Set project inputs and analysis scope

Which inputs define the business question?

Write the brand, approved website, product version, intended audience and the decision being investigated. Add the relevant market and language as separate fields. A company’s headquarters, the language of a prompt and a location supplied in a test are different facts. Define each field’s meaning and scope in the market configuration.

Keep a list of relevant alternatives and why they qualify for the buyer task. A neutral category question can reveal different brands from a named comparison. Preserve those as separate test designs. Do not assume that adding a competitor name to a worksheet automatically changes the product’s fixed measurement instrument or that every requested brand can be configured through a generic competitors tab.

What belongs to a scan and what belongs to a custom test?

A Rankfor project provides the website context for scanning. Open its completed engine scan in Rankfor Index and retain the represented dates and languages. The Index has a fixed instrument. Its headline is not a dashboard where arbitrary custom questions silently replace the underlying sample. Account entitlements and scan availability also constrain what can be executed.

For a question-specific investigation, open Answer Trail and use Your Prompt, optional Your Brand and optional Region. Select from the models that feature offers. For repeated variation, open Dice Roller Test, enter the exact prompt and retain the engine, Memory/Search mode and repeat count. These are explicit custom experiments with their own records, not edits to a completed Index scan.

Where do personas and content settings fit?

Review generated audiences in Discovered Personas and curate validated profiles in Persona Bank. Use Persona Prompts and the audience views according to the account’s available controls and caps. A profile supplies a planning context; it does not identify a real user. If a custom test needs a specific role or constraint, write that context explicitly into the prompt and preserve the full wording.

In Content generation > Content Strategy, choose the relevant Persona and Cluster. The run setup includes Content language, Narrative method and content types. Those inputs shape a content plan and its briefs. Record these settings with the content-plan decision they support and keep their scope explicit.

How do you keep the scope comparable?

Maintain an input register with value, purpose, owner, record type and version. For an illustrative fictional service, changing the targeted audience from office administrators to enterprise IT reviewers creates a new custom test context. It does not justify splicing the new results into the old audience’s series without a visible break. Keep exploratory questions separate from the comparison set.

The query-language and market research supports distinguishing language, location, category and interface. Use those distinctions when planning cross-market work. The public playbooks provide product workflow examples, while your input register records what was actually selected. Before each comparison, verify the available account controls and label any missing or changed condition so the result remains interpretable.