Write a report surveying how small companies are preparing for ESG disclosure

## Role and objective

You are a deep-research analyst. Produce an investigative report surveying how small companies are preparing for ESG disclosure for [FILL IN: intended readership]. Treat preparation as observable activity, such as selecting frameworks, collecting data, assigning responsibility, obtaining assurance, building controls, or responding to customer and investor requests. Do not present a polished company profile or advocacy piece. The report must distinguish documented practices from interpretation and identify differences by company type, sector, jurisdiction, and reporting pressure when the evidence supports them. Completion requires a traceable report in which every headline figure has two independent sources, each abstract term is operationally defined, and every conclusion is limited to what the cited evidence can support.

## Scope and given facts

In scope is a survey of how small companies are preparing for ESG disclosure, including current practices, drivers, barriers, capabilities, costs where documented, and likely reporting needs. The geographic scope is [FILL IN: geographic jurisdiction or jurisdictions]. Define “small companies” as [FILL IN: employee count, revenue threshold, legal classification, or source-specific definition]. Use evidence covering [FILL IN: target reporting period or publication cutoff date].

Out of scope are unsupported predictions, general claims about all businesses, a legal-compliance opinion, and a broad account of ESG performance unless it directly explains disclosure preparation. Do not assume that “ESG disclosure” means a single framework or mandatory regime; identify the framework, stakeholder request, or rule in each source.

Fill the three slots only with information supplied by the user or confirmed by authoritative sources. Do not arbitrarily fill the actual missing items—geography, the definition of small companies, or the reporting period. If no authoritative definition is shared across sources, preserve each source’s definition and explain the lack of comparability.

## Working rules

1. Rank sources in this order: official statistics and microdata; public research institute reports; peer-reviewed articles; local government statistics; international comparative datasets. Use company surveys, professional bodies, and commercial reports only when their sampling, fieldwork, definitions, and limitations are visible.
2. Require two independent sources for every headline figure. Two documents citing the same underlying dataset are not independent. If a figure cannot be cross-checked, mark it “[VERIFY]” and explain why.
3. Operationalize each abstract term. For “preparing,” specify observable indicators such as framework selection, materiality assessment, data-system changes, governance assignment, supplier-data collection, assurance planning, or disclosure drafting. For “small company,” record the source’s threshold and legal or statistical basis.
4. Separate correlation from causation. When sources associate disclosure preparation with customer pressure, financing, regulation, or sector membership, report the association unless a credible design supports causality. Address reverse causality, omitted variables, and confounders where relevant.
5. Record each source’s geography, population, sample size, fieldwork or publication date, methodology, reporting framework, and limitations. Do not combine unlike denominators or time periods without explaining the adjustment.
6. For the United States, search and identify the Census Bureau, including ACS and decennial data where relevant, BLS, BEA, FRED, data.gov, and the relevant state agency. Use the relevant repository only when it contains evidence for this question; do not imply that a repository directly measures ESG preparation if it does not.
7. Label every series by vintage or revision status, including advance, second, or third estimate where applicable. Flag that county and metropolitan-area boundaries can be redefined between decennial cycles, breaking a time series without warning. If the scope is not the United States, do not import these repositories as governing evidence; state the applicable jurisdictional source gap.
8. Do not invent paper titles, authors, DOIs, table numbers, survey results, company examples, or regulatory conclusions. If a source cannot be opened or verified, identify the limitation rather than reconstructing its contents.

## Output structure

Use the following report structure and allocate approximately 10–15% of the report to each of the first two chapters, 20–25% to the evidence and methods chapter, 30–35% to findings, 15–20% to barriers and implications, and 5–10% to limitations and conclusion. Adjust only when the evidence volume requires it.

1. **Executive summary** — State the research question, scope, definition of small companies, principal documented preparation patterns, major barriers, and evidence limitations. Include no unsupported headline figure.
2. **Definitions, scope, and method** — Explain the jurisdiction, period, inclusion criteria, operational indicators, source hierarchy, independence test, and comparability limits.
3. **Reporting context and preparation pathways** — Describe the disclosure frameworks, stakeholder demands, or rules that sources explicitly connect to preparation. Do not infer legal applicability.
4. **Findings** — Organize by preparation activity: governance and accountability; materiality and prioritization; data collection and controls; framework or metric selection; supply-chain and customer requests; assurance or external review; systems, skills, and expenditure. Compare subgroups only where definitions and denominators permit.
5. **Barriers, differences, and implications** — Cover capability, cost, data, uncertainty, and proportionality barriers, separating observed evidence from interpretation.
6. **Limitations and conclusion** — State evidence gaps, non-independent sources, unresolved [VERIFY] figures, and the narrow conclusion supported by the record.

Include required comparison tables and figures only when data exist. Beneath every table, write exactly: `Source: issuing body, dataset, base year / Note: indicator definition, unit`. Where numbers are not in hand, present a table as a design proposal showing column structure, population, denominator, source to collect, and missing data; never fill placeholder values.

## Style rules

Use a hybrid style. Write the executive summary, definitions, limitations, and conclusion in concise narrative paragraphs. Present methods, evidence registers, subgroup comparisons, preparation activities, and source limitations in itemized lists or tables. Use an analytical register, define ESG terms on first use, avoid promotional language and clichés such as “the new normal,” “game changer,” “one-size-fits-all,” and “unlock value,” and qualify claims according to evidence strength.

## Style rules (humanizer v1)

These govern every prose surface in the deliverable. Never alter quotations, code, identifiers, or proper nouns to satisfy them.

- Banned vocabulary: delve, tapestry, testament, showcase, pivotal, crucial, vital, intricate, interplay, meticulous, foster, vibrant, boasts, nestled, groundbreaking, and "landscape" in the abstract sense. Banned inflation phrases: plays a vital role, underscores its importance, evolving landscape.
- Banned constructions: "not just X, but Y" negative parallelism, forced three-item lists, fake ranges ("from X to Y"), signposting ("Let's dive in"), staged staccato ("One goal. Zero compromises."), and synonym cycling. Name a thing the same way every time.
- Punctuation and structure: no em dashes in the final text (rewrite with a period, colon, or parentheses), no emoji, sentence case headings, no heading on every paragraph, no bolding cadence, no "In conclusion" wrap-up. Close on a concrete fact.
- Tone: no flattery ("Great question"), no chatbot residue ("I hope this helps"), no knowledge-cutoff hedging, no stacked hedges. Hold the register the genre calls for and vary sentence length.
- Fact integrity: every instruction to be specific carries one boundary. Use only facts present in the user's input or in a verifiable source. Do not invent details to sound human. Leave anything the user did not supply as a literal [FILL IN] slot instead of a plausible guess.
- False-positive guard: flawless grammar, a single em dash, one "however", or formal wording is not by itself an AI tell. Rewrite only where several signals cluster, and never rough the prose up on purpose.

## Final self-audit

Draft the deliverable in full, then interrogate the draft on two counts. Which passages read as obviously AI-written when checked against the style rules above? Did any line assert a fact absent from the user's input and unverifiable from the sources given? Rewrite what fails and submit only the corrected version. The audit itself never appears in your output.

## Self-verification

1. Confirm that the deliverable is a report surveying preparation for ESG disclosure, not a generic ESG overview or compliance opinion.
2. Confirm that the geographic scope, small-company definition, and reporting period are either supplied, sourced and labelled, or retained as explicit slots.
3. Check every headline figure for two genuinely independent sources; mark it “[VERIFY]” if cross-checking failed.
4. Check that sources are ranked according to the specified hierarchy and that duplicate use of one underlying dataset is not counted as independence.
5. Check that “preparing,” “small companies,” and other abstract terms have observable operational indicators.
6. Check that correlation is not written as causation and that reverse causality, omitted variables, and confounders are addressed where relevant.
7. Check United States evidence for the named repositories, series vintage or revision labels, and the county and metro boundary warning when applicable.
8. Check that every table has the exact required source note and that unavailable numbers appear as design proposals rather than invented values.
9. Check for facts, company examples, figures, paper titles, authors, DOIs, or table numbers added beyond the user input without a verifiable source.
10. Check that no slot—especially geography, company-size definition, or reporting period—was filled arbitrarily.
11. Check that the report does not drift into measuring ESG outcomes, predicting regulation, or giving legal advice unless directly tied to disclosure preparation and supported by evidence.
12. Check that the hybrid format is visible: narrative sections remain narrative, while methods and comparative evidence use lists or tables.