Funnel profiles — group g2

Delivery/funnel profiles from the teardown corpus. One profile per vendor folder; heading is the folder name. Fields: product, surfaces, unit, cadence, mode, curation, free surface, paywall, conversion, actionability, trust record. Facts are from the teardown unless marked (live).

squeezemetrics

product: Market-structure data service — index DIX/GEX plus a per-ticker P/V/G/D measurement suite, spreadsheet downloads, and a keyed API.
surfaces: Web dashboard (under /monitor/), API, CSV/spreadsheet downloads, Yacht Club Sunday notes. No Discord, Telegram, or mobile app evidenced.
unit: A per-ticker state row / chart — P (price trend), V (volatility trend), G (gamma ratio), D (dark ratio), R, I, F, their *_NORM forms, and the P_NN nearest-neighbour forward-P estimate; the Research view adds a conditional 5-day return distribution (count, mean, median, 5th–95th interval).
cadence: Daily. Index series is an end-of-day file (public history 2011-05-02 to 2026-10-23 at teardown); Sunday notes are weekly.
mode: Pull — dashboard/API checked by the customer; the Sunday notes are the only push element.
curation: Algorithmic. Every measurement is a published formula; no human picks.
free_surface: Full public DIX/GEX index chart and downloadable CSV history since 2011, the vendor papers, and formula-complete Docs. No free live/per-ticker tier.
paywall: Login wall gates per-ticker Chart / Research / Ideas / P-scope tools, per-ticker downloads, and the API key. Single subscription covers web + API.
conversion: One plan, $720/month. Invitation-only enrollment, card required up front, $0 due today with first charge 72 hours after start; cancel in Billing before then and there is no charge.
actionability: Low direct actionability. The unit is a measurement/forecast, not a trade: no entry, stop, or target is given, and the customer must supply the interpretation and the decision.
trust_record: No performance claims published on the current site — the only return figure is a dated research pull-quote (high DIX associated with 5.3% vs 2.8% mean 60-day returns), graded UNVERIFIABLE. Docs explicitly disclose that historical files can be revised upstream and are not warranted as a point-in-time as-was record, and flag the shipped outcome labels as outcome labels. Honest disclosure, no auditable archive.

tapeboard

product: Browser market terminal — scanner, flagged options flow, GEX, Short Squeeze Score leaderboard, SEC filings, paper simulator and journal.
surfaces: Web terminal, documented public JSON endpoints / MCP actions (24 actions), push-notification endpoints. Broker connect (IBKR/Schwab) exists in the client for fills sync/journal.
unit: A ranked list row — principally a Squeeze Score (0–100 with its three percentile components and calibrated odds context), a flagged options contract row (volume, Vol/OI, premium, strike-vs-spot lean chip), or a GEX levels card (call wall, put wall, flip, total GEX per 1% move).
cadence: Squeeze leaderboard is a frozen daily record; flagged flow is checked every 15 minutes in market hours; scanner/flow/GEX surfaces are delayed snapshots, labelled as such.
mode: Primarily pull (terminal checked by the customer); custom formula alerts on the top tier are the push element.
curation: Algorithmic. Squeeze gate and weights, flow flag rule, and GEX definitions are all published rules; no human picks.
free_surface: Free tier $0 forever, no card; 7 days of Pro on every new account; public dated GEX/flow pages; documented no-auth endpoints (two verified live at teardown; three others returned forbidden from the test VM, so the "from anywhere" claim is PARTIAL).
paywall: Tier wall — Free vs Pro $39/mo ($374/yr) vs Terminal $79/mo ($756/yr) — gates scanner presets, backtest, simulator depth, and formula alerts. Pricing claims real-time Level 2 for Pro, which the vendor's own Data Disclosures refute (top-of-book only).
conversion: Free-forever entry plus a 7-day Pro trial with no card; Terminal trial requires a card and auto-renews. Annual billing discount. Funnel defect: Terms state trials convert to paid unless cancelled, pricing states the customer moves to Free on day 7 unless they subscribe — opposite mechanics in the vendor's own documents.
actionability: Medium, as context. Scores and flags build a watchlist; the vendor states repeatedly that the Squeeze Score is not a directional call, and no entry/stop/target is given. The paper simulator is where a unit can be practised.
trust_record: Strongest in this group. Scores are frozen into dated daily records and were not re-scored when the formula changed (v5 to v6); the methodology publishes an out-of-sample odds table (top-10: +20% in 5 sessions 16.6%, −20% 12.0%, median 5-day return −1.8%) with the negative median printed against its own product. Defect: the leaderboard explainer still carries the retired v5 weights (70/30) against v6 (60/20/20) on the same page.

optionstrat

product: Options strategy builder/optimizer plus a Live Flow scanner.
surfaces: Web app; saved/shareable strategy links and public strategy screenshots. No order placement — explicitly not a brokerage.
unit: A strategy card — legs, P/L profile sampled across price, breakevens, max profit/loss, net Greeks, chance of profit — or a ranked optimizer result set, or a flow item (type badge, sentiment, premium, reconstructed strategy).
cadence: Chain-driven: delayed by default, live/pro feeds on paid tiers; Live Flow streams over WebSocket in real time on the Flow tier.
mode: Pull for builder/optimizer (customer supplies ticker and view); push/stream for Live Flow.
curation: Hybrid by design — algorithmic enumeration and ranking driven by a human-supplied sentiment/target. The optimizer ranks candidates by a min-max normalised ROI/chance composite set by a user slider.
free_surface: Free $0 tier with delayed data and a limited builder; chance-of-profit and the distribution chart render as an upsell ("--%") for free users. Public tutorials and 58 pre-made strategy pages.
paywall: Feature/data wall in two steps — Live Tools $39.99/mo ($419.99/yr) gates live chain, chance, net Greeks, IV history, commissions and saved tracking; Live Flow $99.99/mo ($999.99/yr) gates paid flow detail, saved filters, alerts, performance tracking, and historical search. Live data also requires the OPRA/Cboe subscriber agreement.
conversion: Both paid plans carry a 7-day one-time trial; annual discount; auto-renewing with no full or partial refunds (stated plainly in the FAQ).
actionability: High for construction, none for selection. A unit gives exact legs, prices (mid by default), breakevens and risk figures a customer can place elsewhere, but no recommendation of what to trade and no follow-through record of what a built strategy did.
trust_record: No win-rate, return, or performance claim published anywhere audited — capability copy only. Flow performance is tracked forward from posting per the vendor, but no as-was archive is published, so history integrity is UNVERIFIABLE server-side. Probability outputs are materially simplified (driftless chance figure; the chart distribution is a bell-shaped proxy, not a density) and are cushioned only by a general "calculations are estimates" disclosure.

marketchameleon

product: Broad subscription options/equity analytics — screeners, unusual-activity and order-flow reports, historical theoretical-value analytics, earnings-event studies, and backtesters, plus a broker-account viewer.
surfaces: Web dashboard (server-rendered reports, downloads), in-product chat assistant, broker-linked account views (Interactive Brokers / TradeStation). No community, mobile, or API surface evidenced.
unit: A report/screener row or a featured-idea card showing market price vs theoretical value, Theoretical Edge %, and historical win rate together; the flow-alert variant is an LLM-written alert with flash headline, trade line, key-metrics table (HFV, Value Edge, win/touch rates, aggression, bid/ask lean), and a Signal Clarity line.
cadence: Intraday reports and screeners (Terms concede quotes may be delayed at least 20 minutes); earnings studies are valued from end-of-day snapshots over at most the last 12 events.
mode: Pull — a reports library the customer queries; no push alert product is the core offer.
curation: Algorithmic statistics with an LLM narrative layer for flow alerts. Featured ideas are ranked by an edge computed from full-history distributions.
free_surface: Basic access with truncated gates — next 3 business days of events, large trades only at quantity 200+ on the current day, 2 downloads per 24h on trial.
paywall: Depth/history wall — Premium/Total Access $99/month unlocks all event dates, large trades at 50+ over 10 trading days, 60 days of per-symbol trade history, up to 3 years of price downloads, and 25 downloads per rolling 24h. A four-tier ladder ($39/$69/$79/$99) is described consistently by third-party reviews but only the $99 anchor is vendor-confirmed (PARTIAL).
conversion: Free first week, then $99/month, auto-renewing; download caps differentiate trial from paid; fees are not refunded on customer-requested termination per Terms.
actionability: Medium. Screeners and per-contract edge/win/touch figures are directly filterable and comparable, but no entry/stop/target or position follow-through is given; the customer still constructs the trade. The headline edge needs care: it is a historical-expiration average, so deep out-of-the-money options mechanically show huge edges at near-zero win rates — the homepage itself shows a 299.1% edge beside a 0% win rate.
trust_record: No product-level win-rate or return guarantee published; win rates are per-contract historical frequencies and are displayed even at 0%. No as-was archive or methodology note establishes point-in-time integrity, survivorship/delisted handling is unaddressed, and backtester fill/slippage conventions are unpublished — server record UNVERIFIABLE.

gammaedge

product: A Discord-centred GEX trading community with frameworks (Wheel, P-Trans+GEX swing) and a gated Terminal v3 web app, sold via Whop. Note: the teardown's main site, joingammaedge.com, is a third-party affiliate funnel for this product, not the operator's site.
surfaces: Discord (the product substantially lives there, including bots/tools), sign-in-gated web app on the operator domain, Whop storefront, public YouTube recap videos. The affiliate site is marketing only.
unit: A GEX levels set (Call Wall, Resistance, Key GEX, Support, Put Wall, plus GEX/DEX/Max Pain figures) and a flow table row with a hardcoded-style bias label in the public preview; in the paid product, framework calls inside Discord.
cadence: Claimed intraday — GEX updated every 15 minutes, flow near-real-time (under 30s), dark pool aggregated every 10–15 minutes (FAQ structured data only; UNVERIFIABLE), plus daily recap videos.
mode: Push — a community/Discord delivery model; the customer is interrupted in-channel rather than checking a terminal.
curation: Hybrid — algorithmic GEX/DEX levels and a cumulative-tick Market Trend Model gate, with human operator frameworks and recaps on top.
free_surface: The affiliate site's Terminal preview is a labelled sample/mock (literal price/flow/GEX arrays and a sine-generated chart — it computes no market data, and says so in its own UI); public YouTube recaps. No free product tier evidenced.
paywall: Community wall at Whop checkout — Discord access and the web app sit entirely behind payment/trial.
conversion: $150/mo, or $125/mo billed as $1,500/yr; 14-day free trial via Whop; every buy link carries an affiliate parameter. A "make $150 in your first month or full refund" guarantee is repeated in marketing copy, but the funnel's own Terms contain no refund clause and disclaim all warranties — the guarantee's obligor and conditions are UNVERIFIABLE. Social proof is stale: funnel shows 4.94/5 across 75 reviews; the live Whop listing at teardown showed 4.93 across 81 ratings.
actionability: Levels give a tradeable support/resistance reference and the frameworks describe tenor and universe (short-tenor cash-secured puts on trend-template, call-dominated names), but every load-bearing threshold (trend gate, "call-dominated", strike selection) is sealed or undefined, so a unit is not executable without the operator's ongoing calls.
trust_record: Hindsight/video display only. Both headline records — Wheel 80%+ win rate with 2.5x profit factor, and P-Trans+GEX 72–73% over 334 trades since July 2023 — rest solely on operator YouTube videos, with no trade list, denominator, win definition, or returns published; both UNVERIFIABLE. No point-in-time archive of levels is published, so recaps cannot be audited as-was.

levelfields

product: Event-driven stock screener/alerts — 24 catalyst scenarios (buybacks, CEO exits, insider buys, activist stakes, government actions) over about 6,300 companies — plus a separate wealth-manager portfolio tracker.
surfaces: Web app, email alerts, SMS alerts (Level 2), separate portfolio web app.
unit: An event alert / scenario table row: the detected event plus per-event forward returns at 1D–4D, 1W–3W, 1M–12M horizons, under scenario statistics (win rate, average 1/10/30-day returns, comparison vs the S&P 500). Level 2 adds human analyst trade ideas.
cadence: Real-time alerts claimed (seconds after a filing or headline — UNVERIFIABLE); statistics are recomputed over whatever date window the user selects.
mode: Push for alerts (email, SMS on Level 2) plus pull for the screener.
curation: Hybrid — algorithmic detection that, at the only visible layer, is CMS keyword search terms (one or two search terms plus an exclusion term per scenario), with human analysts supplying the Level 2 ideas. No ML model, data vendor, or method is named anywhere.
free_surface: No free app trial — disclosed honestly ("data once seen can't be unseen"); instead videos plus one free email alert per week.
paywall: Whole-app subscription wall. Level 1 $299/yr ($25/mo annual framing); Level 2 gates SMS alerts, analyst-called ideas (1–2 per week) with entry/exit targets, and training. Monthly Level 1 and current Level 2 prices ($167/mo / $1,599/yr per directory listings) are not printed on current vendor pages — PARTIAL.
conversion: Annual-first pricing with monthly deliberately much more expensive (the FAQ explains why at length without stating the price); no pro-rated refunds, annual refund requestable at term end; skeptics are offered a "live Trade Log — wins and losses included" by email — an email-gated conversion asset that was not obtained.
actionability: Split by tier. A Level 1 alert is context (event plus history) and the customer decides everything. A Level 2 idea carries entry/exit targets, the most actionable unit in this vendor — and it is the human layer, not the screener.
trust_record: The headline win rate is a 1-day direction hit rate over every event in a scenario — no fills, costs, or benchmark in the rate itself — recomputed per user-chosen window, with no as-was archive and no published sample sizes. Homepage "recent alerts" show five winners only; case studies (+82% in 7 days, +350% in five months) are UNVERIFIABLE anecdotes; the widely quoted ~80% figure does not appear on the current site. One genuine hygiene positive: return cells are horizon-gated in the client and never display before the horizon has elapsed.

edgeful

product: Historical-probability statistics platform — 150+ session-statistic reports, a What's In Play live dashboard, a daily-bias screener, natural-language AI Q&A over the reports, charting-platform indicators, and 7 automated algos on the top tier.
surfaces: Web dashboard, key-gated API plus MCP (API-key and native OAuth connector routes), TradingView and NinjaTrader indicators, algo execution via Tradovate/NinjaTrader/ProjectX.
unit: A report card — an event frequency with sample size and a per-day detail table (e.g. gap-fill %, IB/ORB break and extension rates) — or a Discovery ranked setup (0–100 score with raw hit rate, sample size, and trend), or a per-ticker screener bias %.
cadence: Reports are computed over completed sessions bounded by explicit dates (daily/pull); What's In Play and the screener stream live through the session.
mode: Pull for reports/API; push/stream for the live surfaces.
curation: Algorithmic. Report definitions are fully published in the vendor's own OpenAPI docs; the two layers that convert frequencies into decisions — the Discovery 0–100 score and the screener bias % aggregation — are sealed.
free_surface: No free product tier evidenced; the free surface is the fully public documentation/OpenAPI, worked examples with sample sizes, and a newsletter signup. The limited API (4 tickers, 3 reports, 6 months history) sits inside the paid Essential plan.
paywall: Entitlement wall enforced in the API with named 403 codes (report/ticker/history/live/detailed-row gates). Essential $49/mo ($399/yr); All Access $299/mo ($2,399/yr). An API Pro plan ($99/mo, $799/yr) is published in the vendor blog but does not appear on the pricing page — PARTIAL.
conversion: Annual-discount tiering, an affiliate programme, and upsell from statistics to automation (All Access algos). Platform fees are non-refundable except as required by law. Vendor pages disagree on history depth (7+ vs 8+ years).
actionability: Context, not a trade — frequencies with sample sizes tell the customer how often something happened, with no entry/stop/target, except on the algo tier where the unit escalates to automated execution. The structural trap is named by the design itself: 157 reports times subreport partitions guarantees impressive-looking high-frequency cells in-sample, and no holdout discipline is claimed.
trust_record: No platform win-rate or return guarantee published anywhere — the only percentages are per-report historical frequencies printed with sample sizes, the cleanest claim surface in this group alongside the API documentation. Live What's In Play states have no as-was archive, and the algo optimizer ranks over the same window it reports (the vendor concedes its top-P&L result is not always its top-ranked result).

altindex

product: Alternative-data AI Score platform — a daily 0–100 score per stock blended from Audience, Customer, Employment, and Fundamental pillars — with AI picks, API/MCP data plans, and rules-based Autopilot copy portfolios.
surfaces: Web app, public server-rendered ticker pages, developer API, hosted MCP, brokerage-linked portfolio (via an aggregator; Autopilot executes on the parent company's platform).
unit: A score card — AI Score with four sub-scores and about 25 signal rows per ticker, banded Strong Buy (81+) / Buy (60+) / Hold (40+) / Sell (31+) / Strong Sell — or a pick, or a value-weighted Portfolio Score for the customer's holdings.
cadence: Daily refresh. The app polls (5-minute stale time, no WebSocket); it is a daily-score dashboard, not a real-time product.
mode: Pull — dashboard, ticker pages, and API are checked/queried; alerts exist in the app but are not the core delivery.
curation: Algorithmic with sealed weights — factor definitions ship in the client, the per-factor formulas and the blend do not.
free_surface: Public ticker pages showing scores and signal tables, public leaderboard, and a free account (the earlier "no free tier" note is now stale).
paywall: Breadth wall — Starter $29/mo ($199/yr) caps the watchlist at 25 stocks, 1 brokerage connection, top-25 picks, and a 3-filter screener; Pro $99/mo ($499/yr) unlocks unlimited watchlist, all 200 picks, full screener/export, and API/MCP. A separate API ladder runs Hobbyist $19/mo, Developer $79/mo, Research $199/mo, Enterprise custom.
conversion: Steep annual discounts, per-tier pick counts as the upsell lever, and a 14-day money-back guarantee advertised on pricing — directly contradicted by Terms stating all payments are final and non-refundable "without exception" (REFUTED as consistent). User counts also conflict across the vendor's own pages (80,000+ vs 50,000+).
actionability: A score band is a labelled opinion (Buy/Hold/Sell) with no entry, stop, target, or sizing. The exception is the Autopilot Top-10 rulebook, published in full (universe filters, rank by score, top-10 equal weight, weekly review, delayed entry and rank-collapse exits) — actionable as a system, executed on an external platform whose performance is UNVERIFIABLE here.
trust_record: Contradictory on the vendor's own pages — 61% win rate (home), roughly 70% with ~22% average return (picks page), and 0% on the results page it links to, whose rolling 6–18-month window contained exactly 2 picks (1 loss at −38.6%, 1 flat). To its credit the results page prints the loser unhidden and states an as-was contract (timestamped picks, entry at pick-date close, losers not deleted), but the window is vendor-chosen, statistically empty, and unauditable; showcased winners (+137%, +103%, +85% "since pick") are UNVERIFIABLE, and whether historical scores are vintage or regenerated by the current model is nowhere stated.

optionalpha

product: No-code options autotrading platform — user-built rule bots that execute at the customer's broker, a 0DTE/next-day backtester, a Trade Ideas scanner ranked by expected value, and SmartPricing order laddering. Not a broker and not an adviser, stated plainly.
surfaces: Web app (bot editor, logs, backtester), broker APIs (Tradier, TradeStation, tastytrade, Schwab), webhooks and TradingView/Python/Zapier trigger paths, community bot templates.
unit: The unit is an acting bot, not content: an automation recipe tree that scans/monitors every minute and opens, manages, and exits positions under the customer's rules. Secondary units are a Trade Ideas ranked row (Reward/Risk, EV, Alpha = EV / Max Loss) and a backtest result.
cadence: Real-time automation in market hours (1-minute scan/monitor cadence, cloud-hosted, no login required); Trade Ideas scans 9:35–15:55 ET.
mode: Push by execution — the bot acts on the market without the customer present; the customer is the audience for logs, not for alerts.
curation: Human-authored rules executed algorithmically; Trade Ideas ranking is algorithmic over a published EV method (full-payoff expected value under a published historical-volatility probability model); community templates are human-shared.
free_surface: About 100 pages of public documentation publishing the probability/EV mathematics and the full SmartPricing ladder, example bots with complete rule sets, and a $0 subscription route via qualifying broker accounts (see conversion).
paywall: Capacity wall, not a signal wall — Pro $99/mo on annual billing ($149 monthly) caps at 50 bots and $100k per bot; broker-funded tiers carry identical bot/backtest limits.
conversion: The funnel is broker partnership: $0 via a funded Tradier ($5k fund / $3k maintain), TradeStation ($10k), or tastytrade ($5k) account, disclosed on the pricing page itself. Community "Top Backtests" discovery (pre-filtered by in-sample win rate) drives template adoption. Refund policy is sharp: no refunds on monthly plans; annual refundable only within 7 days of a renewal, not in year one.
actionability: Highest in this group. The delivered unit executes entry, profit target, stop, and adjustments at the broker, with SmartPricing working the order (start at mid, timed steps toward a bounded final price). The customer supplies the thesis and the rules; nothing needs further interpretation at trade time.
trust_record: None published, by stated design — the FAQ argues bot performance cannot be objectively quantified (customisation, toggling, manual overrides) and no win rate, return, or equity curve appears anywhere on the site, marketing included. The backtester is a mid-price fill model on 1-minute data, and the vendor publishes an eight-reason page on exactly why live bots will diverge from it — the most candid limitations disclosure in this group. Residual caveats: "top performing" discovery ranks in-sample backtests with no holdout discipline, and homepage usage counters ($ capital allocated, transaction volume, active bots) are UNVERIFIABLE.

optionality

product: A four-desk Discord trading community (options / day / swing / futures) whose Pro tier adds the Sinux Signals dashboard — a white-label trade-tracking ledger for the analysts' calls, not a signal engine.
surfaces: Discord (primary delivery), web dashboard (white-label), push notifications from the dashboard.
unit: An analyst alert call in Discord free text — entry/average stated by the analyst and managed in the open with adds, trims, cuts, and closes — rendered in the dashboard as a position row with live percentage P&L and its full event timeline. There is no signal generation anywhere: an AI layer only parses the human messages into the ledger.
cadence: Real-time — calls post live in session; the dashboard overlays live quotes polled every 5 seconds where Live Data is enabled.
mode: Push — Discord alerts interrupt the customer; the dashboard is the pull-side record of the same calls.
curation: Human. Named analysts pick and manage every call; curation is the product.
free_surface: Free Discord server, $0 forever, no card — community, daily market context, and education, with the signal channels marked Pro.
paywall: Channel wall — Pro gates the paid signal channels (14 channels marked FREE/PRO across the server map) and the dashboard.
conversion: $1 for 7 days, then $50/mo auto-converting unless cancelled; annual $450/yr. The free server is the try-first funnel. No refunds under any circumstances — the harshest term in this group, but printed in Terms, FAQ, pricing, and a support page, next to the price.
actionability: High as a called trade — entries are called live and managed in the open until they close, so follow-through is part of the unit; there is no separate standardised stop/target field, management updates are the exit mechanism. Dashboard P&L is percentage-only by design (position size is never learned).
trust_record: A tracking ledger, explicitly not an audited track record — the vendor says so in its disclaimer and FAQ, publishes no win rate, labels its homepage sample alert as sample data, and stamps testimonials as not typical results. The shipped scoring code is engineered against flattery: unpriced closes are excluded from every rate, a ±0.05% even band stays out of the win rate, peaks are computed per cost-basis stretch so averaging down cannot inflate them (and refused entirely for shorts/spreads), closed cycles cannot be reopened, and giveback is surfaced in recaps. Structural limit, disclosed by the vendor: the basis is analyst-stated prices parsed from chat, not broker fills.

tradeecho

product: All-in-one options-flow terminal — OptionFlow tape, DealerEdge GEX/vanna/charm levels, dark-pool levels, Discovery setups, AlgoEdge channels, Cortex AI agents, copy trading, and a trading-capable MCP server.
surfaces: Web app, iOS/Android apps, public Daily Echo Brief, MCP server (OAuth or personal token), broker connections for copy trading / Cortex / MCP execution.
unit: A flow print row (Flow Score −100 to +100, tiered Golden / Unusual / Aggressive by published thresholds), a DealerEdge levels set (anchor, flip, defense lines, regime), a Discovery pick held in a daily ledger (first-seen, latest, and peak score), or the Brief's per-index pre-market snapshot card.
cadence: Real-time streaming in regular hours (SSE; quotes polled every 2.5s for live GEX repricing) with published snapshots at 08:00, 16:05, 18:00 ET and a Saturday refresh; Discovery rescans about every minute; the Brief is daily pre-market.
mode: Push/stream inside the app, plus pull — the Brief is a daily digest a non-member can read without an account.
curation: Algorithmic scores and levels with an LLM agent layer (Cortex) on top. Flow Score weights, the 1–5 GEX Rating formula, Discovery scoring, and AlgoEdge classification are server-sealed; tier thresholds and Discovery eligibility rules are published exactly.
free_surface: Free $0, no card — Daily Echo Brief (also public), 5 Cortex chat messages per rolling 24h, MCP limited to the Brief plus SPY/QQQ morning levels (3 briefs/day). No free trial of the paid plan.
paywall: A single subscription wall — the vendor's central claim is that no module is plan-gated; Monthly $199, Yearly $2,199, Lifetime $3,499 differ only in capacity (active copy-trading pilots and live Cortex agents; trade-plan caps).
conversion: Annual saving framed exactly ($189 vs monthly; arithmetic verified), Lifetime framed against two years of monthly ($4,776; verified), affiliate/referral programme, competitions/prize pools. Refund conflict: Terms say non-refundable except where required by law; the standalone Refund Policy says no refunds under any circumstances — material at the $3,499 tier.
actionability: Levels and flow are context requiring interpretation, and Discovery is explicitly stated, twice, to rank move potential and not be a win probability. Actionability escalates inside the same subscription: Cortex trade plans, copy trading, and MCP trading can place real orders in linked brokerages, gated by previews, approvals, per-agent limits, and a drawdown kill-switch (consent plumbing verified in the client; enforcement server-side UNVERIFIABLE).
trust_record: Mixed, with unusually good labelling. The Daily Brief is a genuine point-in-time public artefact (the dated pre-market snapshot as published that morning); the Discovery ledger stamps first-seen vs peak; Wins and leaderboards are member-submitted Discord screenshots ranked by Discord-posted prices, not broker fills — and the vendor's results-sourcing page says exactly that, with results labelled not independently audited and not typical. Against that: every headline score is a sealed server model with no public backtest or as-was archive, one marketing absolute ("computed, not modeled — nothing is estimated") is refuted by the vendor's own methodology (thin-name Greeks are estimated, OI-missing contracts fall back to volume), and the MCP tool count does not reconcile (51 claimed vs 43 in the live public catalog).

tradeideas

product: Real-time US equity scanner platform (TI Pro, Windows desktop) with Holly AI signals, the OddsMaker event-study backtester, and Money Machine auto-trading.
surfaces: Windows desktop app (the core product; Mac via Parallels), a limited web dashboard, broker routing via Brokerage Plus (TradeStation, CenterPoint, Cobra named). No public API — stated directly in the vendor FAQ.
unit: A streaming alert or Top List row (re-ranked every 15 seconds), or a Holly signal with an entry and a stop (usually about 3% from entry, per the homepage FAQ) plus a re-entry point if a stopped-out stock turns back.
cadence: Real-time — alerts fire on individual trades, not on bar closes.
mode: Push — desktop alerts interrupt the day trader; Top Lists are the pull complement.
curation: Algorithmic selection over a human-built library — Holly is a nightly tournament that backtests and optimises a fixed proprietary strategy library, then selects which strategies trade the next session. Parameters are not exposed or customisable at any tier.
free_surface: The web dashboard subset (Top Lists, Races, TreeMaps, Market Pulse, stock detail/news, an Unusual Options Activity channel) and the public Holly Records page. Per the vendor's own comparison table the dashboard excludes Streaming Alerts, OddsMaker, Holly, the AI Assistant, and Money Machine.
paywall: Tier-plus-platform wall — Core $127/mo ($1,068/yr) vs Apex $254/mo ($2,136/yr), with Holly/OddsMaker/Money Machine Apex-only and desktop-only.
conversion: Annual savings framed on the pricing page ($456/yr Core, $912/yr Apex; arithmetic reconciles), "cancel anytime" copy for monthly billing in tension with an EULA stating no early termination and no refunds.
actionability: A Holly signal is actionable without further interpretation — entry, stop, and re-entry are given at fire time — and Money Machine can auto-execute the customer's own lists with per-stock stop/target.
trust_record: Refuted as a track record by the vendor's own footnote. The public Records page (about 404 signal cards at teardown) shows selected examples scored as "profit potential" — signal price to the session's subsequent high (longs) or low (shorts), i.e. the best possible exit, not realised profits, fills, or a complete ledger. No win rate is published by the vendor (a 60%+ / 2:1 figure circulates only in third-party review FAQs), and the nightly re-optimisation means any performance drawn from the process is restated selection history by construction.