanup that users won't tolerate |
| P&L attribution may become disputed | High | If user disputes "violation-attributed P&L", trust dies | User review of source rows and attribution assumptions | Deliver report with row evidence and ask user to challenge it | >2 material disputes in first 5 audits |
| Russian trading teams may be reachable but sales may be relationship-heavy | Medium | Warm intros may work; cold sales may not | Intro conversion rate, trust/privacy objections | 10 warm intro asks through trader colleague network | No warm intros or no export sharing despite interest |
| Telegram signal analytics may be more viral but a different product | Medium | Could distract founder into a market with different user, data, and claims | Separate landing/manual report test in sandbox | One manual public-channel report sold separately | Do not merge; split if stronger evidence appears |
| Founder may keep building infra instead of selling reports | Critical | Code is already ahead of market proof | Calendar/time audit | Freeze feature work; only sales + manual delivery for 14 days | Any new feature before 3 paid reports triggers build pause |
Exact painful moment: after a losing day/week, the trader suspects they broke rules but cannot prove exactly where, how often, and how much P&L came from those breaches. The sharpest version is not "I want analytics"; it is "I violated max daily loss/cooldown/position size, I may lose funded status or payout, and I need evidence."
Who feels it most sharply: prop/funded-account traders, active discretionary crypto/futures traders with real written rules, small teams/coaches reviewing discipline. Broad retail traders with vague "be disciplined" goals feel shame but often lack structure and WTP.
What they do today: export broker/platform history, use Excel/Google Sheets, use TradingView/MT5 reports, use TradeZella/TraderSync/Tradervue/Edgewonk/TradesViz, journal in Notion/paper, ask a coach, screenshot charts, or paste data into ChatGPT/Claude.
Workaround cost: time, inconsistency, shame/privacy friction, manual mistakes, no source-row evidence, no explicit deterministic policy, and reports that visualize performance without forcing "which rule did I break?"
Painkiller or vitamin: currently a conditional painkiller. It is painkiller for users with external accountability or repeated costly rule-breaking; vitamin for broad retail traders who merely want another dashboard.
Behavior proving urgency: user prepays, sends real export/rules within 24 hours, argues about source rows, asks for next audit date, shares report with coach/team, or asks for monthly accountability.
| JTBD | Trigger | Desired outcome | Frequency | WTP signal | Retention strength | Build decision |
|---|---|---|---|---|---|---|
| 1. Upload trades and detect rule violations | Bad day/week, funded account review, coach session | Know exact rule breaches with timestamps/source rows | Weekly/monthly for active traders | Pays for audit, provides export | Medium-high if accountability loop | Build now, concierge-first |
| 2. Quantify P&L damage from rule violations | Suspects discipline costs money | Separate compliant vs violating P&L | Weekly/monthly | Challenges attribution and still pays | High if trusted | Build now, but with careful claim guard |
| 3. Identify repeated discipline patterns | Repeated losses or review session | See recurring rule IDs/days | Weekly/monthly | Asks for trend over multiple audits | High if recurring | Build now in report, not dashboard |
| 4. Produce shareable accountability report | Coach/team/funded review | Show proof without raw data dump | Monthly/after incidents | Sends to mentor/team | Medium | Build now as Markdown/Telegram-ready packet |
| 5. Enforce self-defined risk rules before trading | Impulse after loss, max-loss breach | Stop trading mechanically | Daily/intraday | Will connect account/API after trust | Potentially high | Later/separate decision gate |
| 6. Turn strategy idea into testable hypothesis | New strategy idea | Backtest or falsify hypothesis | Ad hoc | Pays for research/backtest | Medium | Later, only if paid audit users demand |
| 7. Evaluate Telegram/Twitter signal source | Considering paid signal group | Verify channel's real performance | Ad hoc/ongoing | Pays for public channel report | Medium-high but different product | Separate sandbox/product |
Current behavior competitors
| Competitor/substitute | What users use it for | Why they might stay | Why they might switch | What TRA must prove |
|---|---|---|---|---|
| Excel / Google Sheets | Manual P&L, tags, rule notes, charts | Free, flexible, private, familiar | Error-prone, no deterministic source-row audit, hard to repeat | Faster credible report than spreadsheet in <48h |
| Broker CSV/activity statements | Raw truth source | Already available, trusted | Not interpreted against personal rules | Normalize without hiding source evidence |
| TradingView | Strategy tester, charts, export strategy data | Habit, strong charting, strategy workflow | Does not audit executed broker trades against personal risk policy | TRA is post-trade discipline audit, not chart/backtest replacement |
| MetaTrader / MT5 reports | Account history, orders/deals/positions report | Native to FX/CFD workflow | Report is generic, not policy audit | Convert history into rule violation evidence |
| QUIK / MOEX terminals | Russian market trading, exports to Excel/database | Entrenched terminal workflow | Outdated/manual review burden | Support export taxonomy without live integration |
| Notion / paper journal | Reflection and psychology | Low friction, private | No deterministic P&L attribution/source rows | Produce hard evidence, not just notes |
| ChatGPT/Claude | Ad hoc trade review and narrative | Easy and cheap | Cannot be final truth for violations/P&L | Deterministic artifacts and claim guard |
Direct product competitors
| Competitor | What users use it for | Why they stay | Why they switch | What TRA must prove |
|---|---|---|---|---|
| TradeZella | Journaling, playbooks/rules, risk tools, analytics | Polished UI, broker imports, education, reporting | May not audit custom written rules with source-row determinism | TRA gives stricter audit report, not another journal |
| TraderSync | Journaling, trade replay, AI coach, backtesting/replay | Broad features, mobile, imports, AI | Feature-rich but not focused on deterministic rule breach evidence | TRA is narrow, faster, evidence-first |
| Tradervue | Import/share trades, reports, community | Established workflow and free tier | Dated or limited; manual import friction | TRA wins on paid incident audit, not broad journaling |
| Edgewonk | Journal, psychology, mistake tracking, Tiltmeter, performance simulator | Strong discipline/psychology positioning | User wants external deterministic audit/report | TRA produces artifact trail and violation P&L |
| TradesViz | All-in-one journal, 600+ stats, AI, simulators, auto-sync | Feature breadth and free plan | Too broad; user needs policy audit | TRA must be sharper, not broader |
Indirect competitors
| Substitute | Use | Switch condition |
|---|---|---|
| Prop-firm dashboards | Rule compliance, daily/max loss, payout gates | TRA adds independent post-mortem and personal rules, not replacing firm rules |
| Exchange/broker risk settings | Position/leverage/risk limits | TRA handles personal discipline rules after export; live settings remain platform-owned |
| Trading coaches/mentors | Accountability and interpretation | TRA gives coach evidence packet and saves manual review |
| Telegram groups | Social accountability/signals | TRA is private evidence, not entertainment/community |
AI-wrapper competitors: generic LLM review, AI journal assistants, Cypher-like trader coaches. Their weakness is that they can explain and summarize but must not own final violation truth, P&L arithmetic, or source-row traceability.
Non-consumption: many traders do nothing after a breach because of shame, laziness, low account size, or belief that "next time I will be disciplined." This is the real default.
Real enemy: habit + shame + trust + workflow inertia + messy data + lack of urgency. The product loses if it becomes another dashboard instead of a painful accountability artifact.
| ICP | Pain intensity | Ability to pay | Trust/liability requirements | Access path | Fit with current product | Early adopter quality | Verdict |
|---|---|---|---|---|---|---|---|
| Prop/funded-account traders | High | Medium | Need accuracy, privacy, no advice | Prop communities, Discord/Telegram, coaches, warm intros | High | High: explicit rules, frequent exports | Primary |
| Active discretionary crypto traders | Medium-high | Medium | Privacy, exchange export support, no live API in v1 | Telegram groups, Twitter/X, communities | Medium | Good if they have written rules | Secondary |
| Retail discretionary traders | Low-medium | Low-medium | Low formal rules, high churn | Social media/courses | Low-medium | Weak: vague discipline | Defer |
| Systematic retail traders | Medium | Medium | Need backtest validity, OOS rigor | TradingView/Python communities | Medium-low | Good but pulls toward backtesting | Secondary/defer |
| Small trading teams | High | Medium-high | Need confidentiality, repeatability, team workflow | Warm intros, local communities | High | Strong if team lead owns process | Secondary |
| Russian trading teams / small funds | Medium-high | Medium-high | Relationship trust, data privacy, local terminal/export quirks | Trader colleague, warm intros, Russian Telegram communities | Medium-high | Strong but access uncertain | Secondary after warm-intro proof |
| Trading coaches / mentors | High | Medium | Need report they can use with students | Coach DMs, paid communities | High | Strong distribution leverage | Secondary/partner |
| Signal-channel subscribers | Medium | Low-medium | Wants public proof; different data | Telegram/crypto communities | Low | Different problem | Defer to signal sandbox |
| Signal sellers / influencers | Low/negative | Medium | Incentive conflict; may dislike audit | Creator outreach | Low | Bad early adopter | Reject for TRA |
Primary ICP: prop/funded-account traders who already have written risk rules and can export trade history.
Justification: they have explicit external rules, painful consequences, repeat review cadence, and a reason to care about max daily loss/drawdown/cooldown evidence. This ICP also forces the product to stay deterministic and source-traceable. Broad retail will praise the idea and not pay; signal subscribers are a separate product; live risk guard users introduce too much liability for v1.
Where to find prospects
- Trader colleague's warm network first: prop/funded traders, crypto/futures scalpers, small Russian/RU-speaking teams, coaches.
- Telegram/Discord groups where funded-account rules, payout denials, daily drawdown, overtrading, revenge trading are discussed.
- Coaches/mentors who already review journals manually.
- Users posting about TradeZella/TraderSync/Edgewonk pricing, journaling fatigue, max daily loss, funded account failures.
Who founder should contact first
Start with 20 warm or semi-warm prospects, not strangers: 10 active prop/funded traders, 5 coaches/team leads, 5 active crypto/futures traders with written rules. Do not start with signal sellers or broad retail beginners.
Outreach sequence
- DM with a specific ask: "Я проверяю, платят ли трейдеры за независимый аудит нарушений risk rules по реальному export. Не продаю SaaS. Хочу посмотреть, как ты в последний раз разбирал нарушение max loss/cooldown."
- 15-minute call about last real breach and current workflow.
- Ask for anonymized 30-90 day export + written rules.
- Offer paid manual audit: one report within 48-72 hours.
- Review report live, ask user to challenge source rows and attribution.
- Ask for next audit date or referral only after payment/report review.
What not to automate
Do not automate onboarding, broker import, Telegram bot delivery, public dashboard, payment funnel, strategy parsing, or signal analytics. The founder needs to hear objections directly.
First call must test
- Did they have a real recent costly violation?
- Do they already track rules in a concrete way?
- Can they export data quickly?
- Are they willing to show data despite shame/privacy?
- Do they pay for trading tools/coaching now?
- Would an evidence report change behavior, coach review, payout preparation, or team accountability?
Success
- 10 qualified calls in 14 days.
- 5 real exports + written rules.
- 3 prepaid or paid audit reports.
- At least 2 users schedule a second audit.
- At least 1 coach/team lead asks for multi-trader package.
False-positive enthusiasm
- "Крутая идея" but no export.
- Wants live blocking only.
- Wants strategy generator/backtest instead.
- Sends screenshots but no trade history.
- Refuses payment but asks to be notified when SaaS exists.
- Wants Telegram signal analytics, not own-trade audit.
10 discovery questions about past behavior
- Когда ты в последний раз нарушил свое risk rule? Что именно произошло по времени и сделкам?
- Как ты понял, что правило было нарушено: терминал, spreadsheet, журнал, prop dashboard, ощущение после факта?
- Сколько денег, payout, account status или времени это стоило?
- Где были записаны правила до нарушения: документ, Notion, sheet, чат, память?
- Когда ты последний раз экспортировал сделки? Из какого терминала/брокера и в каком формате?
- Как ты сейчас ведешь journal: TradeZella/TraderSync/Edgewonk/Excel/Notion/ничего? Что платишь сейчас?
- Как ты проверяешь max daily loss, max drawdown, cooldown после loss или position size?
- Что ты сделал после последнего нарушения: остановился, уменьшил размер, показал кому-то отчет, проигнорировал?
- Что мешало показать сделки coach/team/friend: стыд, приватность, messy export, отсутствие формата?
- Какой последний отчет/скрин/журнал ты показывал coach/team/prop support? Что они сказали, чего не хватало?
Decision: первый pilot должен быть A. manual trade/risk audit report.
Почему не B: автоматизированный CLI уже есть, но продавать надо outcome, а не tool. Почему не C: Telegram-ready delivery полезен как упаковка, но не core value. Почему не D: live risk guard преждевременен и рискован. Почему не E: backtesting - другой buying trigger.
Core pilot promise
"Пришли trade export и written risk rules. За 48-72 часа получишь audit report: какие правила нарушены, source rows/timestamps, repeated patterns, worst days, violation-attributed P&L, limitations, next-review checklist. Это не investment advice и не live control."
What is manually concierge
- Prospect qualification.
- Export intake and anonymization.
- Mapping user's rule text into policy schema.
- Resolving ambiguous columns/rules.
- Report review and delivery.
- Payment, follow-up, and objection logging.
What must be productized
- Deterministic importer for most common pilot exports.
- Explicit policy schema.
- Source-row traceability.
- Violation records.
- Reconciled attribution.
- Markdown report.
- Claim guard/disclaimer.
- Manifest and local retention/delete.
Эти pieces уже mostly implemented in repo. Do not rebuild them.
What must be cut
Live broker/exchange API, order blocking, Telegram sending bot, public SaaS, dashboard, mobile app, signal parsing, AI rule truth, strategy generator, backtest workbench, private group scraping.
2-week validation sprint
- Days 1-2: compile 50 prospects, prioritize 20 warm/semi-warm.
- Days 2-5: 10 calls, no product demo before past-behavior questions.
- Days 3-7: collect 5 exports + written rules.
- Days 5-10: deliver 3 paid manual reports.
- Days 10-14: live report review, collect disputes, ask for second audit/referral.
6-week concierge pilot
- Week 1-2: first 3 paid reports.
- Week 3-4: repeat audits for same users; refine rule taxonomy/export taxonomy.
- Week 5: coach/team package test.
- Week 6: decide productization based on repeat payment and data normalization pain.
Behavioral success metrics
- 50%+ of qualified prospects provide real export/rules.
- 30%+ pay for first audit.
- 50%+ of paid users review report live.
- 30%+ request or schedule repeat audit.
- 20%+ refer another trader/coach.
Payment test
Take payment before or immediately after data intake. Free sample can be one anonymized example, not a full free audit.
Retention test
Ask for second audit at 2-4 week interval. Retention is not opening a dashboard; retention is sending new export/rules and paying again.
Referral test
Ask: "Кому еще ты показывал этот report or who else has this exact problem?" Referral without prompt is strongest.
Kill criteria
- <3 paid reports from 10 qualified prospects.
- <5 real exports/rules from 10 qualified prospects.
-
2 major attribution disputes in 5 audits.
- No repeat audit scheduled after 6 weeks.
- Most demand is for live blocking/backtesting/signal analytics rather than audit.
- Objective: prove paid pain before more engineering.
- Why now: code is ahead of market evidence.
- Scope in: interviews, manual mock report, payment ask, export/rule collection.
- Scope out: all new code, UI, broker API, Telegram bot, signal analytics.
- Entry criteria: current repo baseline accepted; founder has prospect list.
- Exit criteria: 10 qualified calls, 5 exports/rules, 3 paid reports or explicit failure.
- Main artifacts: call notes, objection log, anonymized workflow examples, payment receipts.
- Engineering tasks: none.
- Validation tasks: past-behavior discovery, payment ask, data access test.
- Risks: founder hides behind building; prospects praise but do not pay.
- Kill/pivot criteria: fewer than 3 paid reports from 10 qualified prospects.
- Objective: deliver audit value manually while using existing deterministic pieces only where helpful.
- Why now: tests core assumption with real inputs.
- Scope in: manual normalization, manual policy mapping, generated Markdown report, Telegram-ready copy if requested.
- Scope out: product UI, self-service, broker sync, strategy/backtest, live lockout.
- Entry criteria: at least 3 prospects agree to paid/manual audit.
- Exit criteria: 3 paid reports delivered; each reviewed with user.
- Main artifacts: audit report, source-row mapping notes, user dispute log.
- Engineering tasks: only bug fixes needed to complete paid reports; no feature expansion.
- Validation tasks: report usefulness, attribution trust, WTP, privacy objections.
- Risks: report looks impressive but does not change behavior.
- Kill/pivot criteria: users do not pay, do not review, or do not request repeat.
- Objective: learn which exports/rules actually appear in pilots.
- Why now: scale risk is importer mess, not evaluator code.
- Scope in: broker/terminal/export taxonomy, rule taxonomy, timezone/session patterns, Russian terminal notes.
- Scope out: broad broker integration library.
- Entry criteria: 5 real exports across at least 2 platforms.
- Exit criteria: top 2-3 export formats and top 6-10 rule patterns documented.
- Main artifacts: export taxonomy doc, rule taxonomy doc, unsupported-data list.
- Engineering tasks: small parser adjustments only after repeated demand.
- Validation tasks: measure cleanup time per audit.
- Risks: every user needs custom cleanup.
- Kill/pivot criteria: average normalization >2 hours per audit with no willingness to pay for concierge.
- Objective: automate only repeated manual work.
- Why now: paid reports reveal repeated formats/rules.
- Scope in: local-first CLI, deterministic CSV import, explicit policy schema, review packet, artifact manifest.
- Scope out: SaaS onboarding, OAuth, live broker credentials, AI-owned rule mapping.
- Entry criteria: repeated export/rule patterns from paid pilots.
- Exit criteria: same audit can be rerun deterministically with stable hashes.
- Main artifacts: normalized trades, approved policy YAML, review packet, manifest.
- Engineering tasks: support top repeated export formats, improve validation errors, local retention/delete.
- Validation tasks: operator time saved, user trust in source-row evidence.
- Risks: overbuilding generic importer.
- Kill/pivot criteria: no repeated formats/rules after 10 audits.
- Objective: make violation truth and P&L attribution defensible.
- Why now: this is the actual differentiated asset.
- Scope in: max daily loss, max drawdown, cooldown, max position size, forbidden assets, leverage warnings, violation-attributed P&L, claim guard.
- Scope out: counterfactual returns, causal-loss claims, strategy adherence without approved deterministic mapping.
- Entry criteria: users have explicit rules and disputes from pilots are understood.
- Exit criteria: users accept source-row evidence and attribution assumptions in review.
- Main artifacts: violations JSON, attribution summary, report sections, golden fixtures.
- Engineering tasks: evaluator coverage only for validated rule types; regression tests.
- Validation tasks: ask users to challenge each violation row.
- Risks: P&L attribution gets disputed or overclaims causality.
- Kill/pivot criteria: material attribution disputes in >30% paid reports.
- Objective: deliver where traders already communicate without changing core product.
- Why now: only if users actually forward/review reports in Telegram.
- Scope in: copyable Telegram summary, report path, limitations, disclaimer.
- Scope out: bot sending, private group scraping, signal parsing, Telegram auth.
- Entry criteria: >50% pilot users ask to receive/share summary in Telegram.
- Exit criteria: Telegram packet improves review/referral behavior.
- Main artifacts: delivery packet text, character-limit tests.
- Engineering tasks: deterministic formatting only.
- Validation tasks: track whether packet gets shared with coach/team.
- Risks: Telegram UX becomes signal analytics scope creep.
- Kill/pivot criteria: users still prefer PDF/Markdown or Telegram requests are about signal channels.
- Objective: turn paid manual audit into repeatable offer.
- Why now: after 3-5 successful paid audits.
- Scope in: offer page, intake checklist, data handling promise, pricing, report SLA.
- Scope out: self-service SaaS, subscriptions without repeat evidence.
- Entry criteria: 3 paid audits and at least 1 repeat intent.
- Exit criteria: 5-10 paid audits delivered with predictable operator time.
- Main artifacts: offer one-pager, intake form, sample anonymized report, privacy note.
- Engineering tasks: reduce operator time, not add features.
- Validation tasks: pricing, objection handling, coach/team package.
- Risks: service business without software leverage.
- Kill/pivot criteria: unit economics fail at prices users accept.
- Objective: decide whether live lockout is separate product, not sneak it into v1.
- Why now: only after audit users repeatedly demand pre-trade enforcement and pay for audit.
- Scope in: research-only, liability memo, read-only monitoring concept, API permission model, security gates.
- Scope out: order blocking, trade execution, broker credentials in product, kill switch implementation.
- Entry criteria: at least 5 paying audit users explicitly ask for live prevention and accept read-only first.
- Exit criteria: legal/security/product decision: reject, defer, separate product, or read-only monitor pilot.
- Main artifacts: ADR, liability memo, API risk map, user demand evidence.
- Engineering tasks: none until separate approval; maybe read-only paper prototype.
- Validation tasks: paid pre-order for read-only monitoring, not blocking.
- Risks: false sense of safety, missed block, false block, trading-loss liability.
- Kill/pivot criteria: users require trade permissions before trust/legal gates.
- Objective: choose next product path after evidence.
- Why now: prevent contamination from backtesting/signal/live-control ideas.
- Scope in: compare paid evidence for audit, signal sandbox, hypothesis bridge, live guard.
- Scope out: merging all ideas into one platform.
- Entry criteria: 6-week pilot data.
- Exit criteria: continue TRA, split product, pivot, or stop.
- Main artifacts: pilot evidence review, decision log, roadmap update.
- Engineering tasks: only after decision.
- Validation tasks: compare revenue, retention, referral, trust, implementation risk.
- Risks: founder follows most exciting feature instead of strongest paid behavior.
- Kill/pivot criteria: audit lacks repeat payment; adjacent wedge shows stronger paid behavior in separate validation.
Не использовать TAM/SAM/SOM. Use competitor pricing only as analogs: TradeZella/TraderSync paid tiers are around journal subscription behavior; Edgewonk lists $197/year; Tradervue has free/import-limited and paid tiers; TradesViz has free + paid all-in-one positioning. These are not proof of TRA pricing, only anchors.
| Offer | Price range | Included | Excluded | Validating payment behavior | Killer objections |
|---|---|---|---|---|---|
| One-time manual audit report | $49-$149 | One export period, written rules mapping, violation report, P&L attribution, limitations, review call | Ongoing journal, live API, backtest, strategy advice, Telegram signal analytics | User prepays or pays before delivery; sends export/rules within 24h | "I can do this in my journal", refuses data, wants free sample only |
| Monthly accountability audit | $99-$299/month | 2-4 audits/month, repeated patterns, month-over-month discipline trend, Telegram-ready summary | Live blocking, full dashboard, broker sync | User schedules next audit and pays month 2 | No second export, no behavior change, report not opened/shared |
| Team/coach package | $300-$1,000/month concierge | 3-10 traders, standardized rule policy, per-trader reports, coach/team summary | Enterprise compliance, SaaS seats, institutional SLAs | Coach/team lead pays for multiple traders after pilot | Privacy blocks exports, rules too vague, team wants full platform before paying |
What to do this week
- Stop product expansion.
- Contact 20 warm/semi-warm prospects.
- Run 10 past-behavior calls.
- Ask for real export + written rules.
- Sell 3 one-time manual audits at $49-$149.
- Deliver with current local-first deterministic workflow and manual operator review.
- Track every objection in a simple evidence log.
What to stop doing
- Stop building features before paid reports.
- Stop discussing live broker/API lockout as v1.
- Stop merging Telegram signal analytics into TRA.
- Stop pitching "AI trading strategy/backtest generator."
- Stop optimizing CLI/importer until real exports force it.
What to build next only after validation
- Support the top repeated export formats from paid pilots.
- Improve policy review packet for common rule wording.
- Add only validated rule types.
- Improve report packaging if users actually share it with coach/team/Telegram.
- Add repeat-audit workflow artifacts, not SaaS dashboard.
Single metric deciding whether project advances
3 paid audit reports from 10 qualified prospects within 14 days, followed by at least 2 repeat audit commitments within 30 days.
Most dangerous founder trap
Confusing technical correctness with market pull. The repo is already good enough to test the business; every extra feature before payment evidence is avoidance.
External sources used for current market, competitor, pricing, workflow, risk, and compliance claims:
- TradeZella pricing/features and explicit "not a brokerage" positioning: https://www.tradezella.com/pricing
- TradeZella trading journal feature positioning: https://www.tradezella.com/trading-journal
- TraderSync pricing/features, replay, AI coach, backtesting positioning: https://tradersync.com/pricing/
- Edgewonk pricing and feature list including risk/mistake/discipline analytics: https://edgewonk.com/pricing
- Tradervue trade import quota and manual import constraints: https://app.tradervue.com/help/quota
- Tradervue supported brokers/platforms and import limitations: https://app.tradervue.com/help/brokers
- TradesViz pricing, free plan, AI analytics/simulators/auto-sync positioning, not investment advisor disclaimer: https://www.tradesviz.com/pricing/
- TradingView official strategy data export docs: https://www.tradingview.com/support/solutions/43000613680-how-to-export-strategy-data/
- MetaTrader 5 official trading report/history docs: https://www.metatrader5.com/en/terminal/help/trading_advanced/history_report
- FTMO official trading objectives/max daily loss/max loss rules: https://ftmo.com/en/trading-objectives/
- Bybit official risk limit documentation: https://www.bybit.com/en/help-center/article/Risk-Limit-Inverse-Contract
- Moscow Exchange terminal docs, Excel/database export and risk parameters: https://www.moex.com/a1530
- Investor.gov definition of investment adviser: https://www.investor.gov/introduction-investing/getting-started/working-investment-professional/investment-advisers
- CFTC automated trading risk controls and safeguards concept release: https://www.cftc.gov/PressRoom/PressReleases/6683-13
- FINRA algorithmic trading supervision/control topic page: https://www.finra.org/rules-guidance/key-topics/algorithmic-trading
- Interactive Brokers API software license risk/liability language: https://interactivebrokers.github.io/
- SignalBlink Telegram signal performance analytics positioning: https://www.signalblink.com/
- Telemetrio Telegram analytics pricing and channel analytics positioning: https://telemetr.io/en/pricing
- X API enterprise pricing/custom pricing dependency: https://docs.x.com/enterprise-api/getting-started/pricing