One conversation tells you what ChatGPT cited once. A hundred tell you what it cites for your market. The Bulk Analysis dashboard aggregates every conversation you have captured into a single view, and exports any part of it.
Purpose
- Aggregate every captured AI conversation into one view: domains, URLs, fan-out queries, products, entities, and brand share of voice.
How to Open
- Popup nav →
AI Insights→ Bulk analysis. - The button is available from every state of the AI Insights tab, including the “unsupported platform” screen, so you do not need to be on ChatGPT to open it.
- It opens in its own browser tab. Opening it again focuses the existing tab rather than making a second one.
The page has two segments: Run and Dashboard. Run drives new captures and is covered in Bulk Prompt Analysis. This page is about the Dashboard.
Filters
Three controls sit above the dashboard and apply to every section below:
- Scope:
This runorAll history. Defaults to All history. - Project:
All projects, or one project. See Projects. - Platform:
All platforms, or one.
Tables have their own page size (10, 25, 50 or 100 per page) which is remembered between visits, and Top domains and Top URLs each carry an independent set of source-category chips so you can look at citations in one and product links in the other.
What You See
Overview
Tiles across the top: Conversations, Links, Fan-out queries, Domains, Carousels, average links per conversation, and average thinking time.
Underneath, coverage bars showing what proportion of your captures contained each kind of data: Fan-out, Images, News, Shopping, Entities, Places. This is the fastest read on whether a prompt set is actually producing the signal you want.
Sourcing pipelines
A doughnut chart splitting every source by the retrieval path it arrived through: Bright Data, Oxylabs, Labrador, or the open web. See Sourcing Pipelines for what each one means and why the split matters.
Reference types
A breakdown of the reference types behind the answers.
Saved conversations
Every capture, searchable by prompt or title. Select rows to add them to a project, remove them from one, or delete them.
Fan-out tag cloud
A packed, size-weighted cloud of the terms across every fan-out query, filterable by All, Search, Images or Products. Reads like a keyword list written by the retrieval system rather than by a keyword tool.
Top domains and Top URLs
Which sites and which pages get pulled into answers, how often, and across how many conversations. Filter each independently by source category: citation, news, image, academia, product, hidden, or other.
Top products, images, news and places
Products with merchant, price, rating and review count. Places open a detail panel with address, phone, categories, rating, price band, opening state, reservations, menu and data provider.
Top entities
The named entities appearing across answers, with product entities called out separately.
Brands and competitors
Your tracked brands against your tracked rivals, with total mentions and share of voice. Set them up in Manage Brands & Competitors. See Brands and Competitors.
Captured data
A raw viewer for the underlying capture, for when you want to check what the dashboard is summarising.
Exports
Every section exports on its own from the Download card: Conversations, Links, Domains, Fan-outs, Products, Images, News, Places, Entities and Brands, each as CSV. Individual sections also carry their own Export CSV button.
Two whole-dataset options sit alongside them:
- Export all (JSON): the complete aggregate behind the dashboard.
- Backup (JSON): the raw captured conversations, for archiving or moving to another machine.
Clear all data deletes every captured conversation. It cannot be undone.
There is no PDF or Excel export. CSV and JSON only.
Workflow Wins
- Answer “which sites does ChatGPT actually trust in our category” with a number instead of an anecdote.
- Show a client their share of voice against named competitors across a real prompt set.
- Turn the fan-out tag cloud into a content brief that reflects how the model decomposes the topic.
- Spot that a competitor owns the product carousel while you own the citations, which are two different problems with two different fixes.
Recommended Plays
- Category baseline: Run 50 buying-intent prompts for a category, then read Top domains with the citation chip on. That is your competitive set in AI search.
- Share of voice tracking: Add your brand plus three competitors, run the same prompt list monthly, and compare the Brands card.
- Retrieval diagnosis: Check the sourcing pipeline split before deciding on a fix. If most answers come from licensed content, being more crawlable will not move you.
- Product visibility: Filter Top URLs to the product category and see whose PDPs are getting pulled in.
Tips & Troubleshooting
- Dashboard is empty? You need captures first. Either browse ChatGPT normally with AI Insights open, or start a run from the Run segment.
- Brands card empty? Add brands and competitors, then use Reprocess all conversations. Reprocessing re-scans everything you have already captured against the new brand list, so you do not need to run the prompts again.
- Changed a filter and a section looks stale? Use Refresh. Only the section whose filter changed repaints, which is deliberate, but Refresh redraws everything.
- Storage. Bulk captures can get large, which is why the extension asks for unlimited storage. Everything stays on your machine.
Frequently Asked Questions
AI Insights: Common Questions
The Brands card shows no mentions. What’s wrong?
Nothing is scored against your brand list until you press Reprocess all conversations in the Bulk Analysis dashboard.
Add your brands and competitors first, then Reprocess. It re-scans everything you have already captured against the current list, so you do not need to re-run any prompts. If counts still look low, add more aliases, the misspellings, the legal entity name, the bare domain, and Reprocess again.
Why did my bulk run stop part-way through?
It almost certainly auto-paused, which is deliberate. A run pauses after 3 consecutive hard failures, or after 5 consecutive answers that never got persisted. Both usually mean ChatGPT rate-limited you or your session expired. Sign back in and press Resume rather than restarting the batch.
Items marked Timed out exceeded the 180-second per-prompt limit. Use Retry unfinished to re-queue just those.
Keep first batches small. Twenty prompts will tell you whether your list is well-formed before you commit to two hundred.
What is a sourcing pipeline, and why does it matter?
It is the retrieval path ChatGPT used to fetch a given source. ChatGPT records one of four and the extension reads the label directly: Bright Data and Oxylabs are commercial web scrapers, Labrador is an allowlist of licensed publishers such as Reuters, WSJ, Wikipedia and arXiv, and Open web / SERP is the search baseline.
It matters because it tells you which lever would actually move you. If answers in your category come mostly through the scrapers, being fetchable and answering the question in the page body is the work. If they come mostly through Labrador, you will not optimise your way in and the lever is coverage by those publishers instead.
Can I capture an old ChatGPT conversation, or only a live one?
Old conversations work. Extraction reads the conversation from ChatGPT’s own backend using the session you are already signed in with, so you can open a chat from last month and capture it in full.
Two things are the exception. Turn intent and live merchant offers are only caught while the answer is streaming, so they will not appear on a conversation you reopen later.
Gemini is different: its capture listens to live network traffic. Reopening an old Gemini conversation produces no traffic to listen to, so it falls back to reading the page and finds fewer queries. Capture Gemini conversations while you are having them.
Which Google domains does AI Overview capture work on?
Currently 17: google.com, .co.uk, .nl, .de, .fr, .es, .it, .ca, .com.au, .co.jp, .com.br, .co.in, .be, .pl, .se, .at and .ch.
On any other Google domain the AI Insights tab will not pick up an Overview, which is the most common reason it looks empty on a SERP that clearly has one. This is being widened to every Google country domain in the next release.