Skip to content
Benford
Start Free

Using Benford

This guide explains how the web app fits together: workspaces, uploads, analysis runs, charts, and exports. For HTTP API automation (Pro), see the Documentation hub and API quickstart. For plans and limits, see Pricing.

Workspaces

A workspace is where your datasets and analysis runs live. You pick the active workspace from the menu in the app header. Dashboard lists datasets in the current workspace; Uploads and Settings apply to that workspace too.

Roles: Owner and Admin can manage workspace settings and (on Pro) API keys. Member can use analysis and uploads according to your product rules.

Plans: Your account plan applies to workspaces you own. Workspaces you only join as a member follow the owner’s plan. Details are on Pricing.

Creating workspaces: From the dashboard you can create additional workspaces (limits apply on Free).

Workspace settings

Open Settings in the header. Here you can:

  • Confidence z-value — Controls how wide the confidence band is around the expected Benford distribution on charts (typical values are around 1.96). New runs use the value saved here; older runs keep the z-value they were computed with.
  • API access (Pro) — Create and revoke Bearer API keys for /api/v1/*. Keys are scoped to this workspace. On Free, the API is not available; you can still read the docs.

Switching workspaces in the header reloads settings and keys for the selected workspace.

Uploads

Uploads accepts CSV on Free workspaces and CSV or XLSX on Pro (with higher size and row limits). The first row must be column headers (names); every row after that is treated as data. The app profiles numeric columns and recommends which columns to analyze.

Free workspaces are limited to a maximum number of datasets per workspace; Pro is unlimited within fair use. Exact numbers are on Pricing.

Processing may finish immediately or return queued while a background worker profiles the file—follow the status link until the dataset is ready or failed.

Datasets and analysis runs

Each upload creates a dataset. When you open a dataset, you can run analysis on the selected numeric columns. That creates an analysis run with status such as queued, running, or completed.

Runs store the tests used (first digit, first-two digits, second digit, value repetition, and so on) and the z-value in effect for that run so results stay reproducible.

Last Two Digits (LTD) is an optional test you can include when you run analysis. It is separate from Benford’s law: each last-two-digit bin (00–99) is compared to a flat 1% expected rate, not a log10 curve. Before the test runs, you choose decimal places (how amounts are scaled to whole numbers, e.g. 2 for cents) and data type:

  • Dollar — amounts under $10.00 are excluded (with the usual decimal scaling).
  • Count — scaled whole numbers under 1000 are excluded.

Under Dataset details, Save LTD defaults stores decimal places and data type for the dataset so new runs (and API-queued runs) start with those choices pre-filled.

A background job usually completes the run within a few seconds after you queue it. Use View results on a completed run to open charts; Process queued runs now is only needed if a run is still queued.

Results: charts, drill-down, exports

Charts compare observed digit frequencies to Benford expectations, with confidence bands where applicable. Drill-down lets you move from a bar (a digit bin) into underlying amounts and row references when your data supports it.

On every Benford drill-down (except Summation Test), you can also expand a Per-bin breakdown table that lists each digit/bin with its count, observed %, expected %, and per-bin difference columns, plus totals for sample size, MAD, and SSD. Next to each SSD value, the app shows a decision (for example Acceptably Close) from Alex Ely Kossovsky’s SSD cutoff bands in Benford’s Law: Theory, the General Law of Relative Quantities, and Forensic Fraud Detection Applications (2014), Figure 3.3, p. 133 — bands differ by test (first digit, second digit, first-two digits, last two digits). Hover the SSD line for the band ranges and citation. These labels describe statistical distance from the expected distribution, not proof of fraud. Use Export CSV on that table to compare row-by-row against a reference spreadsheet or other Benford analytics application. Counts and first-digit / first-two-digit SSD should match exactly; second-digit SSD may differ by up to ~1.0 because many reference spreadsheets round percentages to two decimals before squaring — this is a display artifact, not a counting difference (hover the SSD value in the app for a short reminder).

Last Two Digits results appear below the Benford charts when that test was enabled for the run. The LTD chart shows observed vs 1% expected per bin (no Benford confidence bands). LTD drill-down works like Benford: click a bar, then review amounts and row references. The Per-bin breakdown table uses Obs % for the share of LTD-eligible rows in each bin (00–99); in other workpapers the same figure might be labeled as a proportion or percentage (often shown as a decimal such as 0.1063, which is 10.63% when multiplied by 100). Use Export CSV on that table to line up with external LTD workpapers. Total SSD on the run is summed over all 100 bins; it is not the same as a single-row SSD cell in a spreadsheet. When LTD was part of the run, PDF and CSV run exports and AI interpretation include LTD results alongside Benford output.

For the First Two Digits chart you can also open First Two Digits Summation Test: a separate drill-down that charts the sum of amounts in each first-two-digit bin (10–99), with an optional % of total view, then the same per-amount and row drill-down as the frequency chart. Use the header control to switch between the two views. Summation is not included in PDF/CSV exports.

AI interpretation (dataset page, after a completed run) can generate a plain-language run summary and an executive draft from the same deterministic results the charts show—optional, server-configured, and subordinate to the numeric output.

Exports produce PDF and CSV reports for a completed run. On Free, exports include plan-appropriate branding; on Pro, exports can be unbranded. Export file retention differs by plan (see Pricing).

Account and sign-in

Use Sign in and Create account from the marketing site. If you forget your password, use Forgot password on the sign-in page; you will receive an email link to set a new password.

Where to go next

  • Pricing — Free vs Pro, limits, and contact options.
  • Contact — Email for product questions, billing, and enterprise.
  • Documentation hub — API overview, quickstart, and interactive OpenAPI reference.
  • Legal — Terms, privacy, and acceptable use.

If something in the product doesn’t match this page, the in-app behavior and Pricing take precedence.