Before building a service page, collect evidence that the service, buyer job, proof, and delivery path are real. A keyword list can show language; it cannot confirm that the business serves the audience, has a distinct offer, or can turn a request into a profitable outcome. The Service-Page Evidence Worksheet turns customer and sales knowledge into a release decision.
Google’s people-first guidance asks whether content is made for a real audience. Search traffic guidance can inform query and page research, while Salesforce’s Campaign Influence guide illustrates why sales and opportunity relationships need configured evidence. None replaces direct service-owner validation.
State the page decision
Write the candidate service, audience, market, buyer job, proposed route, owner, effort ceiling, and decision date. Choose among NEW_URL, UPDATE_EXISTING, MERGE, HOLD, or REJECT. A unique phrase is not enough to justify another page.
Record the nearest existing URL, canonical intent, service page family, and any known overlap. This prevents research from quietly becoming a mass-production queue.
Collect buyer language
Ask customers, prospects, support, sales, and delivery owners what problem they were trying to solve, what words they used, what alternatives they considered, what risk delayed the decision, and what evidence helped. Keep direct quotes permissioned and avoid unnecessary personal data.
Group language by decision job: understand, diagnose, compare, implement, verify, or request. Preserve contradictions; they may show different audiences or service boundaries rather than a need for a longer page.
Collect fit and rejection evidence
Define an accepted request: account or person fit, use case, geography, role, urgency, budget or value, compliance, and next step. Collect rejection reasons such as wrong territory, wrong buyer, unavailable service, weak timing, missing information, or unclear offer.
Do not ask sales only for “good keywords.” Ask what makes a request serviceable and which evidence is required before a conversation becomes an opportunity.
Collect offer and delivery facts
Record service scope, deliverables, process, time, price logic, dependencies, exclusions, capacity, staff, tools, locations, and escalation. Mark each fact verified, directional, or unknown. A page should not promise a route that delivery cannot support.
Ask what changed recently: offer, team, market, regulations, case studies, pricing, support hours, or availability. Add an expiry trigger to facts likely to change.
Collect proof and permission
List case outcomes, before-and-after evidence, process examples, credentials, demonstrations, references, customer language, and approved limitations. Record source, owner, date, permission, and intended claim. If proof is missing, design a research task rather than filling the page with generic authority.
Keep negative evidence too. A failed implementation, unavailable location, or common objection can help a buyer self-qualify and reduce unserviceable requests.
Collect page and journey evidence
Map current queries, landing pages, internal links, forms, service hubs, comparison pages, and CRM paths. Identify what the existing site answers, where visitors stop, which page owns the decision, and which route is missing. Use a small sample of sessions or records where approved.
Separate navigation, content, form, routing, and capacity breaks. A new page is not the right repair when an existing page or handoff is the first failure.
Collect commercial evidence
Record accepted rate, response time, opportunity rate, close or mature outcome, value or margin, sales effort, delivery cost, and minimum viable volume where available. Label sample, period, definition, and maturity. A small high-fit cohort may justify a pilot; it does not support a precise forecast.
Ask what decision the page should inform: demand discovery, service qualification, sales enablement, expansion, or retention. Without a decision, the page will be judged by traffic alone.
Complete the worksheet
| Evidence area | Required note | Status | |—|—|—| | audience and job | buyer, problem, decision, language | verified/directional/unknown | | fit and rejection | accepted criteria and reasons | verified/directional/unknown | | offer and capacity | scope, limits, delivery | verified/directional/unknown | | proof | source, permission, expiry | verified/directional/unknown | | current route | existing URL, links, form, CRM | verified/directional/unknown | | commercial result | sample, stage, value, maturity | verified/directional/unknown | | page decision | new, update, merge, hold, reject | named owner |
The lowest-confidence row is the next research task. It is not a reason to publish a page with an invented answer.
Run the pre-brief gate
Review the worksheet with service, sales, content, SEO, analytics, and web owners. Ask what evidence would change the route decision, who can approve claims, what the page must not promise, and how the request will be handled. Set a pilot, effort ceiling, stop rule, and recheck date.
Keep the worksheet with the brief, source register, visual plan, and QA record. Update it when the offer or evidence changes.
Do a short evidence review with the person who can approve the service promise and the person who owns the commercial outcome. Ask which fact is most likely to expire, which customer language is permissioned, and what would make the team choose UPDATE_EXISTING instead of a new URL. Preserve rejected candidates and the reason for the decision; they prevent the same thin idea from returning under a different keyword.
Verdicts
Ready for brief: buyer job, distinct route, evidence, serviceability, proof, owner, and handoff are clear.
Update existing: a current page owns the job and needs focused repair.
Merge: candidates share the same decision and evidence.
Research first: one or more critical facts are directional or unknown.
Hold or reject: no serviceable offer, proof, owner, or decision exists.
The Service-Page Evidence Worksheet is complete when it records buyer language, fit, rejection, offer, capacity, proof, current route, commercial result, route decision, owner, and next test.
How did this article land?
Choose one reaction. You can change it anytime.
