Site evidence
A crawl record can distinguish what was fetched, what was skipped and where coverage stopped instead of treating an incomplete pass as a complete audit.
01 A CMT Web product
SEO operations grounded in evidence.
Kepler keeps the data, reasoning and limits behind each suggested next step together—so a team can decide what deserves attention before it changes anything.
Built for reviewable SEO work—not ranking promises or automatic client-site changes.
02 How it works
A process that keeps the why attached to the work. How Kepler works
The useful part is not a bigger dashboard. It is being able to move from an observation to a decision without losing what made the work worth doing. Five steps from evidence to a measured result.
- 01
Collect
Bring approved site and first-party observations into a bounded record with its source and collection limits.
- 02
Interpret
Turn observations into clear, reviewable signals instead of a score that asks you to trust a black box.
- 03
Prioritize
Rank the work that merits attention, with the evidence, confidence and unanswered questions still visible.
- 04
Decide
Keep a person in charge of what changes, where it changes and whether the available proof is enough.
- 05
Measure
Record what happened after an approved change, including neutral and inconclusive outcomes.
Collect Collect approved site data.
Bring approved site and first-party observations into a bounded record with its source and collection limits.
Interpret Find the signals worth reviewing.
Turn observations into clear, reviewable signals instead of a score that asks you to trust a black box.
Prioritize Choose the work that matters.
Rank the work that merits attention, with the evidence, confidence and unanswered questions still visible.
Decide Review and approve every change.
Keep a person in charge of what changes, where it changes and whether the available proof is enough.
Measure Measure what happened next.
Record what happened after an approved change, including neutral and inconclusive outcomes.
03 What stays visible
A recommendation should be explainable without a meeting. See the reason behind each recommendation
Kepler is designed to keep the limits as visible as the signal, so the next step is easier to defend, defer or reject. The source, reasoning and limits stay attached.
First-party context
Search and analytics inputs are treated as authorized business data with freshness, coverage and source limits—not as universal truth.
Opportunity reasoning
Suggested work carries the observations and conditions that support it, so a team can challenge the recommendation before acting.
Change history
A useful record connects an approved change to its scope, its review trail and the later measurement rather than assuming every change caused a result.
Site evidence What was crawled, skipped or left uncovered.
A crawl record can distinguish what was fetched, what was skipped and where coverage stopped instead of treating an incomplete pass as a complete audit.
First-party context Search and analytics data, with source limits.
Search and analytics inputs are treated as authorized business data with freshness, coverage and source limits—not as universal truth.
Opportunity reasoning Why a suggested change deserves attention.
Suggested work carries the observations and conditions that support it, so a team can challenge the recommendation before acting.
Change history What changed, who reviewed it and what followed.
A useful record connects an approved change to its scope, its review trail and the later measurement rather than assuming every change caused a result.
04 Guardrails
Useful analysis has limits. Kepler says where they are. You stay in control
It is built to support careful work: make an informed decision, preserve the record and know when the available evidence is not enough. Human review. Clear limits. A record of each decision.
-
No ranking promises
Search outcomes depend on many factors outside a tool. Kepler is designed to make the work and its evidence clearer, not to guarantee a position.
-
No automatic client-site changes
Recommendations are not publication authority. A human reviews the proposal, scope and rollback path before any controlled change.
-
Evidence can be incomplete
Missing, sampled or stale inputs stay visible. When the record cannot support a conclusion, the honest answer is that it cannot decide yet.
-
Boundaries stay explicit
The product is designed around organization boundaries, review trails and scoped access so one client’s evidence does not become another client’s context.
No ranking promises Better evidence, without guaranteed rankings.
Search outcomes depend on many factors outside a tool. Kepler is designed to make the work and its evidence clearer, not to guarantee a position.
No automatic client-site changes A person reviews and approves changes.
Recommendations are not publication authority. A human reviews the proposal, scope and rollback path before any controlled change.
Evidence can be incomplete Missing or stale data stays visible.
Missing, sampled or stale inputs stay visible. When the record cannot support a conclusion, the honest answer is that it cannot decide yet.
Boundaries stay explicit Client data and access stay separate.
The product is designed around organization boundaries, review trails and scoped access so one client’s evidence does not become another client’s context.
Make the next SEO decision easier to defend
Open Kepler to see the product, or talk with CMT Web about the work your site needs next.