Use case · Get chosen
Get named in “best” and comparison answers
When a buyer asks an AI assistant for the best tool for a job, the answer is mostly built from comparison pages and roundups. Attensira finds the ones that decide each answer, plans one next move per engine, and drafts it: a page of yours, or a pitch you send. Nothing goes out in Draft mode without your yes.
7 days free · card requiredHow the trial works
- “best AI search visibility tools”ChatGPT linked us/best/ai-visibility-tools
- “best AI search visibility tools for agencies”ChatGPT linked us/best/best-ai-visibility-tools-for-agencies
- “AI search visibility software”ChatGPT linked us/best/ai-visibility-tools
- “AI search visibility for marketing teams”ChatGPT linked us/best/ai-visibility-tools
For the first time since we started reading each question. A first is a reading, not a result we claim.
Analyst · Logged each first, verbatimattensira.com's own workspace, over the Attensira MCP · · n=4 firstsHow does the crew do this job?
You set the direction and answer a few questions. The crew does the rest, in this order, and every step leaves something you can open.
- Step 1Analyst
Measure
Reads the “best” and comparison questions you track on every engine your plan includes, and records who each answer names, in order.
Leaves: The answer grid
Docs: Is a prompt won or lost - Step 2
ResearcherFind what decides it
Builds the win plan: per engine, up to five deciding sources, which rung you are on (proven, better or new) and one next move, never a list of ten.
Leaves: The win plan
Docs: How to win one answer - Step 3
WriterDraft the move
Writes the comparison passage your page is missing, or Outreach drafts the pitch to the roundup that decides the answer, with the answer's own words as evidence.
Leaves: A draft, or a pitch in your mailbox
Docs: Pitches - Step 4Analyst
You approve, then it reads again
Once the change is live, the same questions are read again at +7, +14 and +30 days, against the week before. “Nothing we can prove yet” is a verdict, never dressed up as a win.
Leaves: A re-read verdict
Docs: Before and after
Has it worked on a site you can check?
We have no customers to quote, so the proof is our own sites: attensira.com and modeldiplomat.com, read over the Attensira MCP and committed with the date and n.
No receipt yet for this job
No comparison change of ours has gone live for a “best” question yet. The firsts beside the hero are readings: ChatGPT linking a page of ours for the first time since we started reading, not a result we claim.
How we measureWhat waits for your yes?
Nothing ships without your yes, unless you let a lane ship on its own.
A sentence that names a competitor is a claim, and a claim waits for your yes in every mode. So does every pitch: you send it.
- Every new workspace starts in Draft: the crew works inside the workspace and files everything else for your yes.
- A claim (a number, a price, a product or competitor name, a promise) waits for your yes in every mode, Act included.
- In Act, safe hygiene changes such as structured data, meta descriptions and alt text go ahead on their own. That lane is yours to switch on.
- A question the crew asks waits up to 7 days. Unanswered, the run closes without doing the thing.
Where the work lands
- A draft pull request labelled attensiraDocs: GitHub
- A draft in Webflow or WordPressDocs: Webflow and WordPress
- A pitch saved to your Gmail or Outlook drafts. You send it.Docs: Pitches
What do you see in the first week?
The 7-day Starter trial, card required, includes $30 of credits (3,000 credits): you watch this job's questions, and the agent drafts real fixes for them that wait for your yes. Outreach, automations and top-ups start after the trial.
- TodayCard checked, nothing chargedCheckout states the charge before it takes the card. The Starter plan fee is not charged until the trial ends.Docs: Plans
- Day 1First answers in your workspaceThe moment the trial starts, every prompt you track is read on every engine your plan includes, and you watch the rows fill in. The first report lands once half the reads are in, or after twenty minutes. After that, every prompt is read once a day.Docs: Your first reading
- Day 7The trial ends, and the plan startsThe Starter plan fee, $129 a month, is charged on day 8 unless you cancelled first. Then the plan's own credits start, and with them outreach, automations and top-ups. Cancel in the app before the trial ends and the plan fee is never charged. No call to book, no email to send.Docs: Your monthly budget
For this job, that means the questions behind “best X for Y, X vs Y and alternatives to X” are read every day of the trial.
What does this job cost?
You pay for a plan, and credits pay for the work. We never quote a price for one piece of work.
- The planStarter is $129 a month and includes 3,000 credits. Monitoring and reports are part of the plan.Docs: Plans
- CreditsCredits pay for the crew's work: chats, automation runs and ask calls. Reading your numbers and exporting them never spends one. The more credits you buy, the more work the crew can do, and you can buy more any time.Docs: Credits
- The ceilingA monthly budget you set is the most work can be charged in a month. New work pauses when it is reached. We do the work and charge only the credits it used, never a promise of outcomes or of a set volume.Docs: Your monthly budget
Can I run this from Claude, ChatGPT or Cursor?
Yes, over the Attensira MCP server. Reading your numbers is free; chats, automation runs and ask calls spend credits.
Take the “best” question we lose most and give me the win plan: per engine, which pages decide the answer, which rung we are on, and the one next move.
The tools this asks for
- get_prompt
- deciding_sources
- win_plan
What can't this do?
- It never forecasts a rank or a lift. Until a re-read says moved, the plan says nothing moved.Docs: What the plan will not say
- A roundup's editor decides whether to add you. A pitch carries evidence, not a guarantee.Docs: Pitches
- It never writes a pitch that goes out from an Attensira domain. Mail goes from your own mailbox, or not at all.Docs: Off-site rules
Questions about this job
Will it write about competitors without asking me?
No. Any change that names a competitor is a claim, and a claim waits for your yes in every mode, Act included. A change you edited before approving counts as a claim too.
Why do you ask for a card?
A card keeps the trial for real brands, so every reading in it asks the live engines, the same way a paid workspace does. Checkout states the charge before it takes the card, and the Starter plan fee is not charged until the trial ends.
What does the trial cost, and is the budget charged during it?
Starter comes with a 7-day free trial, card required; the plan fee is charged when the trial ends unless you cancel first. The trial includes credits of its own, and the agent's work in it is drawn from those, never from a budget: the monthly budget starts when the plan does. Outreach, automations and top-ups start after the trial.
What happens when the trial ends?
The trial ends. Unless you cancelled in the app first, the Starter plan fee is charged, the daily readings carry on, the plan's monthly credits start, and the crew keeps working inside the monthly budget you set. Outreach, automations and top-ups start after the trial.
Which assistants does this cover?
Starter reads ChatGPT, Perplexity, Claude and Google AI. Growth and Business add Gemini. Nothing else is called tracked.
Do I need a developer?
No. With the GitHub App installed, a change arrives as a pull request; with Webflow or WordPress connected, as a draft in your own editor; otherwise as new wording you paste in and mark shipped. If someone else owns the repository, the crew drafts the access request for you to send.
See who the “best” answers name instead of you.
Give your website. The trial reads your comparison questions every day, so the week shows who wins them and which pages decide it.
7 days free · card requiredHow the trial works
Tell it where you want to be cited.
7 days free · card requiredHow the trial works
attensira.com's own workspace · read 4 Oct 2026 · n=4,518
197/ 4,518 answers4.4%