FIELDWORK 02 / 06

Build a page that earns its own URL

Decide what a reader can accomplish before choosing the title, structure and a separate URL.

Editorial illustration of developing sketches and a page structure into a useful page
Illustrative process: turn an answer into a useful page.
On this page

Teaching exercise · Synthetic examples · Live-site execution and learner validation NOT_RUN

Public-source research checked on 2026-10-10. This is a proposed exercise, not a completed audit or evidence of results. The example is wholly fictional.

The decision

What useful task will this page finish, and what can it contribute beyond existing answers? Complete a page brief before planning a large set of keyword pages. A template is an implementation choice; page count is not a value metric.

Inputs

Use one opportunity card, a specified reader task, a few existing alternatives, and sources for factual claims. State what you can demonstrate now. A proposed feature is not a tested feature.

Actions

  1. Define completion in reader terms: a usable answer, a comparison that supports a choice, or a valid output. Pick one concrete contribution: original analysis, a transparent comparison, direct experience, source-backed data, or a working tool. Google asks whether content adds substantial value beyond restating sources; Bing calls for clear content that meets intent. (S04, S05)
  2. Draft the first screen: task-specific title, immediate useful answer, and a relevant next action. A title promises what the body and tool actually deliver. Write accurate distinct metadata; title and description character counts are not universal pass/fail limits. Search engines can select different display text. (S25, S26)
  3. Create a claim ledger: claim, source, checked date, test scope and limitation. State how comparisons were made. Attach no “tested,” “best,” “fastest,” expert-review or user-test labels without corresponding evidence.
  4. For a page family, write the per-page distinction: different task, source data, calculation or output. Ask what remains if the variant name disappears. This is an editorial heuristic, not an official ranking test. If only the noun changes, consolidate the answer or keep the variant in one page. Scaled publishing aimed at manipulating rankings without value can violate Google's spam policies. (S06)
  5. Review the draft with an uncoached intended reader when you have consent and access. Ask them to complete the task; record what they did, where they got stuck, and whether the output was usable. Until this happens, learner understanding remains NOT_RUN. A checklist cannot substitute for that observation.
  6. Revise the smallest part implicated by the evidence. A reader misunderstanding one instruction does not prove the whole market, brand or topic is wrong. Keep the previous version and explain what changed.

Worked output — fictional

This independent mock brief explores a rain-commute packing page for SpokeNote; it does not reverse the earlier NO_GO without new real evidence. The page would let a reader edit a checklist and print it. Proposed title: “Make a printable rain-commute packing list.” The draft includes an editable example, a plain-text alternative and explicit exclusions. It makes no claim about preventing accidents, equipment suitability or search demand. “Weekend list” gets no separate URL if it merely swaps a heading.

Deliverable and evidence

Submit the page brief, a claim ledger, one proposed URL, the distinct contribution, a completion definition, the actual function checks when available, and reader observations when run. Mark the mock output SYNTHETIC and unfinished capabilities PROPOSED.

The editorial pass condition is that the title, answer and output serve the same task and supported claims can be traced. Functional readiness, reader comprehension, indexing and performance are separate outcomes.

Failure limits

No fixed word count or number of pages guarantees quality. E-E-A-T itself is not a specific ranking factor; do not manufacture a score. A polished draft does not prove ranking, revenue or AI citation. Do not invent experience, reviews or publication dates to fill gaps.

Bounded agent task

“Draft one task-led page brief from this opportunity card. Specify the unique contribution, claims and sources, promised output, variant consolidation, and checks still required. Treat features, examples and reader tests as unrun unless supplied evidence proves otherwise.”

Public sources

Keep a blank record for this exercise

Header-only CSV. No preset results. Machine fields are shared across languages; preserve unknowns without evidence.

Download blank CSV ↓

Back to lesson 2 · Back to lesson 3