How it works

We run two products that work together. The screener narrows ~4,500 US stocks to a quality-vetted ~150. The analyst goes deep on individual names with AI-generated equity research briefings. Each is priced for what it actually costs to run.

A typical session: use the screener to scan the quality universe and find what's in today's Buy Zone. Spot a name that catches your eye. Open the analyst for a deep dive — view existing analyses for free, or pay for a fresh one if you want today's take.

• Durability over snapshots
  Most screeners filter on TTM ratios. We require 10+ years of consistent performance — revenue growth, EPS growth, ROIC above WACC, positive FCF, manageable debt. We screen for companies that have *proven* the business model, not ones that look good this quarter.
• Five archetypes, five different rulers
  A toll-road business should not be judged on growth rate. A cyclical should not be judged on current-quarter margin. We classify each stock into one of five business archetypes (Reinvestment Growth / Toll Road / Mature Compounder / Cash Cow / Cyclical) and score it on metrics that actually matter for that kind of business.
• Normalized valuation
  Buy Zone uses each stock's own 5-year history, not industry averages. A P/E of 20 might be cheap for one stock and expensive for another. Cyclicals get an additional guard: operating margin must be below 10-year average before they can enter the In or Deep zone — kills the classic 'cheap P/E on peak earnings' trap.
• Daily price gate on top of monthly quality
  Quality changes slowly. Price changes daily. We refresh quality monthly (deliberately — it teaches discipline) and re-grade prices every weekday after market close. The cadence is a feature.

• Eight purpose-built sections
  Identity, leadership, business story, financial snapshot, valuation picture, reverse DCF, archetype deep dive, risk diagnosis, entry verdict. Each section answers a specific decision question, not a generic summary.
• Real-time data and web search
  Every analysis pulls fresh financial data from 5 providers (FMP, Finnhub, Tiingo, Alpha Vantage, Yahoo) and lets Claude search the web for context — news, analyst grades, insider transactions, recent filings.
• Reverse DCF that reveals priced-in growth
  Most DCFs project a growth rate and tell you 'fair value.' Ours flips the math: takes the market price as a given and reveals what growth rate is baked in. Compare to historical CAGR — is the market optimistic, pessimistic, or fair?
• Cached and shared
  Once any subscriber pays for an analysis, every subscriber can read it free, with a chronological view of past briefings. The community benefits from the spend.

We've seen what other research tools charge for AI features — usually a flat $20-$30/month with vague "up to N analyses included." That model has two failure modes:

1. Light users overpay. They pay $30/month and run 1 analysis a year. The flat price subsidizes power users.
2. Heavy users get throttled. Once they hit the "included" limit, the tool either nags or silently degrades quality.

We chose a cleaner model: charge a flat subscription for things that have a predictable cost to run, and a per-use price for things that don't. The screener and ticker detail pages cost the same whether one user or a thousand are looking at them. AI analyses are different — each one is a fresh request that takes real time and real money to run.

Want to see what the full Universe → Buy Zone → Deep Analysis flow looks like with real numbers? Walk through a customer journey with Microsoft as the running example.