🔍 What Is the DKN and Why Might It Need a Reset?
Your Domain Knowledge Network (DKN) is the foundation Search Atlas uses to understand your website's topical structure. It maps your content into interconnected nodes — one per page or topic — to power recommendations across Content Genius and OTTO SEO.
Over time, your DKN may develop issues such as:
- Duplicate or overlapping nodes — the same page is recognised as multiple distinct nodes.
- Stale or poisoned keywords — node keywords carry over incorrect or mismatched data from earlier builds.
- Corrupted titles — node titles contain disambiguation code that was not cleaned up automatically.
- Mismatched data after a project switch — data from a previous project appears in a new project's DKN.
If your DKN contains any of these problems, resetting it and rebuilding from scratch is usually the fastest way to restore accurate topical mapping.
⚠️ Before You Reset: What to Know
Please read the following carefully before proceeding:
- A full DKN reset is irreversible. All existing nodes, keyword assignments, and topical connections for the project will be permanently deleted.
- After a reset, the DKN must be fully rebuilt. Rebuilding time depends on the size of your domain and the number of pages crawled.
- Any Content Strategy work tied to existing nodes — such as briefs or cluster assignments — may be affected. Review your active content work before resetting.
- Do not use the Sync with Brand Vault button on the Content Strategy page as a workaround for DKN issues. A known issue means this action can wipe DKN nodes without triggering a proper rebuild, leaving your network in an incomplete state.
🛠️ How to Request a Full DKN Reset
At this time, a full DKN reset must be performed by the Search Atlas support team. There is no self-serve reset button in the platform. Follow these steps to get it done quickly:
- Confirm which project needs the reset. Note the exact project name as it appears in Search Atlas.
- Decide whether you need a full reset (wipe everything and rebuild from scratch) or bulk deletion of specific duplicate nodes (partial cleanup). Both options are supported.
- Open the chat widget in the bottom-right corner of the platform and type human teammate to connect with the support team.
- Provide the following information:
- Project name and URL
- Whether you want a full reset or targeted node deletion
- A brief description of the issue you are seeing (e.g., duplicate nodes, corrupted titles, stale data)
- The team will confirm the action with you before making any changes.
🔄 How to Rebuild Your DKN After a Reset
Once the reset is complete, follow these steps to rebuild a clean DKN:
- Go to your Project Settings and verify that the correct domain and crawl settings are configured.
- Trigger a new site crawl from the dashboard. This allows Search Atlas to rediscover your pages with fresh data.
- Navigate to the Content Strategy section and allow the DKN to regenerate automatically as the crawl completes.
- Once the DKN is rebuilt, review the new nodes to confirm they map correctly to your pages — look for any remaining duplicates or mismatched titles before resuming content work.
❓ Why Are Duplicate Nodes Being Created?
Duplicate nodes can appear when the system has difficulty determining whether a page already exists in the network. This can happen in several scenarios:
- URL variations — the same page is accessible via multiple URLs (e.g., with and without a trailing slash, or HTTP vs HTTPS), causing the crawler to treat them as separate pages.
- Crawl interruptions — a previous crawl did not complete cleanly, leaving partial node data that conflicts with a new crawl.
- Disambiguation code leaks — a known bug caused internal code used to differentiate similar topics to appear in node titles and keywords. This issue has been patched, and a backfill was run to clean affected data. If you still see symbols or codes in your node titles, contact support for a targeted cleanup.
If you notice duplicates appearing shortly after a rebuild, check your canonical tags and redirect settings in your CMS before running another crawl. Fixing URL consistency at the source prevents the same issue from recurring.
✅ Quick Checklist Before and After a Reset
- Note all active briefs and cluster assignments linked to existing nodes.
- Confirm the correct project is selected before requesting a reset.
- Do not use Sync with Brand Vault as a cleanup workaround.
- After the reset, trigger a fresh crawl before reviewing new nodes.
- Check canonical tags and redirects in your CMS to prevent duplicate nodes from reappearing.
💬 Still Need Help?
If you need further assistance, open the chat widget in the bottom-right corner of the platform and type human teammate to be connected with a member of our team.