Delegation Handoff Completeness: An Auditable Sampling Framework

Table of contents

• Question and method
• Why handoff completeness is worth measuring
• The six fields
• Writing defect rules
• Stratify the sample
• Sample size and review
• Reporting
• Limits
• Sources
• Frequently Asked Questions

Question and method

Research question: How can a small team test whether delegated task briefs contain enough information for an assistant to begin, check, and hand back the work without inventing facts?

This paper describes a sampling framework. It does not report a defect rate, a benchmark, or an outcome study. The unit of analysis is the task brief: the written request that accompanies a delegated task. The framework defines the fields, the defect rules, the strata, and the reporting record. It assumes the organization has a task-management system or another durable place to store the brief.

The O*NET profile for executive administrative assistants describes a broad set of coordination tasks, including preparing correspondence, managing calendars, and arranging information. Those tasks vary in risk and in the information required to start. A completeness measure should reflect that variation rather than treating every task the same.

Why handoff completeness is worth measuring

Delegation failures are often attributed to the assistant's capability when the actual defect is in the request. A brief that lacks a source forces the assistant to choose one. A brief that lacks an authority boundary forces the assistant to guess whether to send, commit, or ask. A brief that lacks a quality bar makes "done" a matter of opinion.

These are measurable conditions. Each can be present or absent in a written brief. Measuring them separates a training problem from a brief-design problem, which leads to different fixes.

The measure also protects the assistant. A record that shows a missing authority field makes visible that the assistant was asked to act without a decision rule. That is more useful than a retrospective conversation about why the assistant "should have known."

The six fields

The framework uses six fields:

1. Outcome: the finished state in one sentence.
2. Source: the authoritative document, system, or person, including version and location.
3. Authority: permitted actions, actions requiring approval, and the backup approver.
4. Deadline: date, time, time zone, and the consequence of missing it.
5. Quality bar: the acceptance test the assistant can apply before handing back.
6. Escalation route: the stop condition and the decision owner.

A brief may contain more, but these six are the minimum for the measure. Additional fields, such as budget or confidentiality level, can be added for specific task families.

Writing defect rules

A defect rule says what counts as missing or unusable. Write it before looking at the sample.

Decide how to treat partial fields. A source named as "the usual folder" may be acceptable if a link resolves and the version is identifiable. A deadline of "Friday" may be acceptable for low-risk work but defective for an external commitment. Record these field-specific rules and keep them stable during the review.

Stratify the sample

Task risk changes the expected completeness and the consequence of a defect. Use at least three strata:

• Low risk: internal scheduling, file organization, research, meeting notes.
• Medium risk: external scheduling, vendor follow-up, report production, inbox drafting.
• High risk: spend, contracts, personal data, legal or personnel communication.

Sample within each stratum. A random sample dominated by low-risk tasks will make the overall completeness rate look better than the experience of the high-risk work.

Also record the task family and the owner. A pattern by owner may indicate a training need; a pattern by task family may indicate a template problem.

Sample size and review

A pilot of 20–30 briefs can show whether a field is broadly missing. A precise rate requires a larger designed sample. Report the method and the uncertainty rather than presenting a pilot as a precise estimate.

Use two reviewers on a subset and compare judgments. If they disagree often, the defect rules are not specific enough. Resolve disagreements by clarifying the rule, then apply the clarified rule to the rest of the sample. Record the clarification.

Reporting

A useful report shows, by stratum:

• briefs reviewed
• defect count and rate per field
• examples of the defect without personal data
• owner of the fix
• action and review date

Do not publish a single "handoff score" that hides which field failed. A team can have a strong outcome field and a weak authority field; the intervention is different.

Limits

The framework measures the presence and usability of written fields. It does not measure whether the task was appropriate to delegate, whether the assistant had the skill, or whether the outcome was good. Those require separate review.

It also depends on the brief being the actual instruction. If the real request happens in a meeting and the written brief is created afterward for the record, the measure describes the record, not the handoff. Note that limitation explicitly.

The NIST Privacy Framework is relevant when briefs contain personal data; minimize what is collected and limit who can read it. The U.S. Department of Labor hiring guidance and Bureau of Labor Statistics occupational information provide context for the administrative role, not a completeness standard.

Sources

1. O*NET OnLine: Executive Administrative Assistants
2. NIST Privacy Framework
3. U.S. Department of Labor: Hiring
4. Bureau of Labor Statistics: Occupational Outlook Handbook
5. U.S. Office of Personnel Management: Delegation

Frequently Asked Questions

Is a high completeness rate proof that delegation works?

No. It shows the briefs contain the defined fields. It does not show the work was well chosen, well executed, or correctly reviewed.

How often should the sample run?

Quarterly is common for a stable process, and after a template or role change. A monthly pilot may be useful while a new brief is being adopted.

Should the assistant score their own briefs?

The assistant can prepare the sample and flag missing fields, but an independent reviewer should apply the defect rules. Self-scoring invites optimistic interpretation.

Review cadence

Repeat the sample after a process change and compare the same fields. If completeness improves but rework does not fall, the missing field was probably not the real cause. If rework falls without completeness improving, check whether the receiving owner simply absorbed the gap. Report both, and keep the sampling rule stable.

A good sample also records the receiving owner’s view. A handoff can look complete on paper and still fail because the owner cannot find the file or does not know which decision is open. Add one question to the review: what did the receiver have to ask for? Those questions are the most useful improvement list.