Six specialists, one job each.

Vera finds the questions your buyers actually type into answer engines — not the keywords a volume tool thinks they search. It reads your category the way a buyer does, then hands back the prompts worth being cited for.

• Maps buy-intent prompts across discovery, comparison, validation, shortlist and attribute
• Separates the questions engines answer well from the ones they refuse
• Scores each prompt by how often your category gets a brand named at all
• Finds the attribute questions where you have a real claim and no coverage
• Hands off a prompt set your monitoring can re-run unchanged

Atlas makes your brand something an engine can resolve. It fixes the structured data, the naming inconsistencies and the missing profile links that stop a model attaching a recommendation to you at all.

• Audits Organization, Product, Offer, Service and review markup across templates
• Finds naming collisions where your brand blends with an unrelated one
• Builds the sameAs graph that confirms you exist to a machine
• Specifies the markup as a template change, not thousands of page edits
• Re-validates after deploy so the fix is proven, not assumed

Juno writes the pages that get quoted. Comparison and shortlist questions retrieve documents that compare things honestly — including where a competitor wins, which is exactly what makes a page usable as a source.

• Writes you-versus-them pages that name the cases a competitor fits better
• Builds the shortlist and best-of pages your category is missing
• Turns product documentation into self-contained extractable answers
• Matches the phrasing buyers use rather than the phrasing you use
• Every page ships with the schema that makes it machine-readable

Orion works the list of domains an engine cited instead of you. It finds which of them accept contributions, which run roundups you qualify for, and what it takes to appear on each.

• Ranks the source domains actually cited in your category
• Separates the reachable ones from the ones that will never link out
• Finds roundups and comparisons you already qualify for and are missing from
• Drafts the pitch, with the evidence that makes it worth publishing
• Re-measures whether appearing there moved the citation rate

Nova checks that engines can read you at all. Client-rendered copy, blocked crawlers, broken canonicals and stale documentation are why sites with good content still go uncited.

• Verifies your copy exists in the HTML source, not only after hydration
• Checks robots rules against GPTBot, ClaudeBot, PerplexityBot and Google-Extended
• Validates schema, canonicals, sitemaps, llms.txt and heading structure
• Finds superseded pages still crawlable and still being answered from
• Reports Core Web Vitals where they affect crawl and rendering

Echo re-runs your prompt set on a schedule and tells you when the number moves. Answers vary between runs, so a single check proves nothing — the trend is the only thing that does.

• Re-runs an identical prompt set across six engines on a fixed cadence
• Alerts when you lose a citation you previously held
• Tracks competitors on the same queries so a drop has context
• Watches which source domains start or stop citing you
• Keeps every raw response, so any number can be re-read months later