↑↓ selectEnter openEsc close

    1. Elbuz

    How to manage website articles and pages

    10 min readUpdated

    The “Articles” window stores your website pages — “Delivery and payment”, “Contacts”, blog, news — plus the header, footer and home page text. The AI chat consultant uses them as its knowledge base.

    When you need this:

    • A “Delivery and payment” page in the footer — link it from the footer article.
    • Moving from an old site with a blog — Elbuz downloads pages from a list of links.
    • The SEO generator must keep your meta title — “empty only” mode leaves it alone.
    • Blog articles compete in search — “Cannibalization” finds pairs and removes the duplicate in one click.

    Where to start: Directories → Articles → “Add”.

    What the “Articles” window does

    How to create a “Delivery and payment” page

    1. In the tree, click where the page should go. For top-level pages, that is the tree root; for a blog article, the “Blog” section.
    2. Click “Add”. If anything other than the root is selected in the tree, the “Add” dialog shows “Parent element” — where the item will go; if the field is missing, the item goes to the root. Enter “Name”: “Delivery and payment” — and confirm.
    3. The new page opens on the right. On the language tab, enter the text and, if needed, a short description and the fields of the “SEO text (meta, h1)” block.
    4. On the “Extras” tab, turn on “Active” and check that “Export to website” is on. Click “Save”.

    A new page is created disabled. The “Add” button does not set “Active”, and until you turn it on, the page is not on the website. The website shows an article only when both “Active” and “Export to website” are on and the article is not in the “Trash”.

    Check what is selected in the tree before you click “Add”. The new item goes inside the selected one. If an article is selected, it becomes a section and the new page its child. The new page also inherits the parent's “Export to website”, “Top of page (header)”, “Bottom of page (footer)” and “On the home page” flags.

    If the same place already has an item with that name, “Add” creates nothing new, but another node with that name appears in the tree — it is the same item. After you refresh the tree (☰ → “Refresh”), the extra node disappears. Give a new page a name this section does not have yet.

    How to add a page to the website footer

    The website footer is not a list of flagged pages but the text of one article with the “Bottom of page (footer)” flag, shown at the bottom of every page. “Top of page (header)” (header of all pages) and “On the home page” (home page text) work the same way. The flags are in the card's “Article display” block.

    So, to get “Delivery and payment” into the footer:

    1. Create the “Delivery and payment” page as a regular article (above) and do not set “Bottom of page (footer)” on it.
    2. Find the article in the tree that has “Bottom of page (footer)” on — that is your footer. If there is none yet, create a “Footer” article and turn on this flag, “Active” and “Export to website”. Do not disable this article: a disabled footer article hides the footer on the website.
    3. In the footer article text, add a “Delivery and payment” link to the new page address and save.

    Only one article should have the “Bottom of page (footer)” flag. The website takes one article with this flag (not a section) for the footer and does not look at which one was flagged last: if there are two, the footer is usually the one created earlier.

    So flagging “Delivery and payment” will not replace the old footer, only confuse things. To switch the footer, first remove the flag from the old article. If the article chosen as the footer is disabled or excluded from export, the website has no footer at all. The same goes for “Top of page (header)” and “On the home page”.

    These flags do not hide the article from the website: the footer article, like any enabled article, shows in its section list and can be found by smart search. Keep it in the root or in a service section.

    In an Elbuz CMS store, these flags do nothing. The “Site page, not blog” flag makes an article a store page (store pages); the store design draws the header and footer with links to them itself. “On the home page” does nothing either: without “Site page, not blog”, an article becomes a blog post unless its type is “Landing page” or “Menu” (table below).

    How to move articles from an old site by URL

    If your old site already has a few dozen articles, you do not need to copy them by hand: Elbuz downloads the pages from their links for you.

    1. In the tree, select the section for the articles. If nothing is selected, the articles go to the root.
    2. Open the “Additional operations” menu (three lines above the tree) → “Import articles from URL”.
    3. In the “Parent section” field, check where the import goes (“Root” is the top level).
    4. In “URLs (one per line)”, paste the page addresses, each on a new line, and confirm.
    5. When it finishes, a table appears under the field: a green mark means the page was processed (an article was created or one with the same name was found, see below), red means an error, and “(exists)” means the address was imported before and was skipped. The tree refreshes by itself.
    Import articles from URL dialog
    1Parent section
    2URL list
    1. Parent section — shows Root because nothing is selected in the tree.
    2. URL list — one address per line; left empty.

    If the section already has an article with the same name, the import does not create one but “links” it. If the section has an article with the same name as the downloaded page, no new article is created, even though the result row is green.

    Instead, the existing article gets the source address and becomes “imported” — and the very first “Sync imported articles” overwrites its text and meta tags with the old page content. Before importing, make sure the section has no manually written articles with the same names, or import into a separate empty section.

    What the program takes from the page:

    • The name — from the heading, otherwise og:title, the title tag or JSON-LD. A page without a heading is not imported. Note: the program cuts everything after the last “|”, “-”, “–” or “—” (to strip the site name from the title), even in an H1. “Wi-Fi router for home” becomes “Wi”. Check names with hyphens after import and fix them in the card.
    • The text — from the main block of the page (article, main). Menus and sidebars of the old site are cut off, and classes and styles are removed from the tags: paragraphs, headings, lists, tables, links and images remain.
    • Meta title and meta description — from the matching tags of the old page; the short description — from the meta description or the first 300 characters of the text.

    An imported article is enabled right away, and the source address is saved in the “Source link” field on the “Extras” tab. The import fills one language; translate the others with the translation buttons next to the fields.

    Images stay on the old site. The program moves links to the images, not the files themselves. If you shut down the old site, the images in the articles disappear — replace them with your own before you close the site.

    Split a long list into parts. The program waits up to 30 seconds for each page, and the window waits up to two minutes for the result. If the old site is slow, even a few addresses can use up those two minutes: the import continues on the server, but you will not see the results table.

    Paste the list in parts and check the result in the tree. Running the same list again is safe — addresses that were already imported are skipped.

    Syncing with the source

    If the old page was updated, you can pull the imported article again: for all of them at once — “Additional operations” → “Sync imported articles”; for one — the refresh button next to the “Source link” field in its card. Only enabled articles that are not in the “Trash” are synced.

    Syncing overwrites your edits. The name, H1, text, short description, meta title and meta description are taken from the old page again. If you have already reworked an imported article, do not sync it.

    How to keep the generator from overwriting your meta title

    An article's meta title, meta description and H1 are written in its card, in the “SEO text (meta, h1)” block on the language tab. The SEO text generator fills these fields in bulk — with the “Fixed text” template or with AI. It leaves a manually written title alone only as long as you do not allow overwriting:

    • in the “Fixed text” template, keep “Fill empty fields only” on when you click “Apply”;
    • in the AI generation run window, keep the “Only update empty fields” checkbox on.

    With these checkboxes, the generator fills only empty fields, and a meta title you filled in by hand stays as is. Without them, the article template overwrites the fields in all articles and sections at once — there is no separate protection for a single article.

    The article card has no “Don't change SEO text via generator” flag. A label with this name exists in the program's language files, but the card does not show it, and the generator ignores it. Only the “empty only” option in the generator itself protects your fields.

    The second source of overwriting is syncing an imported article: it takes the meta title from the old page, whatever the generator settings are.

    How to add internal links and find duplicate articles

    The “Bulk interlinking” window adds internal links between similar articles and finds articles that duplicate each other's topic. You open it from the “Articles” window: click the tree root, “List of sections and articles” appears on the right, and its “Additional operations” menu (three lines) has “Create index for AI” and “Bulk interlinking”.

    Add internal links between similar articles in bulk

    1. Run “Create index for AI” for the articles. Interlinking sees only articles with this index (embeddings): it skips an article without the index and does not link to it either.
    2. Open “Bulk interlinking”, the “Settings” tab. “Only for articles without existing data” is on by default — articles that already have auto-links are skipped; clear the checkbox to recalculate all of them. Choose “Language” and, if needed, enter “Article ID (empty = all)”.
    3. Click “Run”. Progress is shown on the “Log” tab, which also has the “Stop” button.
    Bulk interlinking window, Settings tab
    1Run
    2Only for articles without existing data
    3Cannibalization tab
    1. Run — starts the batch; not pressed during the shoot.
    2. Only for articles without existing data — limits the run to articles with no auto-links yet.
    3. Cannibalization — tab with the search-query overlap report.

    For each enabled article (not a section), the program takes the 15 closest articles in the same language by content, and AI looks for phrases in its text that relate to them word for word. The phrases it finds become auto-links — the article's previous auto-links in that language are replaced.

    Links point only to articles with an . On the website, they are inserted into the article text when the page is shown; the text in the card does not change. You can see the result for one article in its card, “Interlinking” tab → “Smart Links”; the “Generate Smart Links” button there recalculates links for that article only.

    The “No embeddings for link generation” message after “Generate Smart Links” means the article has no AI index in that language. Run “Create index for AI” from the “Additional operations” menu above the article list and generate again.

    Find articles that duplicate each other

    The “Cannibalization” tab of the “Bulk interlinking” window looks for pairs of articles with similar content — they compete in search for the same queries.

    Above the table are two unlabeled lists: in the first, choose the article language (the main one by default); in the second, the similarity threshold: “Medium (≥75%)”, “High (≥85%)” or “Critical (≥90%)”. Then click “Find pairs”. Enabled articles with an AI index are compared, and the table shows up to 200 of the most similar pairs.

    In each pair, the program picks the “Main article” — the one with more clicks in Google or, without Google data, the one with more incoming internal links. For the secondary article, the “Action” column has three buttons:

    • “Canonical” — the secondary article stays on the website but points search engines to the main one as the primary page;
    • “301 Redirect” — after you confirm, the secondary article is disabled and moved to the trash, and its address redirects to the main one;
    • “Noindex” — the secondary article stays on the website but is closed to indexing.

    To pick the main article by real traffic, set up the “Google Search Console” block on this tab: “Website URL in GSC”, your Google service account JSON in “Service Account JSON”, “Save GSC settings”, then “Synchronize” — 28 days of article clicks and positions load. Add the service account as a user of your Search Console property.

    How articles get into the Elbuz CMS store blog

    An Elbuz CMS store has a built-in blog: an article list with categories and tags, and article pages with a cover, table of contents, author and “Read also” block. You write the articles here, in the “Articles” window; they reach the website with the catalog on export with the “Elbuz CMS” template. The store admin panel has no blog articles.

    The blog is on in the “Elbuz CMS” template from the start. Two settings control it: the “Articles” checkbox in “Select data to export” and the “Update articles for the Elbuz CMS Blog module (elbuz_blog)” rule in the additional settings. A new template has the checkbox on; the rule turns on by itself if the site has the blog module.

    In templates created earlier, a program update turned both settings on once: if you turn them off later, they stay off. You don’t see these sections of the Elbuz CMS template: the blog is turned on and off in “Store settings”, the “Blog” card (how to check).

    Where an item goes depends on its flags:

    Item in the “Articles” windowWhat you get in the store
    SectionA blog category; section nesting and order are kept
    An article of any type except the two rows belowA blog post in its section's category
    An article with the “Site page, not blog” flag or with “Article Type” set to “Landing page”A separate store page, like “Delivery and payment”; it does not go to the blog. The “Top of page (header)” and “Bottom of page (footer)” flags do not affect this
    The “Store pages” sectionDoes not become a blog category: it is a service folder for store pages
    “Article Type” set to “Menu”Not exported to the website

    Along with a blog post, the website receives:

    • the name, H1, short description (the teaser in the list), text, tags, meta title and meta description — in all languages the template exports;
    • the cover image — “Link to main image”, or, if it is empty, the first one from “Image link”;
    • the author — name, photo and description from the author card;
    • “Publish date” and the date of the last change;
    • “Do not index (noindex)” — the article is closed to search engines and is left out of the sitemap;
    • “Generate table of contents” — a table of contents appears above the text if the article has at least three subheadings.
    An article in an Elbuz CMS store blog: “Overview” type, title, category, date, reading time, author, summary, cover and the collapsed “Contents”
    1Article Type
    2Category, publication date, reading time, author
    3Cover
    4Contents
    1. The badge is the “Article Type” from the “Extras” tab.
    2. Category is the article’s section in Elbuz; date is the publication date; author comes from the author card.
    3. “Link to main image”; it goes to the site as a link.
    4. Appears with “Generate table of contents” when there are at least three subheadings.

    The blog shows “Active” articles whose “Publish date” has come and that have a name in the page language. An article with a future date can be exported in advance: it appears on that date by itself. Newer articles come first — the Elbuz tree order sets only the category order in the blog.

    An article has one address for all languages, with the language added as a prefix. If a product, category, brand or other page on the website already has that address, a hyphen and a number are added to the article address.

    Images in the article text are not copied to the website. The cover image, author photo and images in the text stay as links to Elbuz storage, so you do not need to export them separately.

    An article deleted in Elbuz cannot be brought back. The blog removes articles and sections missing from the export with the next export, unless article deletion is turned off in the template (deletion on the website). To hide an article for a while, turn off “Active”: turn it back on and the article returns at the same address.

    How to manage store pages and article products in Elbuz CMS

    “About us”, delivery, policy and terms pages

    Elbuz CMS store pages — “About us”, “Delivery and payment”, “Privacy policy”, “Terms of use” — are also managed in the “Articles” window. What makes an article a separate store page rather than a blog post is the “Site page, not blog” flag in the “Article display” → “Store page” block of the article card.

    The article display section of the About us store page card
    1Site page, not blog
    2Page role
    3Top of page (header) and bottom of page (footer) flags
    1. The flag makes the article a store page of Elbuz CMS instead of a blog post.
    2. The role binds the article to the existing store page with that role, so a second identical page does not appear.
    3. For an Elbuz CMS store these flags decide nothing.

    The program sets up nine standard pages by itself — when you create the “Elbuz CMS” template or before the first article export. What each one contains and what to rewrite for your store — Elbuz CMS store pages.

    • With their own role: “About us”, “Delivery and payment”, “Privacy policy”, “Terms of use”.
    • Five more: “Returns and right of withdrawal”, “Warranty”, “Cookie policy”, “Withdrawal form”, “Legal notice”.

    They already have text in eight languages — Ukrainian, English, Russian, Polish, German, French, Spanish and Italian, whichever of them your catalog has — the “Site page, not blog” flag, and they sit in the “Store pages” section. The program sets up each page once: it will not bring back a standard page you deleted.

    If you already have such a page, no second one is created. The program takes the article with the same role, or one in the tree root or “Store pages” with exactly the same name in any language, such as your own “About us”. The article gets the flag and role, empty fields get the standard text, and your text stays.

    • “Page role” links the article to a page that already exists in the store: “Privacy policy”, “Terms of use”, “About us”, “Delivery and payment”. Then the export updates that page instead of creating a second identical one next to it. For other system pages, such as “Warranty” or “Contacts”, leave the role at “—”: they become new store pages.
    • A page without text is not exported. An article with this flag and with empty text and short description in all languages changes nothing: no page is created, an existing page stays as is, and so does its earlier blog post. Add text — the page appears with the next export, and the blog post moves to it with a 301 redirect.
    • Remove the flag and the item moves to the blog. The store page becomes a blog post, and the old address gets a 301 redirect to the new one. It also works the other way: a blog post that gets the flag moves to the store pages with a redirect.
    • The export touches only pages linked to your articles — created from them or linked by role, and it keeps managing them: remove the flag and the page moves to the blog; delete the article and the page leaves the store with the next export. Store pages not linked to articles stay as they are.
    • Some pages are never deleted by the export: “About us” and “Delivery and payment”, linked from the store header and footer, and pages used in store settings, such as the terms accepted at checkout or registration. If such a page's article is deleted or moved to the blog, the page stays as is; the export stops updating it and notes this in its log.

    A new article created with the “Add” button inside the “Store pages” section gets the “Site page, not blog” flag right away. Moving an existing article into this section does not change the flag — set or remove it in the card.

    Products under a blog article

    Under an Elbuz CMS blog post, you can show the products the article is about. You set them in the “Article products” section of the article card:

    • “Add product” opens the product selection window; the selected products go to the end of the list, and products already added are not duplicated.
    • The list shows the photo, name and SKU. You change the order by dragging the handle on the left or from the keyboard: Alt+↑ / Alt+↓ on the selected row. The products appear on the website in the same order.
    • The trash icon on the right removes a product from the article after you confirm. The product itself stays in the catalog.
    The «Article products» section of an article card with a list of four linked products

    Any change to the list marks the article as changed. With the “Elbuz CMS” template, articles go to the website with every export anyway. The blog shows only products already in the store (exported by the same template). The list is available to anyone whose group permissions open the “Articles” window; otherwise the program replies “You don't have permission for this operation!”.

    Scenarios

    Moving from an old site with a 40-article blog

    What to do. Create a “Blog” section, select it, open “Import articles from URL” and paste the addresses in several parts. What you get. Enabled articles in “Blog” with the names, text and meta tags of the old pages. What you don't get. Images are not copied to your server, and no other languages appear — the import writes one.

    “Delivery”, “Payment” and “Contacts” pages in the footer

    What to do. Create three regular articles in the root, turn on “Active”, and link to them from the article with “Bottom of page (footer)”. What you get. The links are at the bottom of every website page. What you don't get. The footer does not list pages by itself: a new page shows there only once you link it.

    Blog articles compete with each other in search

    What to do. Run “Create index for AI”, open “Bulk interlinking” → “Cannibalization”, choose the “High (≥85%)” threshold and click “Find pairs”. What you get. For the secondary article of each pair — “Canonical”, “301 Redirect” or “Noindex” with one button. What you don't get. Articles are not merged automatically: you choose the action for each pair.

    What to do when a page is missing from the website

    What you seeWhyWhat to do
    I created a page, but it is not on the website A new page is created with “Active” off, or “Export to website” is off, or the page is in the “Trash” On the “Extras” tab, turn on “Active” and “Export to website” (creating a page)
    The whole website footer turned into the text of one page This page has “Bottom of page (footer)” on, and the website chose it as the footer Remove the flag, keep it only on the footer article, and add a link to the page there (footer)
    The footer disappeared from the website The article the website uses as the footer was disabled or excluded from export; or several articles have the footer flag and the website chose a disabled one Keep the “Bottom of page (footer)” flag on one article and turn on “Active” and “Export to website” for it (footer)
    An article suddenly became a section It was selected in the tree during “Add” or an import, and the new item went inside it Drag the child item to the right place and clear “This is a section” on the “Extras” tab
    “(exists)” in the import results A page with this address was imported before Nothing: the article already exists. To update it, use “Sync” (syncing)
    A red mark in an import row The page did not respond within 30 seconds, returned an error or has no heading Check that the address opens in a browser and import it again
    Images disappeared from the moved articles The images pointed to the old site, and it was shut down Upload the images again and replace them in the text (import)
    My meta title was replaced with a template one The generator was run without “empty only”, or the article was synced with its source Enter the title again; from now on, run the generator only with “empty only” (meta title protection)
    An article is missing from the Elbuz CMS store blog “Active” or “Export to website” is off, the “Publish date” has not come, or no name in the page language; “Site page, not blog” or the “Landing page” type made it a store page (an update once flagged all articles with “Top of page (header)” or “Bottom of page (footer)”); “Menu” type is not exported; the blog is off in the template Check the date and flags in the article card (Elbuz CMS blog); if the blog is off, turn it on in the “Blog” card (how to check)
    The “About us” or “Warranty” page is missing from the Elbuz CMS store The article's text and short description are empty in all languages — such a page is not exported; or the “Site page, not blog” flag is not set, and the article went to the blog Add text, set the flag and save; the page appears with the next export (store pages)

    FAQ

    Will syncing overwrite my edits in an imported article?

    Yes. Syncing takes the name, H1, text, short description, meta title and meta description from the old page again. If you have already reworked the article, do not sync it.

    Can I import an article in several languages at once?

    No. Import from URL fills one language. Translate the others with the translation buttons next to the fields or with bulk translation.

    What do the “Top of page (header)” and “On the home page” flags mean?

    Like the footer, each is the text of one article that the website puts into the header of all pages or on the home page. Keep each flag on only one article; to replace the header or home page, remove the flag from the old one (how the footer works).

    The program says “No embeddings for link generation”. What should I do?

    The article has no AI index in that language. Click the tree root, run “Create index for AI” from “Additional operations”, wait for it to finish and click “Generate Smart Links” again. Links point only to articles with an SEO URL.

    Does the system check article text for uniqueness (plagiarism) before publishing?

    No. There is no uniqueness check for either manually written or AI-generated articles: an article goes live as soon as “Active” is on. Check the text with an external service before you turn it on.

    Can I run the Elbuz CMS store blog from Elbuz?

    Yes. Sections of the “Articles” window become blog categories, and articles become posts with a cover image, author, date and tags. The blog is on in the “Elbuz CMS” template from the start, and articles reach the website with every export. More details: Elbuz CMS blog.

    Where to find it and who has access

    Menu
    Directories → Articles
    Other ways in
    tree root → “Additional operations” → “Bulk interlinking”
    Access right
    the “Articles” window in group permissions

    On the left is the tree of sections and articles, with the “Add” and delete buttons and the “Additional operations” menu (three lines) above it and the “Search” field below them. On the right is the selected item's card: language tabs with the title, text and “SEO text (meta, h1)” block, and the “Extras” tab with flags.

    Articles window with the Additional operations menu open
    1Add
    2Import articles from URL
    3Article tree
    4Article card
    1. Add — creates a new article; do not press it on a live site.
    2. Import articles from URL — opens the import dialog.
    3. The tree of sections and articles.
    4. The card of the selected entry opens on the right.

    The tree works like folders. A section holds articles and other sections; an article is an end page. Drag items to change order and nesting. An item becomes a section with the “This is a section” flag, or automatically once “Add” or an import puts something into it. Dragging an article onto another does not make it a section.

    Related topics

    Was this article helpful?

    Missing a feature? Tell us.

    We will add it for free if other users need it too.

    If only you need it, we will agree on it separately.

    Article contents

    Suggest an improvement

    Try Elbuz free