dibsresearch Proprietary Data

Our own data, read from the original sources.

We take our data from the original sources — filings, company releases, official statistics — check every figure and keep every version. On that ground we build our own series, reports and charts.

  1. Collect: SEC filings, Company releases, Official statistics, Public registers
  2. Read: From the original source, by code
  3. Check: Every value, before it is served
  4. Keep: With its source and every version
  5. What it becomes: Facts from filings, Company KPIs, Our own series, Reports & charts
  • 2,700+ original filings and releases, kept as filed
  • 160,000+ tables read from them
  • 350,000+ company KPI values, each with its source
  • 5,700+ statistical series, 1,800+ of them calculated by us

Counted in our database on 1 October 2026, the company KPI values on 17 October 2026. Filings and releases of the S&P 100.

From the original document to a figure you can trust

Collect

Filings, company releases and official statistics come from the publishers themselves. Each filing is kept exactly as it was filed.

Read

Code reads the figures from each source, with their period and the exact place they come from. No figure is typed by hand or written by a model.

Check

Every value goes through checks against the rest of the record before it is served. A value that contradicts them is held back.

Keep

Every value keeps its source, its check status and each version that was filed, so it can always be traced to where it came from.

Every version, kept

Companies restate their past — after a spin-off, a merger, a change of method. We keep every version of a figure, show the one you ask for, and compute growth only within one filing, so a restatement never shows up as a collapse.

Four kinds of data, one standard

Facts from filings

Every figure comes with its fact card: the filing it was read from, its period, its versions and its checks. One click opens the original at that spot.

Company KPIs

The figures companies report in their earnings releases — subscribers, segment revenue, regions — under their own names and followed across quarters. A value is marked verified once an independent reading confirms it.

Our own series

Company-level data nobody hands you, calculated from government registers and other public data, with every revision kept. In the search they appear as Data Sheets.

Reports & charts

Economic Reviews, Industry Reports, Earnings Reviews and Chart Notes, written on the library. Every figure is checked before publication; pieces our AI process writes are marked AI-generated.

Fact cards, reports and charts are for signed-in readers.

Trust is built into the data

Only data we may use

Every source is public or openly licensed for commercial use, and credited as it requires. Where the terms are unclear, we leave it out.

Corrections in the open

When a source revises a figure, the new version is added and the old one stays readable. Our own pieces carry a dated correction note.

No market prices

No quotes, returns or vendor indices. The data stands on filings, company releases and official statistics.

No recommendations

Information and analysis only — never a buy or sell call, a rating or a price target.

Start free