Analysis records
Return to the data and judgment at that time.
Keep the site, data window, and full suggestions. When you need to review, you don't have to piece it together from memory.
Open analysis records Read-only Google connection · $9 / monthKeep three records separate: recommendation, change and outcome
AI history retains the earlier analysis. Content revisions and workflow delivery records retain human changes. Neither supplies automatic causal attribution.
On a narrow screen, scroll the table sideways to read the verification column.
| What you observe | What to do | What to retain or verify |
|---|---|---|
| You need to explain an old recommendation | Open its original site, window, filters, input and output before running AI again. | A legacy record without a snapshot stays evidence-limited. |
| A client asks what was implemented | Use content revisions and the assigned task’s actual change/time/verification record. | Share only within an explicitly permitted site; download a selected record/report, not an assumed bulk AI-history export. |
Work through the distinction
A saved recommendation dated Monday, a deployment on Thursday and a later search window answer three different questions. Keep all dates instead of labelling the Monday report “successful”.
First, see what you'll get.
This is an interactive sample. Actual results come from sites you authorize.
Keep context; do not manufacture proof of effect.
Which site
Filter by authorized property to avoid mixing up projects.
Which period
Keep the data window and country, not just conclusions.
What is the evidence
Open the full suggestions and the inputs at that time; no need to rerun the analysis.
A record existing does not mean the suggestion was implemented; there is currently no automatic change tracking or version effect attribution.
Before using it, understand these details.
How it works, data boundaries, and applicable conditions.
What problem does this page solve
Find why a suggestion was made at the time, instead of requesting AI again each time. Select records by site, first verify country, date, and data scope, then read the full output and evidence in new records.
Do not package review as effect tracking
History does not automatically record your site changes, and there is no experiment attribution feature. Please separately note launch times and actual actions, then recheck with consistent criteria; do not treat two analyses before and after as direct growth causation.
Save analysis context
Each record retains the property, data window, country, and full output, so review stays connected to the conditions at that time.
Review by site
Filter by domain and paginate through history, without running another paid AI call just to view old results.
History is not proof of effect
Saving a recommendation does not establish implementation or growth. Content versions and team sharing live in the new operations workspace. Legacy AI history does not automatically track client changes or prove causation.
Review process
Go to analysis records in the workspace, filter by site, find the corresponding window, and view the full output. First verify the property and country, then compare with your own change notes; avoid using global and single-country records directly for trend judgments.
Difference between records and effects
Saving a suggestion only proves the analysis was generated and saved. Implementation requires separate deployment evidence; improvement requires subsequent consistent windows and business data. History is not a customer case study or performance promise.
What notes are worth keeping
Record the official URL, actual actions, launch time, backups, and verification results, and include only necessary information. Do not put credentials, customer emails, or full sensitive data into public notes.
Confirm before deleting
Disconnecting deletes saved tokens, AI history and owned operational data. Preserve needed selected records beforehand. New operations records support site-scoped team sharing and record/report downloads; bulk export of legacy AI history is not claimed. Shared records in other workspaces follow the documented deletion boundary.
Frequently asked questions
Do you need to analyze again to review history?
No, history is read from saved results.
What happens to records after disconnecting?
Disconnecting first deletes this site's token, analysis, and subscription records, then attempts to revoke Google authorization; if revocation fails, it will separately prompt you to handle it in your Google account.
Continue to the next step.
Start with your own site.
Verify the data and analysis first, then decide your next step.