March 2026 Changelog
March was the month SCM moved more clearly toward topic clusters and Page Pipeline workflows. The updates connected topic discovery, clustering, page templates, and generated pages into a more complete SEO content production path.
The work also improved multilingual handling, SERP extraction, Reddit discovery, task logs, and recovery from crashes. If February cleaned up the app, March pushed SCM toward a more structured SEO workflow.
Who this update helps
- Users building topic clusters before creating pages.
- Users moving clustered topics into Page Pipeline workflows.
- Users working with Chinese, Japanese, Thai, or other non-English data.
- Users relying on SERP extraction, Reddit discovery, and local listing collection.
Topic clusters became a stronger workflow
March added options and reliability fixes around topic clustering. SCM became better at finding, adjusting, copying, and moving clustered topics toward page creation.
- Topic clusters can now flow into Page Pipeline.
- Seven AI page templates were added for use with the clustering engine.
- Users can add or remove topics by type during clustering.
- Cluster data can be copied in CSV format.
- A toggle was added to skip Reddit topic clustering when it is not needed.
- Topic cluster templates now use clearer tag styling.
- Topic cluster keyword handling and clustering speed were improved.
Better multilingual and entity handling
Several fixes focused on language edge cases that can break content workflows. This matters for users collecting international SERPs or generating content in languages where tokenization and entity detection are harder.
- Page Pipeline no longer crashes on Chinese language input in the affected path.
- Japanese key term extraction no longer causes a crash.
- Entity detection improved for Chinese, Japanese, and Thai.
- Entity extraction now streams tokens to reduce memory pressure when processing many sites.
More dependable SERP and source extraction
SCM improved how it gathers page and SERP evidence before writing. The changes reduce hangs, improve fallback behavior, and make long extraction tasks easier to stop.
- SERP heading extraction now has a timeout so a task does not hang indefinitely.
- SERP extraction can try a simple fetch first, then fall back to an offscreen browser window when needed.
- Autocomplete extraction is less likely to return zero results.
- Reddit scrolling was improved to find more results.
- Users can abort while SCM is finding clusters.
- Local listings handling of Google language parameters was fixed.
App stability and task-log usability
March also tightened the app shell around crashes, logs, and long-running task visibility.
- The UI window now attempts to restart itself if it crashes.
- Log messages are batched to avoid overwhelming the UI.
- Long log messages no longer get cut off in the log window.
- Task log updating became more reliable.
- Large task logs and job tables became faster to click through.
- The login window and app init error window received fixes.
Related SCM guides and product history
- Reworking SCM around topic clusters
- How to build a keyword clustering system
- SCM is now a programmatic SEO engine
Full developer changelog
The summary above is written for users. The original 49-item developer changelog is preserved below for exact release detail.
March 26, 2026
- fix: Page pipeline crashing on chinese language
- fix: Getting Japanese key terms won't cause a crash
- fix: Old ai prompts that didn't have any AI instructions in them was being ignored.
- fix: The UI window will attempt to restart itself if it crashes
- fix: Performance throttling of log messages, now they batch to avoid overwhelming UI updates
March 24, 2026
- fix: 15 sec timeout when extracting serp headings so task doesn't hang
- feat: topic cluster new options to add or remove topics by type. Also removed pre filter (seems like not needed)
- feat: New reddit scrolling algo for finding max results.
March 23, 2026
- feat: Allow abort while finding clusters
- fix: tweaked clustering algo for speed
- feat: Copy cluster will give you data in CSV format
- fix: Better autocomplete extraction, less likely to return 0 results.
- fix: Added new navigation method for local listings to fix INVALID ARGUMENT error
- style: Update button is blue
March 22, 2026
- fix: topic cluster keyword crash
- style: Updated pageflow icon to be less distracting and styled sidebar more minimally
- fix: Stream tokens when finding entities, lower mem pressure when extracting many sites.
- fix: Better entity detection for chinese,japanese,thai.
March 21, 2026
- fix: Local listings, better =hl handling
- feat: Add toggle for skip reddit topic cluster
- fix: Clustering topic bug fixes
- fix: When adding and removing AI templates, refresh the dropdown select items too
- style: Topic cluster template use tag styling
- other: Add back AI page threads setting
- fix: Add 1 retry when trying to get word embeddings for topic cluster
- fix: Local listings, adding &hl after each retry incorrectly.
March 20, 2026
- fix: Disable browser page extractor while the kinks get worked out
- style: Log window shows contact support button properly aligned
- feat: SERP extractor, try simple fetch, if no content move onto offscreen browser window.
March 19, 2026
- fix: Long messages in log window was getting cut off.
- fix: Task log not updating correctly
- style: Inc font size task log
- fix: Improved ui performance when clicking on tasks with large logs.
- fix: Improve ui table performance when clicking on job rows
- fix: Bing search using wrong url with '//' causing strange SERP results
- fix: Page pipeline inf wait if the queue is empty
- fix: App init error window fixed style
March 18, 2026
- fix: Missing app bar on login window
- feat: topic cluster -> page pipeline
- feat: 7 ai page templates added for use with clustering engine.
- feat: Improved webhook discovery and error handling
March 13, 2026
- feat: Sort AI templates by tag and name
- feat: Resume/Restart on task edit page. Now when task is resuming it will keep track of keywords processed, not just the number of rows processed. If you add or remove keywords resume will handle it better
March 7, 2026
- fix: V align prog bar in task log
- other: rename article creator -> page pipeline
March 5, 2026
- fix: Local listings, remove symbols from about section
- fix: local listings will now correctly add language to direct google place urls
- feat: When extracting page content, SCM will load it via a browser and render it instead of just downloading a html string.
March 2, 2026
- feat: Walking epsilon for topic clustering
See all monthly updates in the SCM changelog archive. New users can also try SCM from the trial signup page.