Reports built only from connected evidence
Three report views — Executive, Revenue, Customer — assembled exclusively from the Shopify, behaviour, Copilot, CRO and Comver module data connected to your store. Every figure is a count of something Comver generated, a sum of estimates it labelled, or a number Shopify synced. No sample data, no industry benchmarks, no modelled recovered revenue: an empty chart says it is empty.
- Assembled from connected data only
- No example reports — ever
- Every figure labelled counted or estimated
Shopify · behaviour · scan → change prepared and applied → measured before / after → ≈ estimate until counted
Executive reports Live
Assembled from Shopify, behaviour, Copilot, CRO and module data
AI daily briefs
51
Generated
Completed audits
6
Storefront scans
AI actions completed
4
Measured executions
Open recommendations
5
Waiting on a decision
Revenue reports
Sum of every estimate · not measured revenue
Customer reports
Lifetime revenue spread
3
Report views
Executive · Revenue · Customer — one tab each, four metrics per tab
5
Connected data families
Shopify, behaviour, Copilot, CRO score and Comver module data — nothing else feeds a report
8 weeks
Evidence generated, plotted
Audits, daily briefs and recommendations per week — plotted even when every week is zero
0
Example reports
“No example reports are shown.” An empty tab offers Generate daily brief or Ask Copilot instead
Figures on this page are illustrative. Your reports show what your store generated and synced, or say “No generated reports” until it does.
Three views of the same evidence, none of it invented
Every tab opens with the outcome-ledger rail — Source → Execution → Evidence → Revenue impact — then four metrics and one chart. The Executive tab counts what the AI produced, the Revenue tab sums what it estimated, the Customer tab reads what Shopify synced.
What Comver produced for this store, week by week
AI daily briefs generated, completed audits, AI actions completed (“Measured executions”) and open recommendations (“Waiting on a decision”) — with an eight-week column chart of audits, daily briefs and recommendations. If nothing has been recorded, the chart says so in words rather than showing a sample.
Executive reports
Assembled exclusively from connected Shopify, behaviour, Copilot, CRO and module data.
AI daily briefs
51
Generated for this store
Completed audits
6
Storefront scans finished
AI actions completed
4
Measured executions
Open recommendations
5
Waiting on a decision
Evidence generated
Eight weeks · counts only
Latest evidence
Newest first
Five connected families — and one deliberate gap
A report can only cite what is connected. Each family below is shown Connected or Unavailable in Copilot’s data-sources grid; anything Unavailable is simply absent from the report, never approximated.
Shopify
Products, orders and customers read live from the Admin API; customers synced to Comver. Lifetime revenue, returning customers and orders per customer come from here.
Learn moreBehaviour
Sessions, events, heatmaps, funnels and traffic sources from the Comver storefront tracking. This is what measures an applied change before and after.
Learn moreCopilot
Daily briefs, recommendations with confidence and monthly potential, prepared changes and completed measurements — the rows behind “AI actions completed”.
Learn moreCRO score
Completed storefront audits, the CRO score, revenue at risk and the leak estimates from the latest scan — the “Completed audits” count and part of the estimate sum.
Learn moreComver modules
Which revenue modules are active, their configured estimates and the orders they attributed by discount code or cart token.
Learn moreWhat is deliberately missing
No example reports, no industry benchmarks, no modelled recovered revenue. “Recovered” stays “Not measured yet” until measured attribution exists — an empty chart says it is empty.
One trace from source to revenue impact
Every report opens with the same rail because every figure follows the same path. A source produces evidence (a scan, a brief, Shopify orders); Copilot turns it into a change that is prepared, approved and applied — reversibly; measurement compares behaviour before and after (at least 24 hours and 50 sessions per window); and the result lands in the ledger as an estimate while it is estimated, and as a counted figure only once it is attributed.
- “Every measured result returns to one revenue trace” — the same canonical revenue-at-risk figure is printed on Overview, CRO score, Revenue recovery, the Action Plan and Reports
- Estimates carry an ≈; counts carry none; “Not measured yet” is written out instead of a zero that could be mistaken for a result
- The rail is the same on all three tabs, so an executive, a growth lead and a CRM owner read the same chain of custody
Source
- Storefront scan · CRO score 64
- Daily brief · today 06:30
- Shopify orders + Comver sessions
Execution
- Change prepared · reversible
- Approved · applied Aug 12
- Rollback snapshot kept
Evidence
- ≥ 24 h · ≥ 50 sessions per window
- ATC rate 8.9% → 10.4%
- Measurement completed
Revenue impact
- ≈ $6,400 / mo while estimated
- Counted only when attributed
- “Recovered” · Not measured yet
No example reports are shown. An empty tab says so — “Eight weeks are plotted and every one of them is zero” — and offers Generate daily brief · Ask Copilot for recommendations · Check the store connection.
What is counted, what is estimated, what is never claimed
The honest version of a reporting page: which figures are rows in your store’s own data, which are labelled estimates, and which the product refuses to show at all.
| Counted from your rows | Estimated, labelled ≈ | Not shown | |
|---|---|---|---|
| AI daily briefs generated | |||
| Completed audits (storefront scans) | |||
| AI actions completed · measured executions | |||
| Open recommendations | |||
| Estimated impact by recommendation | |||
| Synced customers · returning · lifetime revenue | |||
| Recovered revenue | “Not measured yet” until attribution exists | ||
| Example reports · industry benchmarks | Never |
From an empty tab to a report worth forwarding
- 1
Connect the store
Install the Comver Shopify app. Shopify customers sync; products and orders are read live. Add the storefront tracking when you want behaviour data and measured before/after evidence.
- 2
Generate evidence
Run the AI storefront scan, let the daily brief arrive every morning at 04:30 UTC, and let Copilot write recommendations. Each of these is a row that a report can count.
- 3
Read the three tabs
Executive for what the AI produced, Revenue for what it estimated and shipped, Customers for what Shopify synced. Four metrics and one chart per tab, the ledger rail above each.
- 4
Generate new evidence
When a tab is thin, the CTA is explicit: Generate daily brief, Ask Copilot for recommendations, or Check the store connection. Nothing is filled in for you.
When there is nothing to report, the report says so
The product’s own copy for a store that has not generated evidence yet — printed here because it is the clearest statement of what Reports will and will not do.
“No generated reports”
Executive tab, before any evidence
“Generate a daily brief or complete a CRau AI audit. No example reports are shown.”
“Every one of them is zero”
Chart footnote
“Eight weeks are plotted and every one of them is zero — nothing has been recorded for this store yet.”
“No customers synced yet”
Customer tab, before the first sync
Followed by “Check the store connection” — not by a sample cohort
Reports plot counts of audits, briefs and recommendations and sums of estimates. They do not claim recovered revenue, before/after test results or industry comparisons.
Frequently asked questions
Only what is connected to your store: Shopify (products, orders, customers), Comver behaviour data (sessions, events, heatmaps, funnels, traffic sources), Copilot (briefs, recommendations, changes, measurements), the CRO score and its audits, and Comver revenue-module status and attribution. Nothing else feeds a report.
No. The product states it plainly: “No example reports are shown.” Before your store has generated evidence, the Executive tab reads “No generated reports” and offers to generate a daily brief or run an audit; the chart footnote says every one of the eight plotted weeks is zero.
Executive: AI daily briefs, completed audits, AI actions completed (“Measured executions”), open recommendations (“Waiting on a decision”), plus the eight-week evidence chart. Revenue: scored opportunities, estimated impact (“Sum of every estimate”), completed actions (“Shipped and measured”), open recommendations, plus ranked estimated impact by recommendation. Customers: synced customers, returning customers (“Two or more orders”), lifetime revenue, orders per customer, plus the lifetime-revenue spread.
No. It is the sum of the monthly estimates attached to open recommendations and audit leaks, and it is labelled as such. Recovered revenue is a separate stage that reads “Not measured yet” until measured attribution exists — it is never derived from the estimates.
The rail at the top of every tab: Source → Execution → Evidence → Revenue impact. It shows the path every figure took — a scan, a brief or Shopify orders as the source; a prepared and applied change as execution; a before/after measurement as evidence; and an estimate or a counted, attributed figure as impact. “Every measured result returns to one revenue trace.”
Each applied change ships with a measurement plan. After at least 24 hours and 50 sessions on each side, Copilot compares the primary metric before and after (add-to-cart rate, checkout rate, conversion or module click-through) and records the result; the History timeline logs “Measurement completed: … changed from X% to Y%.” That row is what the Executive tab counts.
Reports are read inside the Comver workspace under Workspace → Reports; there is no scheduled email or file export today. The daily brief itself refreshes automatically every morning, and “Generate new evidence” runs a brief or recommendations on demand.
The Executive tab is written for a founder or head of growth who wants to know what the AI did this week; the Revenue tab for whoever owns the action queue; the Customers tab for whoever owns the pipeline. All three read from the same rows, so they never disagree with each other or with the Overview.
Reports you can forward without a footnote
Connect the store, generate the first evidence and read three tabs that only ever cite what is connected. Counted where it is counted, estimated where it is estimated — and blank where it should be.
14-day free trial · Live on Shopify in minutes · Checkout stays in Shopify