← Home
TREN

Documentation

A step-by-step guide to how Zep AutoSEO works — setup, and every feature explained.

1. Overview

Zep AutoSEO produces deeply researched SEO articles for your business every month and publishes them directly on your site. The goal is visibility not just on Google, but in AI search experiences like ChatGPT, Perplexity and Google AI Overviews.

The system is end-to-end automated: it analyzes your business, builds a content calendar, writes the articles, generates cover images, wires internal links, and publishes on schedule. You just approve and watch the results.

2. Sign-up and Initial Setup

During sign-up you enter your website address. The system crawls your site and sitemap to automatically detect your services, industry and language — this forms the basis of your content strategy.

You then pick a plan and create your account with email verification. Your first content calendar is ready in seconds.

  • Site scan: services + industry + language extracted automatically (real sitemap data, no fabrication).
  • Secure account creation via email verification code.
  • First content calendar auto-seeded.

3. Brand DNA

Brand DNA ensures all content is produced in your voice and with correct facts. It has three tabs: your products/services, your brand info, and your writing style/colors.

AI scans your brand to pre-fill these fields; you edit and confirm. Your brand color also drives the main node color in the Link Graph.

4. Content Calendar and Automated Production

The content calendar shows which article publishes on which date. At most 2 items publish per day (10:00 and 16:00 slots) and never two at the same time.

If you change a date manually, that arrangement is locked — automated processes never overwrite your manual schedule. Use 'Publish Now' to publish early (within the daily limit).

5. Content Engine

The engine is a multi-stage AI pipeline: keyword plan → writing → audit → image → internal linking → publish. Every article is produced so it won't cannibalize content already written on your site.

Titles follow proper capitalization rules, meta descriptions are derived from the content, and sentences are never clipped mid-word. You can reprocess an article with 'Fix Article'.

6. Cover Images

A relevant, photorealistic cover image is generated for each article. The image relates to the article's actual topic — no abstract/cliché (neon, sci-fi) visuals.

You can toggle image generation in account settings; it's on by default for new accounts.

8. WordPress Connection

For WordPress sites you install the ZepSeo plugin. You enter an API key generated in the panel; content publishes straight to your site (as Gutenberg blocks).

Use 'Test Connection & Version' to verify the connection and plugin version. If the plugin is outdated, the panel shows an update notice.

9. Non-WordPress Sites (Custom API)

For sites not on WordPress (Next.js, Laravel, custom PHP/Node, headless, etc.) there are two methods:

  • Webhook (PUSH): on publish, a signed article JSON is sent to your endpoint (HMAC signature verification).
  • ⚠️ Deploy rule (mandatory): files the webhook writes to your site — article pages (e.g. blog/), index/data files, sitemap.xml, llms.txt — must be EXCLUDED from your upload/deploy processes. A full-site FTP/rsync push overwrites them with stale local copies, wiping delivered updates (author box, schema fixes). The webhook is the sole owner of these files; styling changes belong in your template, not in the article HTML.
  • JSON Feed (PULL): your site pulls published articles from us using an API key (?since= incremental sync).
  • Categories: every article carries categories[] in the payload (first item is primary), derived from your Brand DNA services. CONTRACT: the receiver looks the category up by name/slug; if it EXISTS on your site it gets selected, if NOT it is left unset — we never create categories on your side. If your category names differ, use the mapping table in Site Connection (an external_id is then included).
  • Author Profile: when the author box is enabled in Brand DNA, the payload includes an author field ({name, title, bio, photo_url, linkedin, page_url, past_roles}). The receiver can render an author box under the article (with the photo/name linked and a "View details" button to page_url when set; past_roles is a newline-separated list of roles — a line containing "present" is the current company; suitable for the Person description) and emit Person structured data; a null field means the box is off.

10. Insights: Competitors, Speed, Rank Tracking, AI Visibility

The panel provides insights such as competitor analysis, site speed via Google PageSpeed, SERP rank tracking, and AI visibility. This data comes from paid APIs and is kept in a domain-keyed cache (no needless repeat queries).

Competitor analysis finds real competitors from target keywords even for small sites with no organic data — no fake/mock data is shown.

11. Visibility in AI Search (GEO)

Zep AutoSEO provides an llms.txt (a content guide for AI) and an AI-bot-friendly robots.txt policy so your site can be found in AI search. On WordPress these files are served automatically by the ZepSeo plugin.

Content is written answer-first, in citable passages — this increases the chance of being cited by ChatGPT/Perplexity/AI Overviews.

12. Resolve Conflicts

If two of your articles target the same keyword, the 'Resolve Conflicts' button automatically de-conflicts them — updating even a published article if needed. It can be used up to 5 times per month; when there are no conflicts the button stays inactive.

13. Plans and Limits

Plans differ by number of sites, number of services/products, and monthly article count. As long as your subscription is active, content production and publishing continue; the subscription gates publishing.

See the pricing section on the homepage for current prices and limits.

14. Google Search Console + Analytics

By connecting your Google account in the panel (OAuth), you can view your Search Console and Analytics 4 data in the panel. Only the required scopes are requested; your data is used to improve your content strategy.

Indexing requests: after publishing, indexing requests are sent to Google from the panel so new content doesn't wait to be discovered; track status in the Indexing tab. Connections are PER SITE — each site can link a different Google account/property.

15. Approval Queue & Notifications

Content routed to human review lands in the Approval Queue. You get an email when something arrives, a red counter shows in the sidebar until you act, and a reminder goes out every 3 days. 'Regenerate' produces a fresh title; approving generates the article with it.

16. Content Tools: Fix, Regenerate, Meta

Each article card offers: regenerate text/image (limited credits), regenerate title, regenerate meta (versions kept separately) and 'Fix per SEO' — it rewrites only problematic paragraphs, never changes your focus keyword, and discards changes that don't improve the score. Fixes to published articles are pushed to your site automatically (same post updated, no duplicates).

17. Site Files: llms.txt, robots.txt, sitemap

The panel generates llms.txt, an AI-crawler-friendly robots.txt policy and sitemap content for you. On WordPress the ZepSeo plugin serves them live; elsewhere you can copy and publish them. They are living files, updated as content ships.

Honest note: Google Search does not use llms.txt (it crawls its own index); this file helps AI services like ChatGPT, Perplexity and Claude understand your site correctly. For Google, classic SEO and content quality remain what matters — the system manages both together.

18. Support and System Status

If you run into an issue, check our System Status page first: you can see live which services are up, and send us a support ticket directly from there.

Even if you can't sign in, you can reach us via the form on the status page. You can also always email destek@zepbilisim.com.