Pages to write, before a competitor does
SEOLift now hands you the pages your site is missing as briefs: the search, an outline in the words people type, the links waiting for the page, and a check after you publish. Here is each piece, and what it refuses to guess.
Start with what you already show for
Searches you already show for with no page behind them.
Search Console already knows searches where Google shows your domain, yet no page on your site is about them. SEOLift collects those under one goal, Pages worth writing. Each row has real impressions behind it, so you write for demand you can see, not for a hunch.
These rows carry no click forecast. A page that does not exist has no history to count from, and the module will not invent one. They never take a slot in the five actions of the week either. They are a backlog you open when the repairs are done.
A brief, not a one-line idea
A page to write, outlined in the words people search.
Each of those searches becomes a brief. It names the search to answer, the searches that go with it, a proposed address, the page the new one must not duplicate, and one sentence on why your current content does not cover it.
The outline is built from the words people typed. The questions they already ask become headings. Every fact is either on your page already or marked as needing a source. A comparison search, such as X versus Y, gets a table for price, features and complaints, with every cell empty and unchecked until a person at your company fills it.
One page per intent, checked before you write
Before you write a new page, we check Google doesn’t see it as your old one.
Before a new-page brief reaches you, SilentStork compares the ten results for its search with the ten for the search your nearest existing page already serves. If 4 or more of the 10 addresses match, the two searches are one intent. The brief is refused, and the action becomes: add a section to the page you have.
Roof repair cost and how much to fix a roof are one page, and you learn that before you spend a day on the second. The check is an intersection of two address lists, arithmetic rather than a model’s opinion. Without paid result-page data, the brief goes on with a visible unchecked mark instead of pretending.
Clusters, with the links planned ahead
Hub first, the rest behind it, with links ready before the page exists.
Briefs that share a search stem form a cluster. The hub is the one with the most impressions behind it, and it is the one page to draft now. The supporting briefs wait, shown as queued, until the hub has been read and indexed. A cluster whose hub has no impressions behind it is never created. That is the guard against pages nobody asked for.
An approved brief also gets a hold list: three to five of your existing pages that could honestly link to the new one, each with the paragraph and a phrase already in it. Once the new page has been read, the Internal Link Dealer judges those links under its usual rules and writes its verdict back. No pasted see-also sentence, no anchor that was not in your text.
What competitors publish, read from their sitemaps
What competitors publish that you don’t. No keyword tool needed.
Paste up to five competitor sitemap addresses. Only the address lists are fetched, never the pages. Each address gets a label read off its slug, marked as inferred. A label is not a search volume and not a position.
You get three piles: page types they have and you do not, pages you have and they do not, and pages you both have where yours is losing impressions. Only the third stands on your own Search Console, so it is the one that points at a page to work on. The first is a backlog, outside the weekly reading.
Searches competitors win, with an address to pass
10 searches a smaller site can win, each with the address to pass.
Sitemaps show what others publish. This check shows what they win. Press one button on the Competitors screen. It finds up to five competitors that share searches with you and lists the first-page searches they rank for and you never show for. By default it keeps those with a volume above 10 and a difficulty below 35.
Each remaining search is read on the live result page. It stays only if that page holds at least two independent sites and one of them has few enough referring domains to be passed. That address is shown next to the search. At most 10 stay, and the list is never padded. A run uses at most 47 checks from your allowance, and every search it keeps gets a brief overnight.
After you publish, the check that it landed
Published? The next reading checks the headings landed.
When the page is live, give the brief its address. The next reading compares the approved headings with the headings on the page. When most of them are there, the brief is marked as landed, visibly, and its cluster moves on to the next page.
Any number in the new title or headings that was not in the brief is reported. It may be right. It was simply not in the brief, and you should know that before a reader does.
What it does not do
It gives no click forecast for a page that does not exist, because there is nothing to count from. It does not write the article and publishes nothing: you get a brief, an outline and a table to fill in. It does not read competitor pages, only their sitemaps and the result pages for a search.
And zero is a valid answer. If no search survives the competitor gap, the screen says at which filter the list ran dry, rather than filling it with guesses. All of this sits inside SEOLift, on the Business and Agency plans.