Skip to content

Stockertim Dictionary · Consulting · Technology

Sanity check

Also called reasonableness check, plausibility check

In short

A quick test of whether a figure is plausible, such as comparing a total with the published accounts or a unit value with common sense.

  • Beginner
  • Process
  • Checked

In detail

A sanity check is a fast, rough test that catches big errors before they spread into analysis and recommendations. It asks whether a number is roughly what it should be: does total revenue in the client's data file match the annual report, does the number of employees agree with the payroll count, is an average order value plausible for the business?

Typical checks include reconciling totals to an independent figure, looking at minimums and maximums for impossible values, checking that percentages add to 100, multiplying a unit figure by the volume to see whether it reproduces the total, and estimating the answer on the back of an envelope before running the model. A result that differs from the rough estimate by a large factor needs explaining.

Quality guidance for government analysis treats this kind of verification as a basic step, alongside more formal review. Sanity checks do not prove a figure is right; they only catch the obvious mistakes. Analysts record the checks they ran, so that reviewers can see what has and has not been tested.

Worked example

A client reports 2,400 orders a week at an average of 50 each and claims weekly sales of 120,000.

  1. Check: 2,400 × 50 = 120,000

The figures are consistent, so the team can move on to the analysis.

In practice

A data file shows annual sales of 4.2bn for a client whose published revenue is 420m. A sanity check exposes a field held in pence rather than pounds before any chart is drawn.

Watch out

  • Passing a sanity check does not make a figure correct, only plausible.
  • Run checks on the data you received before analysing it, not after the deck is built.

Learn it properly

Official sources

Facts last checked on 3 November 2026.

Education, not advice: the Stockertim Dictionary explains what terms mean. It never recommends buying or selling anything.

English