Executive search

A research-request handoff for executive search: question, boundary, acceptance

Commission a bounded search-research task with a clear question, permitted sources, useful acceptance criteria and a review route. Includes a copyable request and fictional example.

By LaisyPublished

“Can you research this market?” sounds like an instruction. It leaves the researcher to decide the question, scope, evidence standard and point at which the work is useful enough to review.

The researcher may produce a substantial document while the consultant was expecting a quick answer to a narrow uncertainty. Both people have worked hard. The handoff failed because they were working toward different decisions.

A retained-search research request should commission an answerable task. It needs a purpose, a boundary and an acceptance route. The request can be short, but the researcher should be able to explain what they are investigating, what they must not assume and who will review the result.

TL;DR

  • State the decision the research will support before listing sources or deliverables.
  • Name the scope, exclusions and evidence gaps that matter to that question.
  • Agree what the reviewer needs to accept the work, including how uncertainty should appear.
  • Keep changes to the request visible; more polished output cannot repair an unconfirmed brief.

Copyable research-request handoff

Complete this with the person doing the work. A request is not accepted merely because it has been sent.

Request elementWhat to agree
QuestionThe specific uncertainty to investigate
Decision supportedWhat the consultant expects to do with the answer
Assignment baselineCurrent agreed brief and the relevant criterion
Inclusion boundaryCompanies, roles, geography, period or other limits relevant to this question
ExclusionsWhat is outside scope and why
Permitted inputsSelected sources and any restriction on access or reuse
OutputThe smallest useful format, with evidence pointers and gaps
Acceptance criteriaWhat the reviewer must be able to check
OwnershipResearcher, reviewer and person who can clarify or change the request
Review pointWhen to compare early findings with the question before expanding the work

The format is an original working suggestion. It is not a professional certification or a substitute for an engagement's actual terms. Keep enough detail to resolve ambiguity without making the form harder to use than the research itself.

Write a question that changes a decision

“Find companies” describes activity. “Establish whether our current market boundary excludes a relevant operating model” identifies something a consultant can decide after reviewing the answer.

Work backward from that decision. Ask what evidence would change the consultant's view and what would leave the question unresolved. If neither person can describe that difference, pause before expanding the research list.

Use the current mandate kickoff record to locate the relevant instruction. Research can expose a reason to revisit the brief, but it should not silently replace it. Mark a proposed expansion as a proposal and identify who can confirm it.

For AESC members, the Client Bill of Rights includes written clarity about the engagement and its scope. Our task-level handoff applies a similar preference for explicit instruction; it is not an AESC form or a claim that all firms are members.

Make the boundary practical enough to use

Specify the aspects that change inclusion. A geography label may describe headquarters, an operating location or the location of the role. An industry label may hide materially different business models. A company-size field may refer to a group or a division.

Choose definitions appropriate to the assignment and state them. Include an example of an obvious inclusion and an edge case that needs review. These examples help reveal misunderstandings before the researcher repeats them across a larger set.

Where the question concerns a role, connect it to actual work. OPM's job-analysis guidance links job tasks with the competencies needed to perform them. Its scope is federal assessment guidance. It does not validate a private firm's sourcing filter or make a title a reliable proxy for experience.

Ask the researcher to flag ambiguous cases rather than force every record into yes or no. The consultant can then distinguish a limited evidence base from a genuine scope decision.

Agree an evidence standard suited to the question

Decide what the researcher should substantiate, where an attributed account is acceptable and which missing information should block a conclusion. The standard should reflect the intended use of the work.

A public company announcement can establish that an event was announced. It may not establish a particular individual's responsibility for it. A current job title can help identify a question; it does not establish willingness to move or suitability for an assignment.

The Government Data Quality Framework's guidance emphasizes identifying critical data and communicating known limitations. That is public-sector guidance, adapted here as a practical prompt to specify what matters to this task. It is not a quality badge for the resulting research.

Agree how the researcher should expose a gap: the question attempted, sources checked, finding and remaining limitation. “Not established from the selected sources” is often a more useful answer than a confident sentence assembled from weak clues.

Define acceptance before requesting volume

An acceptance criterion describes something the reviewer can inspect. “Comprehensive and high quality” provides little help when the consultant and researcher disagree about completion.

Useful criteria might require the output to state its scope, distinguish confirmed findings from interpretations, link consequential claims to sources and identify unresolved edge cases. The reviewer should also be able to see whether the answer supports the original decision.

Specify a format that makes those checks possible. A compact table may work for a bounded market question. A short memo may better explain a conflict between sources. Do not require both by default if one repeats the other without adding useful evidence.

Keep the commissioned task separate from its resulting candidate research notes. The request says what work was authorized and how it will be reviewed. The notes contain the findings. Connecting them is more useful than copying the whole brief into every record.

Fictional example: an early sample prevents a broad detour

This example uses invented organizations and counts. It illustrates a commissioning decision, not a benchmark for research speed, market size or search success.

A fictional search partner asks a researcher to examine a neighboring market. The initial phrase is “look at adjacent companies.” In a short discussion, they identify the real question: whether a particular operating model appears within the agreed geography.

They agree to review an initial set of six companies, selected to test different edges of that definition. Six is a working sample chosen for this exercise, not a statistically representative market sample or a recommended universal quota.

Early findingWhat it establishesReview consequence
Two clear examplesSelected public records describe the relevant modelRetain as examples within the agreed boundary
Three ambiguous casesGroup and division information cannot be separated confidentlyAsk the partner which unit matters before continuing
One excluded caseThe stated operating geography falls outside the requestKeep the exclusion reason; do not expand scope informally

The useful result is the ambiguity surfaced by the three cases. The partner clarifies the unit of analysis before the researcher expands the work. A longer list produced under the original vague instruction would have hidden that decision inside many apparently finished records.

Set a review point before the task becomes expensive

Choose the point at which initial findings will reveal whether the question and method make sense. It can be event-based: after the first meaningful edge case, after a conflicting source or before moving into a new source category.

At that review, compare findings with the request. Ask whether the researcher can explain the inclusion logic, whether the selected sources are sufficient and whether the consultant still needs the same answer.

If a dependency blocks the work, record what is needed and who will resolve it. Avoid leaving the researcher to infer an answer because a client response is late. The client-update workflow can help turn a material dependency into a clear question for the next client discussion.

Use the review to stop unnecessary work as well as to authorize more. If the evidence already answers the decision, collecting additional material may add maintenance and review effort without improving the conclusion.

Make revisions identifiable

Research requests change as the assignment develops. Record which instruction changed, who confirmed it and what existing work is affected. Give the researcher a way to distinguish the current request from a superseded one.

Google Docs' version-history help documents reviewing earlier changes and creating named versions. Those features can help locate the instruction the researcher received. Whatever tool you use, preserve the reason for a consequential change as well as its text.

Do not mark old findings incorrect merely because the question changed. Retain their original scope and decide whether they remain useful. A source-backed finding can be sound while no longer answering the current request.

If the researcher changes, include the current request, accepted work and unresolved questions in the handover. A new colleague should not have to infer the commission from the length of the existing document.

Atlassian’s DACI play distinguishes coordination, approval and contribution. Use that distinction to clarify who can change the research request and who checks the delivered answer, while preserving the engagement’s actual decision process.

Check source visibility as well as task ownership. Clockwork’s internal-note guidance documents client-visibility and notification choices in that product. Verify the behavior of the tool you actually use before placing working research in a shared record.

Accept the answer with its limitations

The reviewer should compare the delivered work with the agreed question and acceptance criteria. Check consequential sources, sample inclusion decisions and read the stated limitations. Distinguish a correction from a new task.

If something fails the agreed criterion, identify the specific gap. If the consultant now wants a different geography or a different question, commission that change explicitly rather than describing the original work as incomplete.

Record the acceptance outcome and any residual action. Acceptance does not make uncertain information certain. It means the reviewer has assessed whether the work is useful for its intended purpose, with the remaining limits visible.

Try one bounded research request

Select a suitable task in a live assignment and compare the reviewed answer with the original question. Include clarification, checking and correction effort in the assessment.

A Laisy discussion can begin with selected permitted inputs and that specific task alongside current tools. Agree the reviewed output, a useful success criterion and when to continue or stop. No automatic research access, fixed delivery time or guaranteed result follows from this guide.

Frequently asked questions

How much detail should a research request contain?

Enough for the researcher to explain the question, boundary and review standard back to the commissioner. Add detail where ambiguity changes a decision. Avoid lengthy background that obscures the task.

Should every request include a target number of names?

Use a count only when it serves a defined purpose. A numerical quota does not establish coverage, evidence quality or suitability. Explain whether it is an exploratory sample, a reporting need or a separate requirement.

What if the researcher cannot answer the question?

Return the attempted method, findings and specific limitation. The reviewer can then change the question, authorize a different approach or stop. An honest unresolved result is preferable to an invented conclusion.

Can an AI-written request be handed straight to the researcher?

Have the commissioner and researcher check it first. Verify the current brief, permitted inputs and acceptance route. Generated wording cannot authorize access or settle an unresolved client instruction.