Actions

An Action is one thing this brand can do to appear in more answers. It names a verb and a target: “Get listed on bestof.example”, “Unblock OAI-SearchBot”. The Actions page is the worklist. Each week it fixes a short list of Actions in order, with the evidence behind each one. Actions come from the stored answers. No model gives advice here. A language model labels what kind of website a destination is, and nothing more.

The tiles

A prompt that asks about one competitor by name gets answers that name that competitor whatever this brand does. Those answers are no losing answers. No dashboard filter applies to this page.

How the worklist is built

FullReach AI builds the worklist again after each daily collection. It reads the analyzed answers of the last 28 days. Branded prompts stay out, as on every page.

1. Losing answers. A losing answer names at least one monitored competitor and does not name this brand.
2. Destinations. A destination is a website that losing answers cite, taken as one place to act. It is the site’s domain. A subreddit, a LinkedIn page and a YouTube channel each count as a destination of their own. This brand’s own site is never a destination.
3. The gate. A destination joins the list when losing answers cite it on at least 50% of the project’s platforms. It also joins when they cite it on at least 25% of the days with answers. In that case they must cite it on 2 days or more.

A destination that no longer passes the gate fades from the list. It returns when it passes again.

The kinds of Action

A monitored competitor’s own site gives no Action. A page of this brand’s own answers the question that site is cited for. A competitor’s website can be an address below its domain, such as rival.example.net. Actions group the answers by the registrable name, here example.net, so its cited pages count under example.net. That gives no Action when another brand of this project, or this brand, has the website example.net. The card “The monitored brands’ own sites” on the Sources tab still counts that competitor’s own address. The evidence grade says how far the finding reaches. The methodology says what each grade rests on.

• A Crawler Action appears for each crawler that feeds a tracked platform and that this brand’s robots.txt blocks. That is a search crawler, or a crawler that reads a page when a person asks. Its answers are the losing answers on the platforms it feeds. The Crawlers tab shows the rule.
• A crawler logs Action appears when FullReach AI reads a Cloudflare zone that logs no request to this brand’s site, from any visitor. Until the logs cover the site, nobody can tell whether AI crawlers fetch its pages. It leaves the list when a daily read finds the site in the logs.
• An Action for your page names a prompt, not a site. It appears when the prompt’s losing answers pass the gate, number at least 10, and hold 1% of the losses. At least 50% of them must cite a list that a tool maker wrote about its own tool. No answer may cite this brand’s own site. Prompts whose answers cite mostly the same pages fold into one Action, and it lists their questions.
  • FullReach AI reads this brand’s own sitemap once a day. When a sitemap page already fits the prompt, the Action asks to improve that page instead. A page fits a prompt that names one competitor when its address holds that competitor’s name. It fits any other prompt when its address holds every word of the phrase the platforms searched most.
  • Its evidence is its brief. The brief holds the questions, the phrase the platforms searched most and the competitors the answers name. It also holds the lists the answers cite instead, and what those lists printed. FullReach AI holds no search volume, so the phrase is no target. No model writes any part of the brief.
• A review site or a publication counts as a best-of list when at least half of its cited pages are rankings. That is because the answers cite its rankings, not its profiles or its stories.
• The kind of a destination comes from its host first. Forums, publishing platforms and known directories have fixed kinds. A domain that the project holds for a monitored competitor gives no Action. For any other domain, a language model reads the titles and paths of its cited pages and labels it. A reference work, such as an encyclopedia, gives no Action. It shows as context in other Actions’ evidence.

Openings

An Opening is an objection that the Perception answers raise against the monitored competitors. It needs at least 3 answers in the last 28 days. Only you know whether this brand does better there. So an Opening waits in Openings to confirm, a fold under the list, until you choose:

• The button that says this brand does better confirms it. It waits under Next in line until a Monday gives it a place on the list.
• Not for us dismisses it. The Dismissed view then shows it by its own words, apart from the Actions.

An Opening’s count is perception answers, not losing answers. The largest confirmed Opening takes one place on the weekly list, with no count of its own. Openings need Perception, so they follow its plan gate. A confirmed Opening says which page of this brand’s own answers it. For a pricing theme, that is a public pricing page. For a feature theme, it is a feature matrix. For a monitoring theme, it is a positioning page. Its evidence sets this brand’s Profile beside the competitors’ quotes. Without a Profile, it asks for one. A quote that states a price never shows. Quote a competitor’s price only from its own pricing page, with a link and the date. Never quote one from an AI answer.

This week’s list

Only an open Action that this brand can work on goes on the list. A tool maker’s own list, an unclassified site and a page cited for a fact never do. The Actions on it are in this order:

1. Crawler Actions, the crawler logs Action and the sitemap Action, because they unblock measurement.
2. The largest confirmed Opening. Its answers come from questions that name a competitor, so it shows no count.
3. Every other Action, by the losing answers it targets that no Action above it targets.

The first build of each week fixes a list of 8 Actions. The list stays the same until the next Monday. An Action that you have not finished keeps its place on the next list. New Actions take only the free places. The others wait under Next in line, in the same order. An Action that the answers no longer support leaves the list. The sidebar badge counts the open Actions on the list.

The list and the detail

The worklist card holds the list, and beside it the detail of one Action. A press on a row opens that Action in the detail. A link from another page opens its Action the same way. A link to one kind of place, such as a group of Your plan on the Visibility tab, shows All sites with that kind alone. The chip beside Show removes that filter. Each row of the list shows:

• Its number in the order of the list.
• An icon: the site’s own icon, or an icon for the work when the target is no site.
• A headline that names the page: for a best-of list or a forum thread, the page the answers cite most, by its title. “+2 pages” says that the answers cite more pages of the same site.
• A line under the headline: the competitors that the losing answers name. From the third week, it also says how long the Action has been on the list. For a Store, a Product kind and an Opening, the count is no count of losing answers. So the line starts with what the count counts.
• Its Action kind, which says what kind of site or work it is. Hover over it for one line on that kind. For a site, it labels the whole website. The page kind on the Sources tab labels one cited page.
• Platforms: on how many of the project’s platforms the losing answers cite it.
• Losing answers: the losing answers that cite it. An Opening counts perception answers instead. A Store counts the answers that sell a competitor. A Product kind counts the shopping answers with no card of this brand. Under Next in line, the column reads “Targets” and counts the losing answers that the Action adds.
• Its state: an open ring, a check mark for done, or a cross for dismissed.

The detail states the Action’s place on the list, its counts and its evidence. One answer can sit behind several Actions. So the detail also says how many of the Action’s losing answers the Actions higher on the list can win too. A Store and a Product kind count shopping answers, and not all of those are losing answers. So their detail says how many losing answers the Action can win that no Action higher on the list can win. Answers can also name brands that nobody monitors. Competitors lists those brands under Detected competitors. Mark done sits at the top of the detail. The ⋯ beside it holds Copy as text, Read the answers and Dismiss. The tiles above the list filter it by the people who do the work. All teams comes first and shows every Action. Under By team, the Outreach, Content and Developer tiles split it. Each tile counts its open Actions and how many of them are on this week’s list. Press All teams to remove the filter. A filtered list keeps each Action’s number in the order.

• Developer: Crawler Actions, the crawler logs and the sitemap.
• Content:
  • a page of this brand’s own
  • an Opening
  • a tool maker’s own list
  • a kind of product
  • an article or a video on a publishing platform
• Outreach: every other site, and the author or the channel on a publishing platform.

Show switches the list between This week, Done, Dismissed and All sites. Under This week, five folds hold the rest:

• Openings to confirm
• Next in line
• Lists a tool maker wrote about itself
• Sites that sell a place
• Cited for a fact

A site whose page offers a paid listing never goes on the list, because a paid place is an advert. Done lists every done Action, a faded one too. All sites lists every Action in any state, and a search box finds a site. A site of no known kind names no job, so no view, tile or CSV export shows it, and neither does Your plan. The three dots on the card export the list as CSV and copy this week’s list as text.

States

The build never changes the state you chose. A restored Opening goes back to Openings to confirm. A faded Opening waits for your choice in the All sites tab.

The evidence

The detail of an Action holds its evidence. It starts with How to do it: three steps for its kind. The steps name its page, the way in that a page read found, and the competitors the page lists. The steps are fixed text, and no model writes them. They end at the work itself, because a page may never list this brand. The rest is the evidence:

• Pages the answers cited, the destination’s pages behind the losing answers.
• Appears in answers to, the prompts.
• Named in the losing answers that cite it, the competitors.
• Fanouts before answering, the fanouts that the platforms ran first. Only some platforms report them. See Fanouts.
• Also cited, as context, reference pages the same answers cite.
• Read the answers, under the ⋯, opens the recent answers behind the Action.

A platform sometimes gives a cited page as its site’s homepage address, under the page’s own title. Such a page has a row for each title, and its link opens the homepage. The evidence states what the answers did. “The answers citing this page named Competitor A” is a claim about the answers. It is never a claim that the page mentions Competitor A. Only a dated page read, below, says what a page printed.

Page reads

FullReach AI reads the top 3 pages of a destination to see which names they list.

• FullReach AI reads the pages of an open destination once a week.
• After you mark a destination done, FullReach AI reads its pages every day for 14 days. After that, it reads them once a week, until a read finds this brand.
• A project spends at most 100 reads a week. The pages of a destination marked done in the last days come first, then open destinations, then older done ones.
• A read obeys the site’s robots.txt. It stores the names the page lists, the page’s own dates, its author, and the link to its way in. The way in counts only on the page’s own site, and never as a generic contact, sales or register page. When the page offers a paid listing, the read keeps that sentence. It never stores the rest of the page’s text.
• FullReach AI reads the pages of this week’s list first.
• A YouTube page is the exception. FullReach AI asks YouTube for the name of the video’s channel, and it does not open the page. It asks for the YouTube pages that losing answers cite, at most 50 a day. The pages with no read come first, then the most cited. These reads are outside the weekly limit.

For most kinds, the evidence shows what the page printed on the day of the read. The kinds are a register, a list, a review site, press, a forum and an unclassified site. It says how many names the page listed, and whether this brand was among them, at which place. When robots.txt refuses the read, the page says so.

The Profile

The Profile is what this brand states about itself: what it offers, who it is for, its prices with a link, and its proof. Proof is awards, rankings, numbers and clients. Edit the Profile under Settings, Project.

• For a project added with the setup wizard, the form shows a draft from the wizard’s read of the website until you save it. An Opening’s evidence shows only a saved Profile.
• Perception reads the offer, saved or drafted, to decide which company an answer describes.

Live, and the outcome

A done Action is finished. It goes Live only when the product can see the change, and nothing waits for that:

• A destination goes live on the first page read after “done” that finds this brand on the page.
• A Crawler Action goes live when the crawler audit no longer finds the block.
• A crawler logs Action goes live when the Cloudflare logs record a request to the site.
• A sitemap Action goes live when an AI crawler fetches a page it names.
• An Action for your page goes live when answers to its questions cite the page on 2 platforms or 2 days. Mark done on a new page asks for its address, which must be on this brand’s site. An Action that improves a page records that page. Its evidence then shows each stage with the first day it held: published, fetched by an AI crawler, cited, and this brand named. Without crawler logs for the page’s host, the fetch reads “not measured”.

An Opening never goes live. Reopen clears the live day, so the next “done” is a new claim. The trend chart of the Visibility tab marks the day an Action was marked done and the day it went live.

How the outcome is measured

For a destination that went live, the outcome compares two groups of prompts. It reads the 28 days before the day it went live and the 28 days after. That day itself counts in neither.

• The prompts citing it are the prompts whose answers cited the destination in the days before.
• The other prompts are every other prompt of the project.

A wording counts only when it has answers both before and after that day. So a reworded prompt counts with the wording that spans the day, or not at all. For each group and each period, the share is the answers that name this brand, out of the answers. The outcome is the change of the first group minus the change of the second. So a rise that every prompt shares, such as a new model version on a platform, cancels out. For an Action for your page, the first group is its questions. The second group is the other questions that qualified for such an Action and that nobody acted on. Its questions need answers on at least 3 platforms. The outcome shows with a 95% interval. It needs 21 days with answers since the Action went live. Each of the four counts also needs at least 30 answers. For every kind of Action, the interval comes from each share’s Wilson interval. So a share of 0 before still gives an interval of real width.

An example

The outcome is 15 − 2 = +13 points. Its 95% interval runs from about +5.5 to +20.5 points. The whole interval is above zero. So the page says that the share rose 13 points more on the prompts citing the page than on the other prompts.

Where it appears