Which analyses run in your Broadcastwell account?
Broadcastwell measures and fixes whether AI search engines name B2B software companies.
Run the free checkThirty-nine analyses run in a Broadcastwell client account, and all of them are available now. Each one reads your own delivered measurement, quotes the receipts it relies on and saves its result for your team to review. Nothing is published, sent or changed for you.
Updated 11 November 2026.
Every analysis below says what it reads, what it returns item by item and, where one applies, the label that ships with every result. A result screen shows when the attempt ran (UTC), its status, the measurement and the passes it read, and how many receipts it read. The receipts come from our measurement of ChatGPT, Claude, Perplexity, Google AI Overviews and Google AI Mode under our method, version 1.1, so an analysis is only as current as the delivered measurement it read.
Until your account holds a delivered measurement, each screen shows the analysis on Kalvenor Systems (Sample data, fictional company), labeled as an example; it is never a result about you. Once it does, the example stays one click away. Analyses that need outside search, ad, shopping, social, traffic or link data are shown on the fictional sample only and read none of your data.
The terms used here are defined in our AI search glossary. Our product documentation covers what you can do without an account, and our service model shows where this work sits in an engagement. The analyses run inside a Broadcastwell account, which you can open with the Free Account and 14-Day Tracking Trial or the Question Tracking Plan.
Read how the answers describe you
Answer tone review
How do the answers that name you describe you? Labels the tone of each answer that names you, engine by engine, and quotes the words in the receipt that set it. A label describes wording; it is not a measured rate.
What it reads. Answers that name you in your latest delivered measurement, up to 60.
What it returns, item by item. Engine; Tone; Why this label. Each item carries the receipt words it relies on.
Label on every result. Tone labels describe wording quoted from receipts; they are not scored and are not part of the Index.
Narrative themes
What story do the answers tell about your category? Finds the recurring themes across answers in your category, such as who is presented as the default choice and what advice buyers are given.
What it reads. Scored answers from your latest delivered measurement, up to 120.
What it returns, item by item. Theme; What the answers say; Answers carrying it. Each item carries the receipt words it relies on.
Tone change alert
Did the way answers describe you change since the last pass? Compares the two latest passes engine by engine. If you are named less, named more or described differently, it reports the change with the receipts from both passes.
What it reads. Answers from the two latest passes of your monthly measurement, up to 100.
What it returns, item by item. Engine; Earlier pass; Later pass; Change; Detail. Each item carries the receipt words it relies on.
Label on every result. A comparison of tone labels between the wording of two measurements, not a scored rate.
Missing brand context
When answers name you, what context about you do they leave out? Compares the answers that name you with your facts file and lists the context they leave out, such as who you serve or what a plan includes.
What it reads. Answers that name you in your latest delivered measurement, up to 60, and your published facts file.
What it returns, item by item. Missing context; Facts file reference; Answers without it. Each item carries the receipt words it relies on.
Entity and subtopic coverage
Which entities and subtopics do answers tie to your category that your facts file never covers? Lists the entities and subtopics the answers attach to your category, such as capabilities, buyer checks and vendors, how many answers carry each, and whether your facts file covers it.
What it reads. Up to 120 scored answers from your latest delivered measurement and your published facts file.
What it returns, item by item. Entity or subtopic; Kind; Answers carrying it; In your facts file. Each item carries the receipt words it relies on.
Entity relationship extraction
What do the answers state about who does what, as subject, relation and object? Extracts the relationships the answers to one measured question state, each as a subject, a relation and an object, such as a vendor and the buyers it is named for, with how many answers state it and the receipt quote behind it. It records what the answers say, not whether it is true.
What it reads. Every scored answer to the question you choose in your latest delivered measurement.
What it returns, item by item. Subject; Relation; Object; Answers stating it. Each item carries the receipt words it relies on.
Market analysis by segment
How are you named in each engine and question type, against the vendor that leads there? Breaks your latest measurement into segments, each engine and each question type, and reads each one: how often you are named, which vendor the answers name most and by how much. The counts come from every scored answer, not a sample of them.
What it reads. The named counts for every engine and question type across every scored answer in your latest delivered measurement, and up to 60 of those answers.
What it returns, item by item. Segment type; Segment; You are named in; Vendor named most; What it shows. Each item carries the receipt words it relies on.
Buyer-intent map
What is a buyer trying to do when they ask each of your measured questions? Labels each measured question with the buyer's intent, from learning about the category to checking terms before buying, names the content that answers it and shows your result on that question.
What it reads. One scored answer for each measured question in your latest delivered measurement, up to 35, and its per-question figures.
What it returns, item by item. Question; Buyer intent; Content that answers it; Your result. Each item carries the receipt words it relies on.
Report summary
What does this measurement say, in one paragraph you can forward? Writes a short summary of a delivered measurement. Every point carries its count and base, and each links to the receipts behind it.
What it reads. The figures and up to 40 answers from your latest delivered measurement.
What it returns, item by item. Point; Count and base. Each item carries the receipt words it relies on.
Scheduled monthly digest
What changed in your answers, written up after every monthly pass? Once you schedule it, runs after each new delivered pass of your measurement, writes what changed from the pass before with receipts, and keeps every digest as a dated history you can reread.
What it reads. Up to 100 scored answers spread across the passes of your latest delivered measurement.
What it returns, item by item. Pass; Named in; Change from the pass before; What changed. Each item carries the receipt words it relies on.
Ask about this view
Can you ask about the view you are reading and get an answer from its receipts? Answers one question about the account view you choose, using only that view's receipts, and quotes each receipt it relies on.
What it reads. Up to 30 answers behind the view you choose in your latest delivered measurement, and your question.
What it returns, item by item. Speaker; Text. Each item carries the receipt words it relies on, where one applies.
Check the facts and the sources behind them
Answer fact check
Do the answers say anything about you that your own facts do not support? Checks each statement an answer makes about you against your own published facts file and marks it supported, not stated or contradicted, with the receipt quote beside the fact it was checked against.
What it reads. Answers that name you in your latest delivered measurement, up to 60, and your published facts file.
What it returns, item by item. Statement in the answer; Against your facts; Facts file reference. Each item carries the receipt words it relies on.
Claim-level evidence check
Which statements in answers about you have a source behind them? Lists each statement the answers make about you and what stands behind it: an independent cited source, only your own page, only competitor pages, or no cited source, and whether your facts file states it. It reads the cited addresses, not the cited pages.
What it reads. Answers that name you in your latest delivered measurement, up to 60, with their cited sources, and your published facts file.
What it returns, item by item. Statement; What stands behind it; Cited beside it; In your facts file. Each item carries the receipt words it relies on.
Source relevance
Which cited sources matter for the question they were cited on? Rates each cited source high, medium or low for the question it was cited on and marks whether it is a third party, a competitor or your own page.
What it reads. Scored answers with their cited sources from your latest delivered measurement, up to 100.
What it returns, item by item. Cited source; Owner; Relevance to the question; Why. Each item carries the receipt words it relies on.
Question bias check
Does the wording of any measured question favor you? Reads the wording of every measured question and flags any that names you, leads the engine toward an answer, or uses your own product terms, with the reason. A favorable wording can raise how often you are named, so a flagged question should be reworded before the next pass.
What it reads. The wording of each measured question in your latest delivered measurement, one scored answer for each, its per-question figures and your published facts file.
What it returns, item by item. Question; Wording; Check; Why. Each item carries the receipt words it relies on, where one applies.
Clean and planted context compared
If a page planted a false claim about you, would an answer built from that page repeat it? Runs one defined test on one measured question: it answers the question from your facts file and the receipts alone, then again with a test passage you write added, and lists each claim that changed and whether the planted claim was repeated, resisted or not used. It shows how a model reading these sources treats the passage, not how any engine answers.
What it reads. Your test passage, your published facts file and every scored answer to the question you choose in your latest delivered measurement.
What it returns, item by item. Claim; From your facts and the receipts; With the test passage added; Planted claim. Each item carries the receipt words it relies on.
Label on every result. A model's reading of your sources, not an engine; it says what your sources support, not how engines answer.
Find the gaps and the next questions
Missing-topic detection
Which buyer topics have no answer that names you? Groups the measured questions where no answer names you into buyer topics, and shows who the answers name instead on each.
What it reads. The per-question figures and up to 40 answers that do not name you from your latest delivered measurement.
What it returns, item by item. Topic; Questions; Named instead. Each item carries the receipt words it relies on.
Follow-up questions
What will a buyer ask next after reading these answers? Lists the follow-up questions the answers prompt, such as the checks they tell buyers to make, and shows how you fare on the measured question closest to each.
What it reads. Scored answers from your latest delivered measurement, up to 60, and its question list.
What it returns, item by item. Follow-up question; Prompted by; Closest measured question; Your result there. Each item carries the receipt words it relies on.
Label on every result. Proposed follow-up questions, not observed answers; our measurements are single turn.
Competitor discovery
Which companies compete with the brand or site you enter, judging by the answers? Starts from a brand or site you enter and lists the companies your measured answers name beside it, with how many answers name each and whether its own site is cited, then suggests kinds of competitor the answers point toward. A suggestion is marked apart from a vendor the answers name.
What it reads. The brand or site you enter, and up to 120 scored answers from your latest delivered measurement.
What it returns, item by item. Company; Basis; Answers naming it; Why. Each item carries the receipt words it relies on, where one applies.
Sales-call questions and objections
Which questions and objections from your sales calls do AI answers already address? Reads one call transcript you are authorized to share, lists the buyer's questions and objections in it, and shows the closest measured question and your result there, with the transcript words and a receipt for each. Remove names and personal details before you paste it; the transcript is saved with the result for your review.
What it reads. The transcript you paste, the per-question figures and one scored answer for each measured question in your latest delivered measurement, up to 35.
What it returns, item by item. Kind; In the call; Closest measured question; Your result there. Each item carries the receipt words it relies on.
Draft pages, corrections and posts for your review
Evidence-backed content brief
What page would answer a question where you are absent? Drafts a page brief for one measured question: the sections a buyer needs, who the answers name instead, and the evidence each section must carry, all tied to receipts.
What it reads. Every scored answer to the question you choose in your latest delivered measurement, and your published facts file.
What it returns, item by item. Section; What it answers; Evidence it must carry. Each item carries the receipt words it relies on.
Third-party correction drafts
What would you ask a cited site to correct or add? Drafts a short request to a cited third-party site, grounded in the answers that cite it and in your facts file. Drafts are never sent; you copy, edit and send them yourself.
What it reads. Answers that do not name you and cite a third-party site, up to 30, and your published facts file.
What it returns, item by item. Site; What to ask for; Draft message. Each item carries the receipt words it relies on.
Label on every result. A draft for you to send from your own addresses; Broadcastwell sends nothing to anyone.
Outreach drafts for approval
What would you send to the sites answers cite, and who approves it before it goes? Drafts a first message and one follow-up to each third-party site cited where answers do not name you, grounded in those receipts and your facts file, and signed by the person you name. Your team sends a draft only after a member accepts it here with a note. We never send it.
What it reads. The person who signs, up to 30 answers that do not name you and cite a third-party site, and your published facts file.
What it returns, item by item. Site; Message; Draft message; Grounded in. Each item carries the receipt words it relies on.
Label on every result. Drafts you send from your own addresses; Broadcastwell sends nothing to anyone.
Repurpose one approved source
How would one approved passage read as an FAQ answer, an email paragraph, a sales note and a short post? Adapts one approved passage you paste into other formats: an FAQ answer, an email paragraph, a sales note and a short post. Each version keeps only facts the passage or your facts file states, and quotes the words it kept. Versions are drafts; nothing is published.
What it reads. The passage you paste, your published facts file, and the answers to the question you choose in your latest delivered measurement.
What it returns, item by item. Format; Adapted version; Facts it keeps. Each item carries the receipt words it relies on.
Social post drafts
What could you post from material you have already approved? Drafts short and long posts from approved material only: your published facts file and the Labs results you accepted. Every draft names its approved source. Drafts are never posted; you edit and publish them yourself.
What it reads. Your published facts file, your accepted Labs results, and up to 20 answers that name you in your latest delivered measurement.
What it returns, item by item. Length; Draft post; Approved source. Each item carries the receipt words it relies on.
Information gain check
Does your draft tell buyers anything the answers do not already say? Compares a draft you paste with what the answers to one measured question already say, and marks each point in the draft new, already said or in conflict, quoting the draft and the receipt.
What it reads. The draft you paste and every scored answer to the question you choose in your latest delivered measurement.
What it returns, item by item. Point in the draft; Against the answers; Detail. Each item carries the receipt words it relies on.
Hold drafts with uncertain facts
Is this draft ready to publish, or does a part of it need a source first? Checks a draft you paste against your facts file and the answers to one measured question, and marks each claim, name, price or date in it ready or held. If any part is held, the whole draft stays a draft, with the reason and the source that part needs.
What it reads. The draft you paste, your published facts file and every scored answer to the question you choose in your latest delivered measurement.
What it returns, item by item. Part of the draft; State; Why; Source it needs. Each item carries the receipt words it relies on.
Direct-answer passage
How would your page answer this buyer question in its first paragraph? Rewrites a passage you paste so it answers one measured question first, then gives the evidence and names the entities a reader needs, using only the passage and your facts file. It quotes the words it kept and lists what it left out for want of a source.
What it reads. The passage you paste, your published facts file and every scored answer to the question you choose in your latest delivered measurement.
What it returns, item by item. Part; Text; Kept from your passage. Each item carries the receipt words it relies on.
Page copy revision rounds
How should this page's wording change to meet your goal, one round at a time? Revises page copy you paste against the goal you set and the answers to one measured question. Each change quotes the original words, gives the revision and the reason, and keeps only facts your copy or facts file states. Run it again on the revised copy for the next round; every round is saved.
What it reads. The copy and goal you write, your published facts file and every scored answer to the question you choose in your latest delivered measurement.
What it returns, item by item. Original wording; Revision; Why. Each item carries the receipt words it relies on.
Business profile description
What should your business profile say, using only the facts you have published? Drafts a description for the profile you choose, within its length limit, from your published facts file only, and lists each fact it used and each fact it left out. It sets the draft beside how the answers that name you describe you, quoted from the receipts.
What it reads. Your published facts file and up to 20 answers that name you in your latest delivered measurement.
What it returns, item by item. Part; Text; Source. Each item carries the receipt words it relies on, where one applies.
Interactive resource ideas
Which calculator, checklist or comparison would answer the questions where you are absent? Suggests interactive resources, such as a calculator, a checklist or a comparison table, for the topic you enter, each tied to the measured questions where no answer names you, with the facts it would need from your facts file. It suggests; it builds nothing.
What it reads. The topic you enter, up to 40 answers that do not name you in your latest delivered measurement, its per-question figures and your published facts file.
What it returns, item by item. Resource; Kind; Questions; Why it fits; Facts it needs. Each item carries the receipt words it relies on.
Plan paid placement beside the answers
Advertising agent
Which measured questions are worth paid placement, and which should the answer win first? Reads your per-question figures, the answers that leave you out and the Labs results you accepted, and advises question by question: worth paid placement, win the answer first, or leave for now, with the reason and the next step. It advises only; it changes no campaign and spends nothing.
What it reads. The goal you write, the per-question figures and up to 40 answers that do not name you in your latest delivered measurement, and your accepted Labs results.
What it returns, item by item. Question; Advice; Why; Next step. Each item carries the receipt words it relies on.
Ad campaign drafts for approval
What would a campaign for your weakest questions look like, ready for someone to approve? Drafts campaign work for the measured questions where answers do not name you: an ad group per question, its search terms, an ad text and the page it should land on, tied to receipts and your facts file. Every draft waits for a member to accept it here with a note, and nothing is sent to an ad account.
What it reads. The approver you name, up to 40 answers that do not name you in your latest delivered measurement, its per-question figures and your published facts file.
What it returns, item by item. Question; Campaign part; Draft; What the approver checks. Each item carries the receipt words it relies on.
Label on every result. Drafts only; you apply any change in your own ad account; Broadcastwell applies nothing.
Advertising context from public sources
What context should an ad beside an AI answer carry, judging by the public sources the answers cite? Reads the answers that leave you out and the public third-party pages they cite, and suggests the context an ad placed beside such an answer should carry: the topic, what it must answer for the buyer, the public sources that set the terms and what your facts file supports. It suggests context only; it writes no ad and changes no campaign.
What it reads. Up to 40 scored answers that do not name you in your latest delivered measurement, the pages they cite, and your published facts file.
What it returns, item by item. Context; Questions; Public sources that set the terms; What the ad must answer; What your facts file supports. Each item carries the receipt words it relies on.
Model and plan, with the basis shown
Naming likelihood
How many answers are likely to name you on each question in the next pass? States, for each measured question, how many answers in the next pass are likely to name you, from the repeated passes so far, with the base it rests on. Its accuracy is unproven: every figure is saved so a later pass can be compared with it.
What it reads. Up to 120 scored answers and the per-question figures from your latest delivered measurement.
What it returns, item by item. Question; Answers likely to name you; Answers in the next pass; Observed so far; How it will be checked. Each item carries the receipt words it relies on.
Label on every result. A modeled estimate, not a measurement. It is shown with its basis and checked against your next delivered measurement, with that check shown beside it once it exists. Until that measurement is delivered, the result says: "Checked against your next delivered measurement once it exists."
Modeled AI influence
Beyond the visits your analytics credits to AI assistants, how many buyers might read an answer naming you without clicking? Keeps two things apart: the visits your analytics credits to AI assistants, which you enter, and a modeled count of buyers who read an answer naming you without a link to your site. The model uses your measured counts and one assumption you set; it is a model, not a measurement.
What it reads. The visits and the assumption you enter, and the per-question figures and up to 60 answers that name you in your latest delivered measurement.
What it returns, item by item. Part; Value; Basis. Each item carries the receipt words it relies on.
Label on every result. A modeled estimate, not a measurement. It is shown with its basis and checked against your next delivered measurement, with that check shown beside it once it exists. Until that measurement is delivered, the result says: "Checked against your next delivered measurement once it exists."
Run your team's own review work
Assistant workspace
What should your team work on next, with a first draft of each task already written? Brings your accepted Labs results, the fixes on your account and the receipts behind them into one working list. Each task names where it came from, the next step and a first draft for your team to edit. Nothing is published or sent.
What it reads. Your accepted Labs results, the points you chose to carry forward, the fixes on your account, and up to 30 answers that do not name you in your latest delivered measurement.
What it returns, item by item. Task; From; Next step; First draft. Each item carries the receipt words it relies on.
Carried context
What should every later analysis remember from the results you accepted? Condenses the Labs results you accepted and your review notes into a short list of points to carry forward. Once you accept the list, the assistant workspace and saved custom tasks read it as context; reject it and nothing is carried.
What it reads. Your accepted Labs results with their review notes, and up to 30 scored answers from your latest delivered measurement.
What it returns, item by item. Point to carry forward; From the result. Each item carries the receipt words it relies on.
Saved custom task
Can a task you write once run again on each new measurement? Runs a task you write, with its name and the answers it should read, on your latest delivered measurement. Every attempt saves the task with its result, and a saved task can run again on newer receipts without retyping it.
What it reads. The answers you choose from your latest delivered measurement, up to 40, the task you write and the points you chose to carry forward.
What it returns, item by item. Finding; Detail. Each item carries the receipt words it relies on.
How a result is checked
A person on your team accepts or rejects each saved result with a note before anyone relies on it, and every attempt runs once: a failed or uncertain attempt is saved and never repeated. Every quote must appear word for word in the receipt it names, or the result is not kept. Repeated runs of the same question can differ, which our note on measurement noise explains, so a result that reads one pass describes that pass.
When the day's allowance is used
Every analysis in an account shares one daily allowance. When it has been used, the screen says, word for word: Today's shared allowance for model-based analyses has been used; it resets at 00:00 UTC and nothing was charged. Nothing is reserved or charged for an attempt that this notice stops, and the attempt can run again after the reset.
Sources
- Broadcastwell, client account analysis screens and their exports at app.broadcastwell.com (sign-in required), read 9 November 2026.
- Broadcastwell, result labels in force from 9 November 2026, as quoted on each entry above.
- Broadcastwell, method v1.1.