Measures the proportion of a large language model's training data that contains references to a brand.
Digital marketers or content strategists reading about AI search optimization and visibility indicators.
01What it is and how it works
When an AI search system builds its knowledge base, it crawls the web, indexes public documents, and ingests partner feeds. Corpus Coverage is the ratio of those ingested items that mention or are authored by your brand. The higher the ratio, the more likely the model will surface your messaging in response to relevant prompts. The calculation happens behind the scenes: each document is tagged with source metadata, then a simple count of brand‑related documents is divided by the total document count.
It is the share of a model's data that includes your brand.
02What to do about it
Take these actions this week to improve your brand's coverage: 1. Audit your website for missing schema.org markup and add it where appropriate. 2. Publish at least one new, high‑quality blog post that includes clear brand mentions. 3. Submit an updated sitemap to Google Search Console and request a recrawl. 4. Add your brand's official social profiles to the robots.txt Allow list if they are blocked. 5. Reach out to industry directories and ask them to include a verified listing.
03How it is measured or noticed
The product dashboard shows a percentage called "Corpus Coverage" next to each brand. Under the hood, the system runs a query against its indexed corpus: SELECT COUNT(*) FROM documents WHERE brand_id = X. The result is divided by the total document count for the model's training slice. You can also see a trend line that updates after each crawl cycle, typically every 24‑48 hours.
04Common mistakes
- Relying on a single page to boost coverage – the model samples many sources, so one page has minimal impact.
- Blocking brand assets with
robots.txt– if the file disallows/blog/, those pages never enter the corpus. - Using generic brand mentions without context – vague references are filtered out during relevance scoring.
05Limits
Corpus Coverage only reflects the presence of brand content in the training slice, not the quality of the answers. A high coverage score does not guarantee top‑ranked placement in AI search results. The metric also excludes private or pay‑walled content that the model cannot access. Do not confuse it with SEO rankings; it is a visibility indicator for generative models, not a traditional SERP metric.
06Worked example
"After adding structured data to our product pages and publishing a quarterly whitepaper, our Corpus Coverage rose from 12 % to 27 % within two crawl cycles. The AI assistant started quoting our specs in three new user queries."
Frequently asked questions
How does Corpus Coverage differ from other brand visibility metrics?
It depends on what you are measuring. Corpus Coverage looks at the share of a language model’s training data that mentions your brand, while typical visibility metrics track impressions or clicks on published content. The former tells you how often the model can draw on your owned material, not how often users see it.
Should we invest in increasing Corpus Coverage for every brand in our portfolio?
Usually you prioritize brands that drive the most revenue or have strategic importance. Raising Corpus Coverage for low‑impact brands may consume resources without a clear return, whereas focusing on key brands can improve AI‑driven search relevance where it matters most.
How is Corpus Coverage calculated on the product dashboard?
It works by comparing the number of indexed documents that contain references to your brand against the total number of documents in the model’s training slice. The system reports the result as a percentage next to each brand, giving you a quick view of coverage depth.
Does a low Corpus Coverage mean our brand will never appear in AI‑generated answers?
No, a low percentage only indicates limited presence in the training data, not an absolute ban. The model may still surface your brand if other signals, like strong contextual relevance, trigger a mention.
What are the risks of ignoring a low Corpus Coverage score?
The main risk is that the AI may default to competitors or generic information when answering user queries. You might notice a drop in brand recall in AI‑driven search results, which can affect perception and lead generation.
How long after we add new brand content will Corpus Coverage reflect the change?
Typically the update cycle ranges from a few days to a couple of weeks, depending on the model’s ingestion schedule. In the meantime you can monitor interim signals, such as crawl logs, to gauge when the new content is likely to be indexed.
Asked out loud
spoken, not typedThe same term in the words somebody uses speaking to an assistant rather than typing into a box — written from the situation, which is why each one carries the situation it came from.
Yes, you can see the current Corpus Coverage percentage on the dashboard, which tells you how often the model can reference your brand. If the number is low, consider adding recent press releases or product pages to improve the score before the presentation.
Usually the mobile view of the dashboard displays the same Corpus Coverage metric as the desktop version. Open the app, select your brand, and you’ll see the percentage instantly, so you can act before the meeting starts.
It depends on the current coverage figure; a value above 70 % generally indicates solid representation, while anything lower may need reinforcement. Review the dashboard now and, if needed, push a fresh blog post or partner feed to boost the score before the call.