For teams with an AI agent

What your agent can do with Keats

Each week your agent finds the questions where ChatGPT and the others leave you out, drafts the page or the outreach that fixes them, files the tickets and tells you what moved. You review and publish. Keats proves whether it worked.

For one product, that’s about 8½ hours of work a week turned into 30 minutes of review, by our estimate below.

The work it takes off your plate

01
Finds the gaps
It reads every answer the assistants gave this week and tells you, in a paragraph, where you're left out, who is named instead and which gap is worth fixing first.
02
Drafts the page or the outreach
It writes from Keats's brief: the questions to win, how the page should be built, and what counts as done. Where the fix is a review site or a list, it drafts the pitch.
03
Files the tickets
Each piece of work lands in your tracker once, with the brief attached, so nothing is filed twice and nothing slips.
04
Reports back what moved
At the end of the week it tells you what changed, what shipped and which results came in, ready to paste into a team update or a client email.

One week, with your agent

An example on the demo workspace of Prepco, a made-up exam-prep brand. The numbers and the conversation are Keats’s real output for it.

Monday
Keats asks Prepco's 24 buyer questions on five assistants: 312 answers.
Monday, 9:00
You ask your agent for the week. It reports Prepco named in 55% of buying answers against an aim of 80%, with Testwise at 62%, and picks the gap worth most.
Monday, 9:20
It drafts /sie/pass-guarantee/ from the brief and files 5 tickets, one per piece of work. You get a draft to review, not a blank page.
Wednesday
You edit the draft and publish it. Your agent tells Keats the page is live, which starts the clock on its result.
Friday
Your agent sends the week's summary: what moved, what shipped and what's next. It goes into the team channel as it is.
Day 35
An earlier change, a comparison page, gets its result: +35 points on the questions it targeted, against +2 points on the ones it didn't touch.

Monday morning’s conversation

You ask one sentence. Your agent does the reading and comes back with a plan and a draft.

You
Run this week's Keats review for SIE.

Reads this week's numbers from KeatsShow what came back

{
  "offering": "SIE exam prep",
  "answers": 312,
  "namedWhenBuyersChoose": {
    "percent": 55.4,
    "aim": 80,
    "changePoints": 0.8
  },
  "topRival": {
    "brand": "Testwise",
    "percent": 62.3
  },
  "belowAim": {
    "count": 18,
    "of": 24
  },
  "topCard": {
    "id": "33544d90…",
    "type": "new_page",
    "title": "Publish a page answering “Is there an SIE prep course with a pass guarantee?”"
  }
}

Your agent
Prepco is named in 55% of buying answers, against an aim of 80%; Testwise is named in 62%. 18 of 24 questions are below their aim. The top card is “Publish a page answering “Is there an SIE prep course with a pass guarantee?””. I'll get its brief.

Gets the brief for the top piece of workShow what came back

{
  "gap": "Named in 8 of 26 answers on 2 questions, aiming for 80%. Testwise and Passmark are named instead.",
  "page": "https://prepco.example/sie/pass-guarantee/",
  "targetQuestions": [
    {
      "text": "Is there an SIE prep course with a pass guarantee?",
      "rate": 23,
      "bar": 80
    },
    {
      "text": "What's the cheapest SIE prep that's good enough to pass?",
      "rate": 38,
      "bar": 80
    }
  ],
  "requiredStructure": [
    "One page per question: the H1 is the question (\"Is there an SIE prep course with a pass guarantee?\") or its closest natural title, at the brief's proposed path.",
    "Answer first: the first two sentences under the heading answer the question directly, with the brand named in them."
  ],
  "…": "quotes, pages cited, rivals named instead, acceptance criteria"
}

Your agent
Drafted /sie/pass-guarantee/ from the brief: the question as the title, the answer in the first two sentences, and a comparison table. It's in your review queue. Once it's live I'll tell Keats the URL, and Keats will read the result at day 21 and day 35.

The brief it wrote the draft from

Content brief

Publish a page answering “Is there an SIE prep course with a pass guarantee?”

Named in 8 of 26 answers on 2 questions, aiming for 80%. Testwise and Passmark are named instead.

Page: prepco.example/sie/pass-guarantee/

Target questions

Is there an SIE prep course with a pass guarantee?named in 23%, aim 80%
What's the cheapest SIE prep that's good enough to pass?named in 38%, aim 80%

How it should be written

One page per question: the H1 is the question ("Is there an SIE prep course with a pass guarantee?") or its closest natural title, at the brief's proposed path.
Answer first: the first two sentences under the heading answer the question directly, with the brand named in them.
Every number, price, date and requirement is checked against a primary source (the regulator, the exam body, the brand's own pricing page) and the source is linked.
Plain HTML text, not an image or a script-rendered widget, so the engines can read and quote it.
Then the supporting detail: a short comparison table where the question compares options, steps where it asks how, and a FAQ of the related questions.
Link it from the offering's main page and include it in the sitemap.

Done when

The change is live on the public site, indexable (no noindex, in the sitemap), at the URL reported.
The goal the evaluation reads against: "Is there an SIE prep course with a pass guarantee?": named in at least 80% of answers; "What's the cheapest SIE prep that's good enough to pass?": named in at least 80% of answers.

Named instead today: Testwise, Passmark, Kapstone.

Time

The hours it gives back

What a week of this takes for one product, by hand and with an agent doing the reading, drafting and filing.

Each week	By hand	With your agent
Read the week's answers312 answers at half a minute each, against a five-minute summary	2½ hours	5 minutes
Work out why you're left out18 questions below their aim at five minutes each	1½ hours	none
Write the fix for the top gapthree hours to write a page, against twenty minutes to review a draft	3 hours	20 minutes
File the work as tickets5 cards at five minutes each	25 minutes	none
Write up what moveda weekly update, against reading the one it writes	45 minutes	5 minutes
A week	8½ hours	30 minutes

Our estimate for one product with 24 questions, like the example. Your own times will differ; the minutes left are the ones that need your judgement.

You stay in charge

Your agent does the legwork. Every decision that reaches the public stays with a person.

Nothing goes live without you. Keats never publishes, sends or edits anything on your site, and your agent drafts for you to review.
It can't run up a bill. Your plan is a flat monthly price, and going past its questions is refused, never charged.
It sees one workspace. A key works only where it was made, and you can make it read-only.

Questions

Questions people ask first

What if I don't have an AI agent?: Then your team does the work from the app. Every gap comes with the same brief an agent would get: the questions it targets, what to write and what counts as done, ready to copy into a doc or hand to a writer. An agent saves the hours; it isn't required.
Which agents work with it?: Any agent that can use MCP tools, such as Claude, Cursor and Windsurf, and any script that can call a REST API. It takes about two minutes once you have an account.
Will my agent publish anything without me?: No. Keats never publishes, sends or edits anything on your site, and your agent drafts for a person to review. You decide what goes live.
How fast will I see results?: You see where you stand within your first week, once you've approved your questions. A change you publish takes longer: assistants pick up new pages over weeks, so Keats reads its result 21 days after it goes live and again at 35. An earlier read would mostly measure chance, and we'd rather tell you that than show you a number that means nothing.

Already have an account? Connect your agent.

Give your agent the AEO work

Request early access. Once you're in, connecting your agent takes about two minutes.

Request early access Sign in

The work it takes off your plate

Finds the gaps

Drafts the page or the outreach