Manual data entry is the recording of a measurement by a person rather than by an integration. In security measurement it is treated as a stopgap on the way to automation. For a substantial part of any security programme, it is the permanent and correct answer.
Whether a tabletop exercise actually stressed the response plan or confirmed that everyone knew their lines.
Whether a third party closed what it committed to, when the evidence is a PDF attached to an email.
Whether a budget line bought the capability it was approved for.
Whether an audit finding is genuinely closed or has been closed on paper.
Whether the team has the skills for the estate it now runs.
Board and committee decisions, exception approvals, policy waivers and their expiry.
None of these live in a security tool. Several of them are the most consequential things the function does.
Because automated data feels more objective, and because manual entry is associated with the spreadsheet-based measurement everyone is trying to escape.
The result is the failure described under the McNamara fallacy. The measurable half of the programme gets measured, the rest gets omitted, and omission is read by every audience as zero.
Attach evidence. A number typed in with the report, screenshot or approval that supports it is a different object from a number typed in alone. Where the entry matters, evidence should be mandatory rather than optional.
Record who entered it and when, and keep the history of changes.
Give it an observation date distinct from the entry date, so a figure recorded in July describing a June exercise ages correctly.
Set a cadence and an owner, so a missing entry is visible as a gap rather than as an absence nobody noticed.
Treat it as a first-class input rather than an annotation. If it lives in a side spreadsheet, it will lose every argument against an automated number, regardless of which one matters more.
Metric Maestro treats manual entries as facts with the same standing as anything pulled from an API. Entries carry supporting evidence, an entry author, an observation date and full change history, and they feed metric computation the same way automated facts do.
That is also what makes measurement possible on day one rather than after an integration programme. The manual path is the fast path, and integrations replace individual sources over time without changing the metrics built on them.
From the blog