Write the decision before the title
State who is asking, the recent situation, decision they must make, consequence of error, geography, time horizon, and what a useful answer enables. Use reader language and evidence rather than a hypothetical persona. GOV.UK recommends basing needs on research and focusing on the problem rather than a proposed solution. A query such as “best budgeting app” is not yet a brief; “which method lets a variable-income freelancer reserve quarterly tax without losing bill visibility?” defines a more testable decision.
Evidence: Government Digital Service
Set scope, exclusions, and escalation at the outset
List what the article will answer, adjacent questions it will link to, readers it cannot serve, and conclusions it will not make. Assign a risk tier to financial, health, legal, safety, privacy, or other consequential claims and name the required specialist. Set publication date, research time, source cutoff, and safe stop. Google’s people-first guidance warns against producing across many topics merely for search traffic. A narrow unresolved answer is better than confident scope the team cannot maintain.
Evidence: Google Search Central
Decompose the answer into claim classes
Create rows for definition, mechanism, prevalence, eligibility, feature, comparison, performance, typicality, price, transaction, risk, and recommendation. For each, write the exact proposition and why it changes the decision. Separate sourced fact, direct experience, constructed example, and editorial inference. Headings should follow the decision sequence rather than keyword variants. This decomposition reveals when a single sentence secretly combines a documented feature, predicted outcome, and universal suitability, each needing different support.
Evidence: Federal Trade Commission
Assign an evidence standard before searching
Specify primary document, official guidance, independent study, qualified review, current policy, first-party observation, or contradictory case required for each row. FTC substantiation policy illustrates that the reasonable basis depends on claim type, product, consequence, cost of proof, and expert expectations; objective claims should have support before dissemination. An affiliate editor can adapt that discipline even when not the original advertiser. Do not lower the standard because the first search result is weak or the deadline is close.
Evidence: Federal Trade Commission
Run a source-family and contradiction plan
List likely organizations rather than favored URLs: regulator, program owner, manufacturer documentation, independent research, professional body, complaint source, and alternative provider. Search for evidence that would narrow or disprove the tentative answer. Record publication date, access date, methods, product correspondence, jurisdiction, conflicts, and limitations. SPJ’s ethics principles on accuracy, original sources, context, corrections, and independence provide useful editorial checks. Three pages repeating the same press release remain one source family.
Evidence: Society of Professional Journalists
Design the concrete output before prose
Choose a decision table, calculator, workflow, diagnostic tree, comparison matrix, checklist, or annotated example that lets the reader act. Define its fields and evidence links. For the variable-income question, an illustrative cash-reserve worksheet might include income timing, fixed obligations, tax estimate source, buffer, transfer rule, and professional-review trigger without giving individualized tax advice. An artifact exposes missing variables earlier than paragraphs do and prevents the outline from becoming generic explanation followed by a product link.
Evidence: Government Digital Service; Google Search Central
Complete the question-to-evidence outline
The finished document contains reader decision, direct answer hypothesis, exclusions, claim rows, evidence standard, sources to seek, contrary evidence, artifact, section purpose, reviewer, uncertainty, commercial disclosure location, related next question, owner, and expiry. Mark rows ready, narrow, unresolved, contradicted, or out of scope. Do not draft recommendation language for unresolved material rows. The outline should let another editor understand why every section exists and what evidence would cause it to change.
Evidence: Society of Professional Journalists; Federal Trade Commission
Stop research with a publish, narrow, split, or decline decision
A second editor traces claims to sources and challenges the most convenient inference. Publish when the core question can be answered within scope; narrow when evidence covers only a subgroup; split when two decisions require different mechanisms; decline when consequence or missing evidence exceeds authority. Record gaps and recheck dates. This workflow cannot guarantee rankings or completeness. Its success is a defensible reader answer whose evidence, limits, and next action survive the transition from planning into prose.
Evidence: Google Search Central; Government Digital Service
Sources and further reading
These references informed this article. A source supports a claim; it does not imply endorsement of TenMultigure or any future product reference.
- Learning about users and their needsGovernment Digital Service · Accessed August 10, 2026
Provides the evidence-based user-need and solution-neutral framing used to turn a keyword into a concrete reader decision.
- Creating helpful, reliable, people-first contentGoogle Search Central · Accessed August 10, 2026
Grounds intended-audience focus, original value, expertise, scope discipline, and the choice to narrow or decline search-driven topics.
- FTC Policy Statement Regarding Advertising SubstantiationFederal Trade Commission · Accessed August 10, 2026
Defines claim-specific prior-support logic used to assign evidence standards before research or recommendation drafting begins.
- SPJ’s Code of EthicsSociety of Professional Journalists · Accessed August 10, 2026
Adds independent accuracy, primary-source, context, correction, and editorial-independence principles to the source-family plan.
Reviewed by TenMultigure Editorial Review. See an error or a source that has changed? Tell the editorial team.
Review method: AI-assisted desk research with editorial checks. Reviewed ; next scheduled review . Built an eight-step question-to-evidence workflow with claim classes, precommitted standards, contradiction search, an artifact-first variable-income example, and publish, narrow, split, or decline stops.