Publish Research and Thought Leadership

13.1 Choose Topics Where the Firm Has Substantive Expertise and the Audience Has Meaningful Questions

Begin with a question the intended reader needs to answer. “The future of finance” is a subject area. “How should a CFO compare business units that allocate shared costs differently?” identifies a decision and gives the author a concrete job.

Collect candidate questions from permitted notes of client discussions, proposal questions, recurring practitioner explanations, audience feedback, and the monitoring program. Remove confidential particulars. Group variants that reflect the same underlying decision instead of commissioning a separate article for every wording.

For each candidate, ask whether the firm has a useful contribution beyond an existing explanation. That contribution might be a new observation, an analytical distinction, a worked example, or a synthesis that makes scattered evidence easier to apply. Originality does not require a new dataset for every article.

Google’s content guidance asks creators to contribute original information or analysis, add value beyond their sources, and make relevant expertise and authorship clear. This supports a useful editorial priority; it does not establish a formula that guarantees visibility across AI products. 

Reject a topic when the firm lacks the expertise or evidence needed for its central claim. Narrow it when only part is supportable. If a proposed industry survey is infeasible, a clearly labelled practitioner explanation may still answer a useful question.

Build a small agenda around a coherent problem

For an initial program, select one priority audience and a few related questions. A consulting practice might examine how manufacturers resolve disagreements in monthly planning. Its publications could explain decision rights, report a descriptive benchmark, and discuss exceptions raised by readers.

Set the pace according to expert availability and the work required. A technical analysis may need rapid review after a development; an original study may need months. Protect time for substantive review before committing to a publication date. A calendar should organize the work, not determine its findings.

13.2 Turn Practitioner Knowledge Into Useful Frameworks, Diagnostics, and Explanations

Experienced practitioners often explain a problem better in conversation than in a first draft. An editor can help recover the reasoning by asking about choices, exceptions, and evidence. The practitioner remains responsible for the professional substance.

Interview for the reasoning

• What decision does the reader need to make, and what do people commonly misunderstand about it?
• Which facts or conditions would change your answer?
• What are the plausible options, and when would you advise against each?
• Can you work through an example using invented or appropriately permitted information?
• Which statements are supported by evidence, and which express your judgment?

Turn the discussion into a proposed argument, a worked example, and a list of claims needing support. Ask the practitioner to review those elements before an editor polishes the prose. Otherwise, an attractive draft can conceal that the reasoning is incomplete.

A useful framework makes distinctions that affect a decision. Explain its inputs, how the reader applies it, what the resulting categories mean, and where it stops being useful. Credit an existing framework when adapting it. A new name or diagram does not make an established idea the firm’s invention.

Be explicit about the status of a diagnostic

A checklist can help a management team identify questions for investigation. A numerical diagnostic makes an additional claim about measurement. Explain how its scores and thresholds were chosen, what has been tested, and what an assessment can support.

If scoring reflects practitioner judgment, say so. Do not describe a maturity score as a validated predictor of business performance without the relevant validation. Include a meaningful path for a reader whose circumstances do not fit the categories; every response should not lead automatically to purchasing the firm’s service.

Accounting example: comparing business-unit results

An accounting firm could publish a management-reporting guide for a CFO comparing units that use different cost allocations. The guide preserves each unit’s reported result, explains the definitions and period, and shows a clearly fictional reconciliation to a more comparable measure.

For every proposed adjustment, the guide explains its reason, supporting information, and remaining limitations. It might show why reallocating a shared cost changes the comparison while leaving differences in service mix unresolved. The reader sees the professional reasoning rather than just a normalized total.

Describe the example as a management-reporting illustration. Identify the reporting basis and intended use, and distinguish management adjustments from any required financial-reporting treatment. The publication should not imply that the illustration establishes compliance or provides assurance.

Before release, ask a colleague unfamiliar with the draft to apply the explanation to a different example. Confusion about inputs or exceptions is a reason to revise the explanation. This is a usability check, not proof that a diagnostic predicts outcomes.

13.3 Design Original Research, Surveys, Benchmarks, and Technical or Regulatory Analyses

Choose a method that can answer the question. Interviews can explore how people reason. A survey can describe reported practices or views within the limits of its design. Operational records can support defined measures. A technical analysis can explain what follows from a standard, rule, dataset, or set of assumptions.

An original synthesis can also be valuable. State which sources were examined, why they were selected, and what the firm adds. Do not call a literature scan comprehensive or systematic unless the work supports that description.

Define what is being studied

For a benchmark, specify whether the observation is a company, office, project, transaction, or person. AAPOR’s guidance for establishment surveys emphasizes clear units and eligibility rules, including situations where several people or systems supply information for one organization. Three executives describing one company do not automatically create three company observations. 

Define metrics before collecting them. “Planning cycle completed” might refer to a meeting, an approved forecast, or an agreed set of actions. Without a shared definition, a precise-looking table may compare different things. Document the measurement period, units, exclusions, and what counts as a usable observation.

Design the questionnaire and sample together

AAPOR recommends specific, neutral questions addressing one concept at a time and pretesting with people similar to the intended respondents. Test how respondents interpret the question as well as whether the form works. Keep skipped answers distinguishable from substantive responses such as “no” or “don’t know.” 

For a survey drawn from clients, subscribers, or a professional network, explain how volunteers entered the sample. Such recruitment does not become representative of an industry merely because many people participate. AAPOR’s disclosure guidance permits precision measures for nonprobability studies only with an explained model, validated assumptions, and calculation. A conventional margin-of-error formula is not a substitute for that work. 

Use a research specialist when the intended claim requires population estimates, complex weighting, predictive validation, or comparisons the team cannot competently design. For a small program, a transparent descriptive study may be the appropriate scope.

Agree on the analysis before choosing the headline

Write down the primary question, planned comparisons, treatment of incomplete records, and reasons for exclusions before reviewing the results. The American Statistical Association’s ethical guidance opposes efforts to predetermine results and selective interpretation, and calls for transparency about planned and unplanned analyses. Additional exploration can be useful; identify it as such. 

Keep the research capable of producing an inconvenient finding. If respondents do not support the premise behind the project, report what the evidence supports or explain why the study cannot answer the question. Do not change exclusions or combine categories merely to produce a stronger marketing claim.

For recurring benchmarks, preserve comparable definitions and collection methods or explain the break in the series. AAPOR cautions that changes in question wording, context, and mode can mimic a change in the underlying subject. A different mix of participating firms can also change a descriptive benchmark. 

Law-firm example: analyzing a regulatory change

When a rule affecting its clients changes, a law firm can publish a dated analysis of what changed and which organizations should assess its relevance. The publication should identify the jurisdiction, instrument, version, and status. Distinguish a proposal, final rule, guidance, and a court decision.

Separate the source’s publication date from its effective date and any distinct compliance or transition dates. Explain what the primary text expressly says, the firm’s interpretation and reasoning, and questions requiring an organization-specific assessment. Link to the relevant provisions so readers can examine the basis.

The contribution may be an explanation of operational choices, unresolved interpretations, or dependencies that a brief announcement omits. State what would trigger an update, such as revised guidance or a relevant decision. Do not convert a conditional assessment into an instruction that every reader must act immediately.

13.4 Document Methods, Sources, Dates, Authorship, Limitations, and Relevant Interests

Publish enough information for a reader to assess the contribution. A prominent claim should have a visible route to its support. Keep detailed working records internally, and give the public an intelligible account of how the conclusions were reached.

Use a compact methods and source note

For survey findings, AAPOR’s transparency standards cover sponsorship, instruments, recruitment, collection, sample counts, processing, limitations, and relevant AI use. Use the applicable standards to determine the full disclosure required for the study; a short marketing summary should link to that fuller account. 

For other publications, adapt the note to the work. A reporting guide should identify its basis and assumptions. A regulatory analysis should identify the primary authorities and the date through which they were checked. A framework should explain whether it is a practitioner proposal, an adaptation, or something that has been evaluated.

Name the author and the actual reviewer, with their respective roles. An editor who shapes the prose, an analyst who calculates results, and a partner who reviews professional interpretation make different contributions. A partner’s name should not imply review that never occurred.

Identify the publisher, funder, and relevant commercial relationships. If a technology provider financed a benchmark about the problem its product addresses, readers need that context. Explain the sponsor’s role in the design, analysis, and release. Commissioning an external reviewer does not by itself make a sponsored publication independent.

Preserve source and version information

W3C’s Data on the Web Best Practices recommends documenting data origins and changes, identifying versions, retaining version history, and stating reuse terms. Apply those principles proportionately: retain the source release, calculations, approved manuscript, and correction record. A publication does not need to expose restricted records to explain its evidence. 

Give a publication a date and, where relevant, a distinct data period or research cutoff. Label a historical edition clearly. When a material correction changes a finding, explain what changed and update affected summaries and downloads. Changing a displayed date alone does not update the underlying analysis.

Make AI assistance accountable

AI can help propose questions, organize approved material, or prepare a draft for review. Check source references, quotations, calculations, and substantive interpretations against the actual evidence. Describe material use in research collection or processing and the human checks applied.

AAPOR’s current code distinguishes human research participants from AI-generated cases and requires generated cases to be identified. Simulated CFO responses are not additional interviewed CFOs. Keep a simulation visibly separate from empirical findings and from the reported human sample. 

Apply Chapter 12’s permissions and confidentiality review before using client material or participant information. Research participation should have an honestly described purpose; do not turn an invitation to contribute evidence into undisclosed sales prospecting. 

13.5 Use Newsletters, Podcasts, Webinars, and Conference Contributions to Share Substantive Expertise

Choose a format for the contribution it enables. A newsletter can explain one finding to an existing audience. A podcast can expose the reasoning behind a judgment. A webinar can work through an example. A conference contribution can invite professional scrutiny and discussion.

Start each adaptation from the reviewed publication or evidence record. Give the presenter or editor the supported finding, essential qualifications, relevant source links, and matters that remain uncertain. The purpose is to make the substance useful in another setting.

Design the discussion around a real question

For a consulting benchmark, ask why participating firms organize planning differently and when a practice might be inappropriate. For an accounting guide, work through the reconciliation rather than reading the article aloud. For a legal update, distinguish questions the published analysis can answer from those needing an assessment of particular facts.

Invite thoughtful disagreement. If a guest challenges an assumption, explore the reason and record any resulting correction. Do not transform every conversation into endorsement of the firm’s position. Obtain the necessary permissions for recording, attribution, excerpts, and later publication.

Create a usable written record

W3C distinguishes a basic transcript of relevant audio from a descriptive transcript that also conveys visual information needed to understand a video. Make the transcript easy to find from the recording. A highlights article serves a different purpose and should be labelled accordingly. 

Review automated transcripts against the recording, particularly names, figures, technical terms, and words such as “not” that reverse meaning. Identify speakers. For a webinar, explain an essential diagram or calculation that the presenter points to without describing aloud. Captions and other accessibility needs require their own attention.

Preserve qualifications when preparing short excerpts. “Among participating companies” should not disappear from a survey finding, and a jurisdiction should not disappear from a legal summary. If an excerpt is too short to remain accurate, use a narrower statement or direct readers to the explanation.

Keep a record of published derivatives so corrections can reach them. Chapter 14 addresses earning external coverage and recognition. Here, the priority is a substantive contribution and a dependable record of what was said.

13.6 Publish Accessible Text, Evidence, and Summaries While Making Deliberate Decisions About Gated Material 

Give each substantial publication a stable page on the firm’s website. Explain the question, contribution, authorship, date, principal findings, and limitations in readable text. Link to the fuller report, methods, related practitioner profiles, and relevant service information without interrupting the explanation with repeated sales claims.

Publish a report in a format that suits its users. A downloadable document can support printing and internal discussion. An HTML page can make the main explanation easy to navigate and link. Check that critical information remains understandable without downloading a file or interpreting an image.

W3C’s guidance on complex images calls for a text alternative that conveys the essential information. For a benchmark chart, provide an intelligible finding and an accessible table or adequate description. Retain the population, denominator, units, period, source, and relevant exclusions. A label such as “benchmark chart” is insufficient to explain the result. 

Choose what to make public

Do not assume every registration form makes its content invisible to every search or AI system. Google documents arrangements for indexing subscription and registered content when the appropriate bots can access it. Actual access depends on implementation and the service involved. 

As a practical default, make the main contribution and its necessary qualifications public. Gate an additional asset when there is a clear reason, and explain what the reader receives. Apply the firm’s current approved access settings and Chapter 10’s technical checks. Google explicitly states that meeting its requirements does not guarantee crawling, indexing, or serving. 

Before release, have someone open the public page without a special account. Can they identify the finding, understand its limits, and reach the supporting explanation? Keep protected information protected; improve the public account when that account is incomplete.

Add the publication’s topic to the existing monitoring program. Track whether assistants cite the work, attribute it correctly, preserve material limitations, and describe the firm’s relevant expertise accurately. Also record useful reader questions and qualified inquiries. A research citation can be worthwhile even when it does not produce an immediate service recommendation.

Practical Output: A Thought-Leadership Agenda and a Research or Publication Brief

Choose one priority audience and commission one substantial contribution before expanding the calendar. The following entries illustrate alternatives for different practices; they are not a requirement to publish all three.

For the selected entry, record the accountable expert, editor or analyst, evidence required, intended formats, access decision, review dependencies, target date, and update owner. Identify the specific audience question for each follow-on item so republishing does not become repetition.

Worked research brief: monthly planning practices

This hypothetical consulting brief proposes a modest descriptive study. The numerical illustration below shows how a possible result would be reported; it is not a completed survey or an expected finding.

Purpose and audience: Help operations leaders at privately held US manufacturers consider how responsibility for monthly planning is assigned.

Working title and scope: “Monthly planning practices among participating manufacturers.” Define eligibility as companies with 50–250 employees and an established monthly process for reconciling sales and operational plans. The company is the unit; collect one coordinated response per company.

Recruitment and design: Invite 60 eligible companies from a professional network through a designated operations leader. Participation is voluntary. Explain the sponsor, purpose, intended public reporting, data handling, and optional follow-up separately. The study is unweighted and is not a probability sample of US manufacturers.

Primary question: “For your most recently completed monthly planning cycle, was one person explicitly designated as responsible for securing approval of the final cross-functional plan?” Offer yes, no, don’t know, and prefer not to answer, while retaining blank responses separately. Pretest the meaning of the question and revise it before launch if respondents interpret it differently.

Collection plan: Set and record the fieldwork dates and the month covered by each response. Track eligibility, duplicate-company submissions, withdrawals, and item completeness. Finalize definitions and exclusion rules before analyzing results. Do not collect customer names, product-level forecasts, or other details unnecessary to this question.

Hypothetical reporting example: Suppose 30 companies participate: 18 answer yes, 9 answer no, and 3 leave this question blank. The yes/no base is 27. An accurate statement is: “Among the 27 participating companies answering yes or no, 18, or approximately 67%, reported designating one person to secure approval. Three other participating companies did not answer this question.”

Interpretation and limits: Report the missing answers and voluntary recruitment alongside that statement. The result describes reported practice among participants. It does not establish industry prevalence, verify how authority works in practice, or demonstrate that assigning an owner improves performance. Discuss those questions separately as practitioner interpretation or future research.

Public release: Publish the finding, counts, exact question, fieldwork and reference periods, recruitment description, methods, and limitations. Add a practical explanation of decision rights that distinguishes the survey evidence from the firm’s judgment. A worksheet can help readers identify their own unresolved decisions.

Responsibilities and review: The research lead checks the design, records, and calculation. A qualified methods reviewer assesses whether the intended claims fit the design. The practice lead checks interpretation; the editor checks headlines, summaries, and attribution. Complete the applicable permission and disclosure reviews before release.

Distribution and follow-up: Prepare one newsletter explanation and a practitioner conversation addressing what the study leaves unresolved. Link both to the public release. Assign the owner responsible for questions, corrections, and any decision to repeat the study.

Move the brief into production

Reserve time for design, source collection, analysis, writing, review, publication, and maintenance. Approve the question and feasible method before investing in production. If evidence or participation is inadequate, revise the scope or defer the empirical claim.

At release, the team should be able to show what the publication contributes, which claims the evidence supports, who takes responsibility for them, and where a reader can inspect the explanation. Use the questions that follow publication to improve the next piece of work.