Backstage! · Podcast research & planning
The work behind a great conversation.
Discover guests, prepare source-linked interview questions and build a timed run sheet for your next episode.
Explore regional cultural signals, then compare your imported audience reports before and after an episode. Observed changes do not establish what caused them.
Open worked example →Source-linked brief, timed agenda and clearly labelled synthetic regional results
Preparing your workspace from the saved study…
Find an unexpected guest.
Explore guests for your episode, then build the interview.Choose a guest. Build the interview.
References
Profile & provenance
Your workspace
Your first episode
Your next episode starts here.
Start with your show and episode goal, or try a recorded study. The example uses captured Qloo evidence; any sample analytics are labelled synthetic.
These shortcuts make no Qloo or model requests.
Regional audience goalOptional · no baseline saved
Choose a target before comparing guests. Channel activity, producer objectives and Qloo cultural signals remain separate. No guest-growth forecast is calculated.
Includes saved case history. Unfinished edits are excluded; source notes and daily reports use their own exports.
Your next step
Choose the episode goal
Outstanding review
Qloo · cultural discovery
Cultural signals
Guests in this study
YouTube counts are rounded channel snapshots checked 2 Nov 2026, not active podcast listeners. Unknown counts stay unknown.
Regional focus uses cultural affinity to compare each recorded signal with itself across four sample windows on shared cells. It is not measured listenership or a global popularity rank. Some creators use an explicitly labelled associated-show signal.
How the categories work
Editorial visibility categories: flagship personalities, established specialists and independent expert voices. Sources explain each profile. Categories do not measure listener counts, Qloo affinity, booking cost or willingness to appear. Availability needs confirmation.
Build a regional guest mixQloo evidence · three creator categories
Optional regional visibility mix
Set an episode question first. This optional experiment then applies regional evidence and a one-per-category rule. Book breadth is a research aid, not an interview-quality score.
Fixed researched guest menu. Past guests and local rejections are excluded. Categories describe editorial visibility; they do not measure audience size or availability.
Viewing a proposal leaves your shortlist and interview notes unchanged.
Saved proposals in this workspace
Latest eight frozen proposals. Downloads preserve their original evidence and exclusions.
No people match these filters. Try a name, show or broader topic, or clear the filters.
AI-illustrated portraits · not photographs
Identity corrections and sources
Geographic context
Guest geography
Cultural observations, not guest locations or audience counts.
Drag to pan; use + / − to zoom. Keyboard: arrows to pan, + / − to zoom. Navigation does not fetch new Qloo cells.
Land: Natural Earth · local public-domain coastline, generalized; no street or venue data. Points: Qloo cell centers, ordinal bands within the selected query. No audience counts or cross-person score comparison.
Show context & captured references
Current show · host-authored context
Geographic evidence is about cultural signals. It does not locate guests or establish residence, availability, audience size or attendance.
Capture limits and missing information
Interview angles and questions are authored examples, not Qloo outputs or live model generation. No outreach is sent.
Show history & episode goalPublic feed or reviewed XML
Read a podcast feed
This Week in Startups is the suggested public example. Confirm the show identity before use.
Next episode
These are your declared criteria. Availability and other guest attributes are not automatically verified.
Your shortlistNo guests pinned
Your guest mix
Choose a guest in Discover and use Pin to shortlist to start. Your source notes and unfinished drafts stay with each identity.
Compare all guests
Guest decisions
Compare the evidence
| Guest | Cultural fit | Geography | Interview sources | Decision | Editorial case | Review guest |
|---|
No identities match these filters. Your research is retained.
Ranks apply to one query. Cells are observations, not audience counts. Source approval records your review.
Decision support
Compare finalists
Choose up to three reviewed identities. Cultural rank comes from one fixed-menu query; geography and editorial judgment stay separate.
Select at least two guests to build a comparison.
What does each cultural reference change?Qloo reference test
Hold the guest menu fixed. Compare the full brief with one cultural reference withheld at a time.
02 / Geographic evidence
The same study area
Show references
Highest thirdMiddle thirdLower thirdOrdinal bands within each displayed map
Coordinate grid, not a street map. Equal-size points show observed cell centers. Colors do not measure audience size or differences between people. Missing cells stay unknown.
Inspect a cell and its source
Show / Cultural references
Show & cultural references
Confirm cultural references
Select up to four exact books, films, artists, brands or other supported entities. Resolve namesakes before adding a signal.
Compare one guest menu
Past guests are excluded locally. Discovery suggests prospects; a comparison ranks only the confirmed people you selected in one query.
Optional / Live person research
Find and confirm another guest
Checking whether live search is enabled. No Qloo request is made automatically.
Every search requests person records. Occupation context checks only the tags in the returned matches; it is not a comprehensive upstream category filter. Namesakes remain visible.
Identity confirmation does not establish availability. Geography uses only the fixed corridor. The reference map must match the current confirmed show references.
Returning to the captured study hides added evidence and preserves your edits. These controls do not reset the server request budget.
03 / Interview preparation
This prevents pinning/exporting the guest. In an edited show, the comparison menu changes and its previous ranking becomes historical. Saved source evidence remains.
Past guests · saved local exclusions
Qloo cultural-fit evidence
Authored editorial example · editable
These freeform notes have no question-level source review. Use the source notebook below to attach premises and approve questions. Your edits do not change Qloo results.
Producer notes
The editorial case
References behind this result
Qloo source contributions within this result. They do not measure audience shares or establish the guest’s opinions.
Why this guest for the target region?
Original-source preparation
Research
Fetches this public URL only. No web search or model runs. Source text and its relationship to the guest require your review. Previews and passage choices stay with each guest; switching guests or replacing a preview requires a fresh check.
Paste source text or import a text file
For sources you can read but this app cannot retrieve. The URL is a citation you supply; the text is not checked against it.
Prepare a question from a saved note
Read the saved source note
Templates use your focus and premise. No model execution or automatic approval. Edit and review the drafts in the existing notebook.
Anthropic · evidence workflow
Research the interview
Choose reviewed source notes. The agent inspects the confirmed guest’s saved Qloo evidence, then proposes a cited case and tailored questions.
Authorship is producer-declared. Review each source’s details before attributing a passage to the guest; leave uncertain authorship unknown.
04 / Source notebook
Interview sources
Keep the fact, its source and your interview angle separate. Approval records your editorial review; it does not certify the source.
Two concise notes from Norman's original article, checked in prior project research. Manually entered example, not retrieved by this app. Questions start as drafts.
Public source records
Interview questions
Qloo captures, where available, remain separate from interview sources. Source notes are manually entered or explicitly retrieved and reviewed; receipts identify retrieved documents. Interview angles are editorial interpretations. Source content is not executed. No outreach is sent.
Episode production
Interview run sheet
Arrange saved questions into a timed working agenda. Question wording and source revisions are captured when added; refresh them explicitly after edits.
Opening
Add saved questions to begin. Draft questions keep their review status.
Closing
Producer reviewFrozen participant packet · no ratings yet
Share the participant file for a review with Qloo ranks and input order withheld. This workspace itself displays those ranks. The single-menu review freezes your selected finalists and has no second arm. The comparison mode uses an independently supplied control declaration and a matching captured Qloo shortlist; it does not generate a control or verify independence. Export key & ratings includes previous saved packets. No benefit claim follows from a packet alone.
Use a control shortlist collected independently for the same saved brief, constraints and eligible menu. This workspace displays Qloo ranks; send the participant file to an independent reviewer. Control origin and reviewer independence are declarations, not verified facts.
Source operations · actual requests & file imports
Episode results
Save a guest hypothesis, then review your channel’s regional activity.
Your audience evidence
Bring your YouTube data
CSV import · no Google account connected
Choose an exported daily report, review its columns, then compare your channel’s activity. The file is processed in this browser.
Study setupGuest, target region and publication
Guest & hypothesis
Choose a confirmed guest or load the real Qloo capture.
Saved snapshots
Publication & period
Both periods use channel-level activity. The publication day begins the after period; days are Pacific calendar dates.
Observed channel activity
Observed change, not attribution. Qloo scores stay separate.
Geographic context
ReportsLocal CSV import · no account connected
Import YouTube reports
Nothing is uploaded. Import a daily channel report for the target state, then optionally a whole-channel report for context. No authentic channel telemetry is bundled.
1. Export a daily report from YouTube Studio
- In YouTube Studio → Analytics → Advanced mode, select the entire channel and the date range covering both periods. Choose a daily breakdown.
- Apply the target-state geography filter if your report supports it. Select Views and/or Watch time. Export the current view as CSV.
- Review the exported report’s filters, units and completed dates. Do not label country-only, video-only, realtime or incomplete data as a state-level channel report.
Studio documentation supports export, but does not guarantee a state selector or fixed CSV headers for every account. Map the actual column names below. If your export contains country totals or video rows, it cannot be used as a daily state report. State-filtered daily reports are documented in the Analytics API; direct Google sign-in is not enabled here.
Practice with invented data
Two sample CSVs use 13–26 October 2026, with an example publication on 20 October. Expected regional views: 700 before → 1,050 after. Whole channel: 7,000 → 7,700. These are test figures, not YouTube results.
Use the separate walkthrough above to practice without changing your main workspace.
Privacy and saved reports
This CSV flow sends no file contents to Backstage’s server, Google, Qloo or the application AI. No Google password, token or account permission is needed.
After you confirm, normalized daily figures, your channel label, filename and declared report scope are saved in this browser profile. Extra CSV columns are discarded. This storage is not encrypted by Backstage; someone with access to this browser profile may be able to read it.
Exported packets can include daily figures and source metadata. Check them before sharing. “Remove report” hides a report recoverably; the button below permanently deletes saved CSV rows and previews in this workspace. Saved study/planning totals, other workspaces and files already downloaded remain.
Saved on this device. Exports include sources and coverage.