SEO Automation

Automated Content Refresh Workflow

Update existing pages with fresh evidence without losing what made them good.

SEO Automation EditorialPublished March 20, 2026Updated March 20, 20264 min read

Intermediate · Level 2–3 · Tools: GSC, Crawler, LLM, CMS

  1. Select page
  2. Pull performance
  3. SERP diff
  4. Draft updates
  5. Review
  6. Publish
  7. Monitor

A content refresh workflow is a repeatable way to update pages you've already published: new facts, current examples, gaps filled. The automation gathers the evidence and drafts the changes. A person approves the final edits.

Think of a marketing team with 300 blog posts and one writer. Everyone knows half the posts are out of date: old screenshots, dead links, advice that's no longer true. Nobody has time to reread 300 posts. The decay workflow tells you which pages are slipping. This one does the actual updating, one page at a time, the same careful way every time.

At a glance

  • What it does: Update existing pages with fresh evidence without losing what made them good.
  • Difficulty: Intermediate. Needs a few connected data sources and some care with prompts and review steps. Build a beginner workflow first if you can.
  • Automation level: Level 2–3 on the levels of SEO automation. In plain words: anything from a fixed chain of steps (Level 2) up to AI helping with judgement calls (Level 3), depending on how far you take it.
  • Steps: 7, from "Select page" to "Monitor".

What you'll need

Tools

  • GSC: short for Google Search Console, Google's free tool that shows which searches your pages appear for and how many clicks they get.
  • Crawler: a tool that visits every page on your site the way a search engine would and records what it finds: titles, links, errors and more.
  • LLM: a large language model, the kind of AI behind ChatGPT, Claude and Gemini.
  • CMS: the system you publish pages with, like WordPress or Webflow.

What goes in

  • A page to refresh
  • Its Search Console performance
  • Current search results for its main queries
  • A crawl of the page: its text, links and headings

What comes out

  • A side-by-side list of proposed changes
  • An updated draft
  • The refreshed page, published
  • A follow-up note on how it performed

How it works, step by step

  1. Select page. A page gets picked, either by the decay workflow, a rule (say, every post older than a year), or by hand.
  2. Pull performance. It pulls the page's clicks and top queries from Search Console, and crawls it for broken links, missing titles and old dates.
  3. SERP diff. It compares what the top search results (the SERP) cover today with what your page covers, and lists the differences.
  4. Draft updates. An AI model drafts specific changes, section by section, rather than rewriting the whole page.
  5. Review. An editor sees each change next to the original and accepts, edits or rejects it.
  6. Publish. Approved changes go live and the updated date changes.
  7. Monitor. The workflow tracks the page's clicks over the next month and notes whether things got better or worse.

A worked example

Say the page is a 2024 guide to email subject lines. The crawl finds two broken links and a recommendation for a tool that's since shut down. The search results comparison shows that the top pages now include a section on subject lines for mobile previews.

The AI drafts a new section on that, swaps the dead tool for a current one, and fixes the links. It also suggests rewriting your opening, which tells a personal story readers liked. The editor rejects that one and keeps the story. The review takes about half an hour.

Where AI helps, and where it doesn't

  • Spotting what the page is missing compared with what ranks today.
  • Drafting targeted changes instead of a full rewrite.
  • Summarising the changes so the reviewer can check them quickly.

AI tends to smooth everything toward the average. Left alone, it will strip out the opinions, stories and original data that made your page worth reading, and it may stuff in extra keywords. It should propose changes, not overwrite the page.

What to keep human

  • An editor approves the final list of changes before anything is published.

For more on deciding how much to hand over, see the SEO autonomy dial and what not to automate.

What can go wrong

  • The refresh strips out the unique angle, story or data that made the page stand out.
  • Keywords get stuffed in unnaturally during the update, so the page reads worse than before.

Start small

Start with a refresh checklist, not an AI rewrite. For one page at a time, have the workflow report broken links, outdated years, and the headings the top five results use that yours doesn't. Make the edits by hand. Add AI-drafted changes once you know what a good refresh looks like on your site.

Where to go next

Related pages