OPERATING SEQUENCE
Use the AI Search Operating Manual
Place the map inside a repeatable baseline, evidence, rewrite, and retest loop.
Open the manualBOFU PLANNING · SIX DECISION CLUSTERS
Map the questions a serious buyer might need answered.
Create an editable hypothesis set for comparison, alternatives, evaluation, risk, implementation, and vendor discovery, then move it into your evidence workflow.
INPUT DOSSIER
Define the decision context
Use plain category and role language. You can edit every generated line before copying or exporting it.
Named competitors or alternative categories
Optional. Add up to three public comparison anchors; do not enter confidential deal data.
Buying-stage emphasis
All six clusters remain in the map. Emphasis only changes which clusters appear first and receive priority labels.
Inputs and edits stay in this browser. Analytics receives only the selected emphasis ID and aggregate counts: never your text.
Turn hypotheses into an evidence workflow
Generation is the starting point. Use the operating records and human review to decide which hypotheses deserve observation or publication.
WORKING FILES
Move rows into the Operator Templates
Use the Prompt Coverage CSV to assign engines, evidence, owners, and observation status.
Open the templatesPAID EVIDENCE REVIEW
Scope a 30-day GEO audit
Have Tenten validate the decision landscape, public evidence, and accountable next actions.
Review the auditWhat this builder does, and does not know
Are these the prompts buyers actually use?
No. They are deterministic planning hypotheses created from your inputs. Validate wording through sales, research, controlled observation, and documented evidence.
Does the map show search volume, live frequency, or AI visibility?
No. The tool has no search-volume data and does not query any model. It cannot measure prompt frequency, mentions, citations, rankings, preferences, or buyer behavior.
Is the text sent to Tenten or an analytics provider?
No. Generation, editing, copying, and CSV export happen locally. Analytics events contain only stable action IDs, the emphasis ID, competitor count, and candidate count.