Target sitemap: which pages your website needs.
The target sitemap sets your existing website next to the target state, row by row. It shows which pages stay, are optimised, merged or newly created – and every suggestion cites at least one piece of evidence.
A page plan backed by evidence instead of gut feeling.
Websites often grow out of habit: a blog here, a service page there. Which page is missing for an economically relevant topic, which pages compete with each other and where new content belongs remains unclear.
The target sitemap connects the search market, your website and AI visibility in one plan. A suggestion without evidence is discarded.
Calculate first, then ask.
A run works in a fixed order. Steps with a fixed basis – such as a snapshot or SERP run – are only recalculated when that basis has changed.
Glossary, guides, services, fact directory: every detected section must exist on the website, otherwise it is discarded.
Your own pages are read with the same analysis as competitor pages – the basis for a fair comparison.
View content analysis →Checked before the plan is calculated; the requirements can feed into the page plan and brief.
View Grounding Page →Mandatory evidence for every new page.
Every suggested page cites at least one confirmed keyword or measured AI question as evidence. The strength of the evidence is visible on each suggestion – a new page does not appear just because a model thinks it makes sense.
A confirmed topic without a page.
A requirement that none of your pages meets.
A prompt with no page to answer it.
A hub that is too thin.
Rules for a usable page tree.
The planning rules live in the code, not in a prompt.
A surplus page is downgraded, not discarded.
With low ranking chances or weak business relevance, BrandZaps suggests fewer subpages.
Anything that is only a variant of a planned topic becomes a section, not a page of its own.
A guide on “SEO consulting” belongs in the guides section, not under the service page. The topical relationship is recorded in the brief and created through linking.
The unit of work is the hub.
A topic cluster can be decided while collapsed. Rows that require approval, such as merging, splitting or excluding, remain individual decisions and are explicitly not covered by a bulk confirmation.
Existing pages that need no action stay in the tree but are not tasks. A rejected row can be confirmed later or set back to open.
Every run creates a new version; the previous one is kept as history. Decisions survive the run as long as their evidence is unchanged.
If the evidence changes, the row goes back into review.There is no score and no ranking by importance. A run starts when the user triggers it, not automatically after every update.
A GEO profile and a content brief – in one step, stored separately by discipline.
Not every good idea needs its own URL. Every planned URL needs a reason.
An evidence-based page plan with roles, page types and linking relationships – row by row, next to the current state.




