Changelog

Engines occasionally return no answer to a prompt: they decline, time out or return an error. Until this release, Cited counted those runs as prompts where your brand wasn’t cited, which lowered Share of Answer by up to 3 points for some workspaces.

• Runs with no answer are now excluded from the Share of Answer denominator

• A new No answer column shows how many runs were excluded, by engine

• Historical Share of Answer has been recalculated back to each workspace’s first run

• CSV and Snowflake exports include a new answer_status field

If you report Share of Answer externally, you may see small upward revisions to past weeks. The largest change across all workspaces was 2.8 points.

Content Briefs are now available on every plan. Each brief targets a group of prompts where competitors are cited and you aren’t, and shows what the cited pages cover that yours don’t.

• Briefs ranked by opportunity, based on prompt count, citation concentration and your domain’s existing citations

• Suggested structure, including headings, a summary block and comparison tables

• An update-or-create recommendation when an existing page covers part of the topic

• Export to Notion, Contentful, WordPress and Framer as drafts

Read the launch post for results from the beta, where 61% of briefed pages earned a first citation within 30 days.

Crawler Insights now groups AI crawler visits by declared purpose, using the user agents each operator publishes. You can see which crawlers fetch your pages to answer questions in real time and which collect data for training.

• New purpose filter: live retrieval, training and undeclared

• Response-code breakdown per crawler, including blocked and timed-out requests

• A list of pages with the most crawler errors in the last 30 days

• A robots.txt check that shows which crawlers your current rules allow

Requests that claim a known crawler’s user agent but can’t be verified against the operator’s published IP ranges are now flagged separately.

Cited now exports your data to Snowflake and Google Sheets. Warehouse exports include every prompt run and the full answer text, so data teams can build their own models.

• Daily Snowflake sync with documented, versioned tables

• Backfill of up to 12 months on first connection

• Scheduled Google Sheets exports for any report, refreshed daily or weekly

• Column selection and engine filters for each export

Setup takes about ten minutes. See the Snowflake and Google Sheets integration pages for step-by-step instructions.

Engines sometimes cite the same page under several URLs: with tracking parameters, with or without a trailing slash, or through a redirect. Cited treated these as different sources, which split citation counts and understated your top pages.

• Citations now resolve to the canonical URL after following redirects

• Common tracking parameters, such as utm_ tags, are removed before matching

• Trailing-slash and www variants are merged

• Citation Sources history has been reprocessed for the last 12 months

Some pages will now show higher citation counts than before. Totals across your domain are unchanged.

Answer Diff now shows two versions of an answer side by side, with added and removed text highlighted and citations listed beneath each. You can compare any two weeks, not just the latest pair.

• Side-by-side view with inline highlights for added and removed text

• Citation changes listed separately from text changes

• Alert rules for lost citations, new competitor mentions and changes to brand claims

• A weekly digest of the largest changes by prompt group

Alerts are delivered by email and Slack. Existing alerts have been migrated to the new rules automatically.