Platform Support

⚡ AI Recommendation Generation: Batch Size Limits & Timeout Fixes

Overview When you run AI recommendation generation across a large set of pages, jobs that include too many pages in a single batch may time out before completing. This article explains the cause and the recommended approach for working through large page sets without losing progress. Why Large Batches Time Out AI recommendation generation sends each page through a processing pipeline that analyses content, evaluates on-page signals, and produces actionable suggestions. Each page adds to the total processing time for the job. When a single batch contains too many pages, the cumulative processing time can exceed the system's job timeout threshold, and the job may be cancelled before all pages are completed. - Pages already processed before a timeout are not lost — their recommendations are saved and visible once the job ends. - Remaining unprocessed pages will show no recommendations until you rerun generation for them. Recommended Batch Size To ensure every job completes successfully, limit each AI generation run to a smaller subset of pages rather than selecting your entire page list at once. Smaller batches process well within the timeout window and give you predictable, reliable results. You can run multiple batches back-to-back — each time selecting only pages that have not yet been processed. How to Process Large Page Sets in Batches Follow these general steps to work through a large number of pages without hitting the timeout limit. 1. Open the AI generation tool you are using. Navigate to the relevant tool in your Search Atlas account where you run AI recommendation generation. 2. Filter for pages without recommendations. Use any available filters to display only pages that have not yet received AI-generated recommendations. This keeps your working list clean and prevents duplicate processing. 3. Select a small subset of pages. Choose a manageable batch of pages from your filtered list. Keeping batches small reduces the risk of hitting the timeout threshold. 4. Trigger AI generation for your selection. Start the generation job for your selected pages and allow it to run to completion before starting the next batch. 5. Wait for the job to finish. Once the batch is complete, the recommendations for those pages will be available. 6. Repeat for the next batch. Return to your filtered list, select the next set of unprocessed pages, and repeat the process until all pages have been covered. What to Have Ready if You Need Support If you continue to experience timeouts even with smaller batches, please have the following information ready when you contact support: - The name of the project and the specific tool where the issue is occurring - The approximate number of pages you were processing when the timeout occurred - Any error messages or job status indicators you observed 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.

🔍 Fix an Empty Domain Knowledge Network

🧩 What Is the Domain Knowledge Network? The Domain Knowledge Network is a visual map that shows how your business entity connects to related topics, people, places, and concepts across the web. It is powered by the data you enter in your Knowledge Graph, but it relies on additional processing steps before it populates. ⚠️ Why It May Appear Empty A common point of confusion is assuming that completing 100% of the Knowledge Graph fields will immediately fill the Domain Knowledge Network. In reality, there are a few reasons why the network may still appear empty even after full completion: - Processing delay: After your Knowledge Graph reaches 100% completion, the platform needs time to process and build the entity connections. This can take up to 24–48 hours. - Entity recognition pending: The system must verify and match your business entity against external knowledge bases. If your business is newer or has limited online presence, this step may take longer. - Missing or inconsistent NAP data: Your business Name, Address, and Phone number (NAP) must be consistent and complete. Even small discrepancies can prevent entity matching from completing successfully. - Schema or website data not yet crawled: If your website was recently updated with structured data or business schema, search engines and the platform may not have re-crawled it yet. ✅ Steps to Resolve the Issue 1. Confirm 100% Knowledge Graph completion. Navigate to your Knowledge Graph dashboard and verify every section — including business categories, social profiles, founding details, and key people — is fully filled in. Partial entries in optional fields can still affect processing. 2. Check your NAP consistency. Make sure your business name, address, and phone number exactly match what appears on your website, Google Business Profile, and other online listings. Even minor differences (e.g., "St." vs. "Street") can cause issues. 3. Verify your website URL is correctly entered. An incorrect or redirecting URL can prevent the platform from crawling your site and confirming your entity data. 4. Wait 24–48 hours and refresh. If everything looks correct, allow up to two full days for the Domain Knowledge Network to populate. Refresh the page after this period to check for updates. 5. Save and re-submit your Knowledge Graph. Try opening your Knowledge Graph, making a minor edit to any field, saving the changes, then reverting and saving again. This can trigger a re-processing of your entity data. 💡 Tips to Improve Entity Recognition The strength and accuracy of your Domain Knowledge Network depends on how well your business entity is established online. Consider the following best practices: - Ensure your business has a Google Business Profile that is fully verified and up to date. - Add structured data (Schema.org markup) to your website, especially LocalBusiness or Organization schema types. - Maintain consistent business listings across major directories such as Yelp, Bing Places, and industry-specific platforms. - Link your social media profiles (LinkedIn, Facebook, Twitter/X) within the Knowledge Graph to strengthen entity signals. 🔄 When to Expect Results Most customers see their Domain Knowledge Network begin to populate within 24–48 hours of completing a fully consistent Knowledge Graph. Businesses with a stronger existing online presence tend to see faster results. If the network remains empty after 48 hours and you have followed all the steps above, it is worth reaching out to our team for a manual review. 🛠️ Still Seeing an Empty Network? If your Domain Knowledge Network is still empty after 48 hours and you have verified all the steps in this article, our team can investigate your specific entity data and processing status. 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.

How to Reset Your Domain Knowledge Network

Understanding the Dynamic Knowledge Network The Dynamic Knowledge Network (DKN) powers your OTTO SEO content generation. It scans your domain to understand your business, products, and existing content. This ensures that when the AI writes new articles, the information is accurate and highly relevant to your brand. When You May Need a DKN Reset Sometimes, your website might undergo significant changes. You may need to request a DKN reset to reflect new services, updated branding, or a complete site overhaul. A reset ensures your future content perfectly matches your current business profile. - Updated Content: You recently added new service pages or a new blog. - Business Pivot: Your core business focus or target audience has changed. - Duplicate Prevention: You want to ensure the AI does not generate articles that are too similar to content you already published. Requesting a Full DKN Reset If you need a completely fresh start—such as deleting the existing DKN entirely to clear stale data or wipe the slate clean—you will need our team to process a full reset. Since this action completely wipes the current knowledge base for a specific project, it must be handled manually by our support staff to ensure no data is lost accidentally. To request a full reset, please prepare your Project Name, Domain Name, and DKN ID. 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.

Order Troubleshooting

Q: What is a broken link and why does it matter? A: A broken link points to a missing page (e.g., 404). Left unfixed, it wastes crawl budget, interrupts link equity, frustrates users, and can erode trust. Regular checks keep your site healthy and your rankings resilient. Q: What causes broken links, and how do I prevent them? A: Common causes include deleted/moved content, URL changes, typos, removed external resources, protocol/server issues, redirect chains/loops, and soft‑404s. Prevent issues by keeping URL structures stable, testing links before publishing, and running routine scans. Q: How do I find and fix broken links? A: Run scans with OTTO SEO or a Site Audit to identify errors and redirect chains. Then: fix typos, 301 redirect to the closest relevant page, restore removed content if appropriate, or ask external site owners to update broken backlinks. After fixes, monitor GSC crawl errors, bounce rates, organic traffic, and rankings for improvement. Q: How often should I scan for broken links? A: Large/fast‑changing sites: weekly to bi‑weekly. Smaller sites: monthly. All sites should run a comprehensive audit quarterly and within 48 hours after migrations or redesigns. Q: My links aren’t indexed, can you help? A: Yes. Share your Order ID and live URL(s) and we’ll submit them for indexing. Typical turnaround is 3–5 business days once we have the URLs, though final timing depends on search engines. We’ll notify you as soon as indexing is detected.

🗂️ Search Atlas Module Ownership Guide

Purpose of this guide This guide helps you direct questions and feedback to the right area when working with Search Atlas. Module ownership can change as products evolve, so use the module name and workflow details when requesting assistance. Module responsibility overview When raising a question or issue, identify which of the following Search Atlas modules it belongs to, then include that module name along with the full context of your request: - OTTO (Automated SEO): Questions about reviewing, approving, or monitoring automated on-page and technical optimization recommendations generated by OTTO. - Content Genius (Content workflows): Questions about article generation, AI-assisted writing, content planning, and bulk content tasks within Content Genius. - On-Page Audit: Questions about page-level analysis, topical terms, questions, audit reports, and content score tracking. - Rank Tracker: Questions about keyword performance, visibility trends, rank movement, and report sharing. - Site Auditor (Technical SEO and site health): Questions about crawling, technical findings, indexing checks, and website performance insights. - Projects and account settings: Questions about project setup, website connections, user access, permissions, and account configuration. How to route a question clearly 1. Identify the specific module from the list above that best matches your question. 2. Note the project name, the page or keyword involved, the action you were attempting, and the result you expected versus what actually happened. 3. Include all of that context when submitting your request — for example: "I attempted an OTTO automated SEO action in Project A and the recommendation did not apply as expected." 4. If you are unsure which module applies, describe the workflow you were using and the outcome, and the support team will route it appropriately. Product ownership and escalation Search Atlas does not maintain a fixed public list of individual owners for every module. Internal ownership may change, and a feature can involve multiple teams, such as product, engineering, SEO, or customer education. The most reliable route is to identify the affected module, gather the full context (project name, exact action taken, expected vs. actual result), and submit that detail through the platform support channel so it can be routed to the correct team without back-and-forth. If you need further assistance, open the chat widget in the bottom-right corner of the platform and type human teammate to be instantly connected with a member of our team.