Product Roadmap
Where Grade My Investments is headed — published openly because a platform's direction is part of what you're evaluating. A living document; order and scope evolve with what users teach us
flowchart LR
subgraph DONE[Shipped foundation]
D1[Reports + ML grading]
D2[Billing + free tier]
D3[4 client surfaces]
end
subgraph NOW[Live growth layer - Aug 2026]
G1[2,500+ public grade pages]
G2[Report sharing +<br/>verbatim dashboards]
G3[Social card pipeline]
MACRO[MACRO ADVISOR<br/>regime, grade, playbook,<br/>ML forecasts]
end
subgraph BUILT[Built Sep 2026 - in final testing]
ONB[Guided first report]
FAIR[Grade fairness +<br/>style lenses]
EARN[Earnings Setup +<br/>GMI's Read]
REC[Recurring reports +<br/>Claude comparison]
TOOLS[Watchlist, benchmarking,<br/>screener]
LEG[Investor Legends +<br/>Smart Lists]
ALERTS[PAY-AS-YOU-GO ALERTS<br/>metered, capped]
CARD[Forecast report card]
end
subgraph NEXT[Next - the open seams]
N1[Push, ML-health,<br/>tripwire alerts]
N2[Screen entry and<br/>exit alerts]
N3[Transcripts, implied move,<br/>measured prior]
end
DONE --> NOW --> BUILT --> NEXT
FAIR -. formula guard kept the<br/>regrade from alerting .-> ALERTS
LEG -. change log shaped<br/>for alerts .-> N2The Flagship: Macro Advisor SHIPPED
GMI answers "how healthy is this company?" The Macro Advisor answers the question that sits above it: "what kind of economy is this, and what tends to work in it?" It shipped in August 2026 — regime and Macro Grade, the Regime Playbook, Sector Grades, the Macro Lab, Commodities & Hard Assets, the release calendar, and machine-learning projections for both the economic series and the commodities, in all three applications.
The vision: pull the government's own economic firehose — employment and payrolls (BLS), GDP and income (BEA), inflation, rates and yield curves, housing, industrial production, consumer sentiment — the myriad of official series the pros watch and retail rarely sees assembled. Then do what GMI does best: turn hard data into a legible roadmap. Where is the cycle? Which regimes do these indicators resemble? Which sectors and styles have historically done well from here — and which of your graded holdings sit with or against that tide?
It's the same architecture philosophy as everything on this site: authoritative data with provenance, deterministic-where-possible analysis, AI for synthesis and explanation — and the result connected directly to your portfolio, not left as an abstract chart. Company grades told you what to look at; the Macro Advisor will help with when and where.
Built in September 2026 IN FINAL TESTING
Ten roadmap waves plus the macro forecast report card were built on parallel branches, merged, and deployed to production for final testing on 29 September 2026. Waves 12 (price & technical alerts) and 22 (tracked-symbol watch) were built inside Wave 24 as event types, and Wave 15 (earnings calendar & events) turned out to be covered by the Earnings Lab, Wave 21 and Wave 24. Each feature is on all three surfaces (web, admin, mobile) with its own tests and a design doc. They're listed as built, not shipped. Shipping is a separate step that comes after final testing.
| Wave | Feature | What was built |
|---|---|---|
| 20 + 23 | Guided First Report + positioning line | "Pick a company you know" builds one real report in a tap; four first-view callouts; a public walkthrough over a dated, real grade; a GA4 funnel. The home page now reads "For investors who read the numbers, not the headlines." |
| 19 | Grade fairness | Formula 1.1: debt-to-equity and current ratio score a neutral 0 for Financial Services; value and growth style lenses sit beside the grade and never feed it |
| 21 + 21b | Earnings Setup | Per-symbol reaction table, beat habit and base rates with sample sizes and an 8-print floor, plus a daily consensus-estimate journal. 21b added the run-up chart and GMI's Read, whose first version failed a walk-forward backtest and was rebuilt |
| 24 | Pay-as-you-go Alerts | Email and SMS alerts metered per message with a spend cap enforced inside the ledger transaction; a free catalogue on by default; five compliance gaps closed first. Absorbs Waves 12 and 22. Negative-news alerts ship switched off |
| 25 | Investor Legends + Smart Lists | Eight screens from investors' published criteria with per-criterion pass/fail/not-evaluated results; any screen saved as a self-refreshing list that logs why each symbol came or went |
| 11 | Recurring reports + Claude comparison | Weekly to quarterly series on top of one-shot schedules; automatic change analysis; one combined notification; charges-apply consent |
| 13 | Watchlist dashboard | Every tracked symbol in one grid (price, RSI, 52-week position, grade, sparkline) with no new data feed |
| 14 | Portfolio benchmarking | Position reports vs. SPY and other index ETFs, labelled a "held-through" comparison, in the app and in the Excel download |
| 16 | Stock screener | Filter the graded universe from an in-memory snapshot; built-in presets that state their numbers |
| — | Forecast report card | Every ML projection recorded when it's made and graded after its month prints: band calibration, direction, and skill over "nothing changes" |
What's Next
| Wave | Feature | What it does |
|---|---|---|
| 24 tail | MAUI push, ML-health and regime-tripwire alerts | The three pieces of the alerts design not built: push as a near-zero-cost third channel, an event for ML health flips (needs per-symbol ML output in the grade cache first), and holdings-vs-regime tripwires |
| 24 + 25 | "Entered or left your screen" alerts | Smart List change rows are already shaped as alert source records; what's missing is the detector that reads them |
| 21 seams | Call transcripts, options-implied move, a measured prior | The first two wait on a probe of the market-data contract. The third replaces GMI's Read's flat 50% prior with a cross-sectional rate measured on real cached symbols |
| 25b | Screener follow-ons | Nightly universe-wide prices, sector-relative filters, a macro-fit filter, and, only with a survivorship-bias disclosure, screen backtesting |
| 17 | Broker integration | Read-only holdings sync so positions arrive without CSVs. Not built, deliberately: it needs broker partnerships and API agreements before any code |
| 18 | Cryptocurrency coverage | Extending the grading discipline to digital assets. Not built, deliberately: it needs confirmation that the data contract licenses crypto data for display |
Order reflects current thinking, not promises. Waves have been reordered before when reality called for it: report sharing and social cards jumped the whole queue in a single week of August 2026 because the growth need was immediate, and two live demos in September moved onboarding to the front. What won't change: the data-provenance, transparency and lean-economics principles every wave has to meet.
Shipped — the Track Record
Ten planned waves delivered, plus the unplanned ones reality demanded: position reports · symbol-list templates · Claude AI analysis & report comparison · ETF/mutual-fund holdings import · HTML reports · blog & content system · contextual help & UX polish · the AI Report Dashboard · admin billing queue & history · auth/ownership hardening — then the free tier, the 2,500-page public grade universe, report sharing with verbatim dashboards, and the social card pipeline.
The delivery pattern to judge us by: features here ship in days-to-weeks, fully across all three surfaces (web, admin, mobile), tested, documented on this site, and dogfooded on real money before they're announced.