Skip to content

Act on results

To-do

Each item on your To-do list is one change to make. Items are ranked by how many of your questions each could help you win.

It might be a site fix, a new page, or a site to get cited on. It might also be a rival's win, or a wrong fact AI models say about you.

Where to find it

The same items are the rows of Improve, in five sections. The Improve card on the Overview shows the top few, and View all in Improve opens the rest.

Your AI reads the whole list through its connection to remake, best first. Reading it and marking items costs nothing. It's built from your stored runs.

What lands on the list

The list is rebuilt from your runs each time it's read. It holds 6 kinds of item.

Kind What it means Most at once
Rivals A rival is named on prompts where the AI models never name you, in your latest run. 3
Get cited A site the AI models often cite for your prompts that isn't yours or a rival's, across your stored runs. 3
Site fixes A website check that failed or is almost there on your latest run. 1 per check
New pages A page your fix suggestions say to write, which your site doesn't have yet. 4
Wrong facts Something the AI models say about you that looks wrong, from Perception's fact-checking. 3
Your fixes A fix your own AI assistant wrote and added to the list, tied to the questions it's meant to win. 50 per brand

Every flagged website check adds an item, named as My website names it. For example: Show your phone and email, Add a sitemap, Make your pages load faster or Describe your images.

A check with a ready-made change shows once, as that change. A check that passes adds nothing.

Each title says what the item is. A rival item reads like "Apple wins questions you miss". A get-cited item reads like "gsmarena.com takes 34% of the source links".

How the list is ordered

Each item says why it matters, with the number it's ordered by inside the sentence.

Kind What the item says, for example
Wrong facts "AI said this in 2 runs · your fact: two campuses, BKK1 and Toul Kork". These come first.
Site fixes "Affects every page AI reads · No business information tags". The fix changes how the AI models read your whole site. Those with High impact come next.
Rivals "3 of your 12 questions name them, not you".
Get cited "1 of your 5 questions · cited where Soundskool is named, not you".
New pages "Answers 3 questions you're missing, e.g. “best piano lessons for kids”".

More questions rank higher, and ties go to the item where more AI models miss you.

A new page that answers none of your questions says what it rests on, like "Comes up in 4 AI answers". It ranks after the items with questions. One with no evidence at all says Idea from your site review and comes last.

Each item also carries an effort chip: how much work the change takes.

Effort Items
Low Write a clear page title, Write a page description, Let AI crawlers read your site, Show your phone and email, Add a sitemap, Describe your images
Medium Add business information tags, Add a questions-and-answers section, every Rivals item
High Make your pages load faster, Add more text to your pages, every Get listed and Articles & pages item

Item statuses

Status What it means
New Not started yet.
Doing Work on it has started.
Done Your AI marked it done. remake shows what moved since. If your latest run no longer shows the reason for it, it reads Done · no longer showing: the problem went away. A started (Doing) item that goes away this way counts as Done too.
Dismissed Put aside. Your AI can put it back on your open list.
Expired It was only started or noted, and your latest run no longer shows the reason for it.

Open counts both New and Doing items.

Work an item with your AI

Your AI works the list for you. Ask it to make the next change on your To-do list.

  1. Your AI reads the item whole: why it matters, its questions, its steps and your notes.
  2. It makes the change with its own access to your site.
  3. It marks the item Doing or Done, or Dismissed if it doesn't apply, and can add a note.
  4. When it marks the item Done, remake confirms the change is live at once and answers found or not found. That check costs the Fix verification price. A listing on another site costs the Outside page check price per page. See Improve.

Your AI can also add a change it made, which then gets its own "since" line. See What your assistant can do.

Once an item is done, remake says what moved since that day. For example: "Since 5 Oct: named on 0, 1, then 2 of these 6 questions". It reads the run before that day and up to 5 runs after.

Before your next run it reads "Marked done 5 Oct · first result after your next run". It says what changed since, not that the change caused it.

A live check reads only your own site's existing pages. For rivals, sites to get cited on and new pages, watch Prompts or Sources instead.

Marks are saved per brand. Marking your first item done ticks Apply your first fix on the sidebar's Start here checklist.

FAQ

Does the To-do list cost anything?

No. It's built from your stored runs, so reading it and noting progress is free. Reporting an item done starts remake's check that the change is live, at the price card's row. Your AI tells you first.

Why did an item leave the list?

A newer run no longer found that gap, or other items of its kind now come first. Rivals, site fixes and new pages follow your latest run. Get cited items follow your recent runs, up to the last 12.

What happens to a done item if the problem comes back?

It stays Done. A mark stays with its item from run to run. Ask your AI to set it back to New to put it on your open list again.

Why doesn't every failed check get an item?

Only 5 checks add an item, and only when they fail. Every check that needs work is listed under All site checks on My website. Website checks explains each one.

Open Improve