Evidence Intelligence

Evidence before opinions.

Every conclusion Growthract makes traces back through a visible chain — from what was actually observed, to the evidence behind it, to what that evidence means.

Evidence

Everything on your site itself — pages, metadata, structured data, and whether it can even be crawled.

Observation

What we actually found.

Evidence

The receipt behind every claim.

Finding

What that actually means for you.

The short version

A source can prove one thing.
We don't ask it to prove ten.

See it live

Every report, kept and revisitable.

The evidence trail behind a finding doesn't disappear after one scan. Saved reports keep it available to open again.

Real Growthract Reports interface
Live capture · saved reports, opening a report's evidence

Evidence model

Different layers answer different questions.

Growthract separates what was observed from what can reasonably be inferred and from what should happen next.

Observation

A direct result from a page, integration, AI response, log window or public source.

Evidence

The retained source material or structured result supporting the observation.

Finding

A grounded interpretation of what the available evidence establishes.

Recommendation

A prioritized response to one or more findings, not a substitute for the evidence behind them.

Verification

A later comparable check against defined acceptance conditions.

Evidence sources

The web leaves different kinds of evidence.

Growthract can combine multiple sources while preserving the limits of each one.

First-party website evidence

Page content, metadata, structured data, internal links, crawler controls and other directly observable site conditions.

Search evidence

Connected search performance data when available, kept distinct from inferred visibility.

Observed AI responses

Point-in-time responses preserved with provider and prompt context.

Crawler activity

Observed requests from supplied server logs, separate from robots configuration.

Common Crawl

Evidence that a URL or content appeared in a public crawl corpus. This is not proof of model training, indexing by every AI system or recommendation likelihood.

Wikidata

Public entity evidence when an official-site match can be established. Absence from Wikidata is not, by itself, evidence of poor entity clarity.

Wayback

Historical snapshots that can help establish how public pages changed over time. They are not ranking evidence.

Truth boundaries

A source should never be asked to prove more than it can.

Common Crawl

Presence in Common Crawl is public-web crawl evidence. Growthract does not present it as training proof or as evidence that an AI system will cite or recommend the business.

Wikidata

A matched Wikidata entity can strengthen public entity evidence. Missing Wikidata evidence remains neutral unless other evidence independently supports a problem.

Wayback

Historical snapshots can establish prior public states. They do not establish search ranking, AI usage or business impact.
Ready when you are

See what Growthract already knows about your website.

No setup marathon. Point it at your site and see what is already there.

Analyze your website