Example two
A monthly report
The same business, two weeks later, as a monthly plan client. It shows the report card, the sources behind it, the numbers — including the ones the sources did not provide — and the decisions it hands back.
Example — inventedReport card for the invented monthly report: the client, the period, the plan, the versions, the issue date and the strategy call.| Client | Anystreet Family Medicine (DEMO — fictional practice; no such practice exists) |
|---|
| Location(s) | One: Anystreet Family Medicine, 1200 Placeholder Way, Suite 210, Anytown, TX 00000 (fictional address). Single location, so no figure in this report blends two places. |
|---|
| Period | 2026-10-01 .. 2026-10-31 inclusive, America/Chicago |
|---|
| Plan | AI Visibility |
|---|
| Report version | v1 — first report for this client; no correction and no reissue |
|---|
| Dictionary version | METRIC-DICTIONARY-v1, dated 2026-09-22. Unchanged this period; nothing was applied retroactively. |
|---|
| Prepared by / approved by | DEMO placeholders — no such people; the business currently has no named reviewer, which is an open decision, not a fact about this report |
|---|
| Date issued | 2026-11-06 (Friday) = 5 business days after period close (2026-10-31, a Saturday) |
|---|
| Strategy call | 2026-11-10 (Tuesday) 11:00 CST, 30 minutes — within 7 business days of the report landing; the agenda is fixed and the same every month |
|---|
How to read the table
“If a metric is missing from this table, it is missing — not zero. Missing reasons are in section 5.”
One row per metric that should exist, including the metrics that got no data. A row is never removed because the value is absent, never set to 0, and never carried forward from another period. Every value carries one quality status: observed-public (we saw it ourselves on a public page, dated), observed (ours, fully documented), not-measured (access is pending, so the value is not printed as a number) or not-obtainable (no source exists for it today). A snapshot metric is a single-day observation with no monthly value — so its “previous period” cell carries an earlier observation date, not a month’s figure.
Data sources used, and the date each was obtained
The invented report’s data sources, what each was used for, and the date each was obtained.| Source | Used for | Date obtained |
|---|
| The business’s public listing page (the map product’s profile) | §2 listing.field_state, review.*; §3 | 2026-10-26 (and earlier: 2026-10-16, 2026-09-16) |
| The business’s own website, public pages only, logged out | §2 listing.field_state (comparison), §3; evidence URLs | 2026-10-26 (and earlier: 2026-10-09, 2026-10-16) |
AI-assisted search assistant, one sitting, logged out (placeholder tool id AITool-1 (DEMO)) | §3 only | 2026-10-23 |
| BravData’s own task and publication records | §1, §2 work.*, site.published_items | records closed 2026-10-31 |
| The free visibility check of 2026-09-16 / 2026-09-17 (our own worksheet) | the previous-observation values in §2 and §3 | obtained 2026-09-16 and 2026-09-17 |
Sources not obtained, and when we last asked
The sources the invented report does not have, when each was last asked for, and what its absence means for the numbers.| Source we do not have | Last asked | Consequence |
|---|
| Listing platform performance report (impressions, views, actions) | asked 2026-10-13; reminder 2026-10-27 | listing.impressions, listing.views_maps, listing.views_search, listing.actions.* are not reported (§5, item 1) |
| Website analytics (sessions) and form/CRM records (submissions) | asked 2026-10-13; reminder 2026-10-27 | site.sessions, site.enquiry_form_submissions are not reported (§5, item 2) |
| Inquiry log or practice-management aggregates (Growth measurement set) | not asked — outside the AI Visibility plan, and none supplied | §4 is empty this period; inquiry.*, appointment.*, spend.amount, derived.* are not reported (§5, item 4) |
What we changed this period
Two tasks were completed, both answering findings from the free visibility check of 2026-09-16/17 — the check’s next steps were written to be doable by the business alone; the business chose to have us do them. Both were approved by the client before we started them.
The two invented tasks completed in the period: what changed, who approved them, when they were delivered, the work estimate and the evidence.| Task | What changed | Approved by / on | Delivered | Work estimate | Evidence |
|---|
| T-101 | Home page first screen rewritten: “family medicine”, the town, the listed hours and a tap-to-call number now appear before the first scroll. No claim about patients, results or rankings was added. | R. Placeholder (practice manager) / 2026-10-05 | 2026-10-09 (one revision round used) | 2.0 staff hours (one revision round included) | Live URL, verified by us 2026-10-09 and re-verified 2026-10-26; publication log P-101 |
| T-102 | Saturday hours reconciled and the listing’s services field filled. We drafted the exact field values; the business applied them in its own listing manager because we hold no access (§5, item 1). The business confirmed its site’s Saturday hours were the correct ones; that confirmation is reported by the client, while the field values are observed by us. | R. Placeholder (practice manager) / 2026-10-13 | applied 2026-10-16; re-observed 2026-10-16 and 2026-10-26 | 1.0 staff hour | Field values observed on the public listing page 2026-10-16 and 2026-10-26; publication log P-102 |
Coverage, reconciled.
Both tasks are implementation work the plan’s published monthly scope covers. The earlier implementation-credit model is retired, so no credit count appears in this report. Nothing off the price list was auto-charged, and no task ran without the client’s approval.
What we did not do.
The check’s third finding — the “Book online” link on the Services page failing to load — is not in this list: it is a change to the business’s own booking link, and it needs the business’s decision. We did not fix it ourselves and will not describe it as fixed.
The term this section used to quote:
Earlier reports quoted the implementation-credit term word for word. That model is retired, so the quote is gone and the work above is what the plan’s published monthly scope covers.
The numbers — eight rows, eight different states
Eight invented metric rows, each in a different state: a real zero, a rating not shown, a snapshot, not-reported values with reasons, and one metric that is not obtainable at all.| # | Metric key | This period | Source (date obtained) | Quality status |
|---|
| 1 | review.count_total | 0 — a real zero: the profile displayed no review text and the reviews area read “No reviews” (snapshot observed 2026-10-26). Zero here is counted, not missing. | Public listing/review profile | observed-public |
| 2 | review.rating_displayed | Not shown on the profile on 2026-10-26 — no rating is displayed, because no reviews are published. This is a third state: not zero, not missing. | Public listing/review profile | observed-public |
| 3 | listing.field_state | 9 fields recorded on 2026-10-26, the hours field read in its three parts; field values, agreement with the website and the two fields that changed are in §3. Snapshot on that date, not a monthly figure. | Public listing page + the business’s own site, both read logged out | observed-public |
| 4 | review.new_count | 0 — a real zero: no review carried a published date inside 2026-10-01..10-31. Counted, not missing. | Public review profile | observed-public |
| 5 | review.response_count | Not reported — with no reviews published in the period there is no population to count, and a rate over a zero denominator is not a number (§5, item 3). Printing 0 would read as “you responded to none of your reviews”, which is a different and false statement. | Public review profile | observed-public |
| 6 | listing.impressions | Not reported — nothing has been estimated and nothing from another period has been substituted. Reason: no listing performance access and no export (§5, item 1). | Listing platform performance report — not obtained (asked 2026-10-13, reminder 2026-10-27) | not-measured |
| 7 | site.sessions | Not reported — nothing has been estimated and nothing from another period has been substituted. Reason: no analytics access and no export (§5, item 2). | The business’s analytics — not obtained (asked 2026-10-13, reminder 2026-10-27) | not-measured |
| 8 | inquiry.count + the Growth set | Not reported — this is a AI Visibility client and no aggregate log or export was supplied (§5, item 4). call.tracked_count is not obtainable at all: no source exists unless the business already pays for call tracking, and such subscriptions are excluded from both plans. | Would be the business’s own log or vendor export — not obtained, and not in this plan | not-measured |
The report’s own rule, stated once: a metric with no source prints as a blank with a reason — and a metric whose source shows nothing prints as the value the source showed. Both sentences are the method working, not a gap in the work.
Visibility and reputation observations
Listing field snapshot — observed 2026-10-26, logged out, from Anytown, TX. The same nine fields every check reads. Two of them changed since the free check, shown as an observed-on-date pair, never as a before/after claim:
The invented listing’s field values observed 2026-10-26 and their agreement with the business’s website read the same day.| Field | Value observed 2026-10-26 | Agreement with the website (read 2026-10-26) |
|---|
| Name · Address · Phone · Website link | as published | match (phone formatting differs) |
| Category | “Family practice” | mismatch in wording only — the site says “family medicine clinic”; same meaning |
| Hours — weekdays | Monday–Friday, as published | match |
| Hours — Saturday | “Saturday 9:00 a.m.–1:00 p.m.” | now matches — on 2026-09-16 the listing showed Saturday as “Closed” while the site published 9:00 a.m.–1:00 p.m.; the business confirmed the site was correct (reported by the client) and applied the listing change on 2026-10-16 (observed by us 2026-10-16 and 2026-10-26) |
| Holiday hours | not shown on the listing or the site on 2026-10-26 | cannot be reported as a match or a mismatch; see §5, item 6 |
| Services | annual physicals; same-day sick visits; chronic-condition follow-up | now matches — the listing’s services field was empty on 2026-09-16; filled by the business on 2026-10-16 (observed by us) |
| Photo count / most recent photo date | 3 photos; most recent shown as 2026-03-11 | n/a (the site carries no photo count) |
Confidence: high — each value was read from the page and is quoted with its date, and two values were reproduced on two separate dates. What this does not show: which of two disagreeing values is correct (that is the business’s call), and whether anyone read these fields.
Reviews.
Review count displayed: 0 — a real zero, counted. Rating displayed: no rating displayed. New reviews in the period: zero new reviews — a real zero, and not the same state as missing. Reviews carrying a response: not reported (§5, item 3). Observed themes: none — with no published review text there is nothing to read, so no theme was recorded and none was inferred; no sentiment score and no sentiment index exists in this method.
What this does not show: nothing about the business’s care, and nothing about how many people looked. A profile with no reviews is not evidence that nobody is happy; it is evidence of an empty reviews area on the date observed.
AI-assisted search observations — one sample set, 2026-10-23.
Five questions, the frozen set, asked once in one sitting while logged out. This is the only sample set for October. The questions are recorded with the answers, the tool and the time. Nothing here is scored, ranked, averaged or turned into a percentage, and nothing here is compared with any other month. Tool id: AITool-1 (DEMO — a placeholder id; in a real sample the tool’s own name is recorded, so that no invented answer is attributed to a real product).
- “Who are good family medicine doctors in Anytown?” — asked 2026-10-23 14:02 CDT. Answer, verbatim: “I don’t have a ranking for that. For family medicine in Anytown, check who’s accepting new patients, the opening hours, and how appointments are booked.”
- “I’m looking for a family medicine practice near Anytown that takes Medicare.” — asked 14:06. “I can’t confirm which local practices take Medicare — ask the practice directly and confirm before you book.”
- “What should I know before choosing a family medicine practice in Anytown?” — asked 14:11. “Ask about same-day availability, who you’d actually see, and what happens if you need care outside opening hours.”
- “Anystreet Family Medicine — what can you tell me about it?” — asked 14:15. “I don’t have reliable information about that practice.”
- “Where can I find same-day sick visits in Anytown?” — asked 14:19. “I can’t confirm which Anytown practices offer same-day sick visits. Directory listings [two directory sites named by the tool — recorded in the internal sample log; not reproduced here, because this report names no other practice and measures none] may help.”
What this does not show: whether any patient asks these questions, whether the answers resemble what a patient sees on their own device, or what the same tool would say tomorrow. Quoted, dated, and nothing more.
Inquiry and access — empty, and this is why
This section is empty this period, and this is why. It applies to Growth, or to an AI Visibility client that supplies its own inquiry data. This client is on AI Visibility and supplied no inquiry or appointment aggregates for the period, so there is nothing to show. We have not filled the gap with a zero, an estimate or another period’s figure.
What would put numbers here, and who unblocks it: an aggregate monthly count of inquiries by channel and by the source the patient stated (a spreadsheet is enough — no patient-level file, ever), and, where the business’s systems can produce them, aggregate bookings, no-shows and cancellations.
What we could not measure this period, and why
Six short blocks, each with the metric, the reason, whose action unblocks it and what happens if it stays blocked. Two of the six carry no action at all — the honest case:
- Item 1 — listed impressions, views and actions. Not measurable yet: we do not have listing manager access and no performance export was received (asked 2026-10-13, reminder 2026-10-27). Until it is granted this metric is not measured, not zero. Action on your side: invite us as a manager on the business’s listing, or export the platform’s performance report and send it to [email protected]. If it stays blocked: these rows remain blank next month, and the report will not substitute an earlier period or an estimate.
- Item 2 — website sessions and form submissions. Not received by 2026-11-06. Nothing has been estimated and nothing from another period has been substituted. Status: awaiting your data. Action on your side: grant read access to the analytics property, or have the web vendor export October sessions and form-submission counts as aggregates. If it stays blocked: we cannot show whether the first-screen change produced anything measurable, and we will say that in the next report rather than imply an effect.
- Item 3 — the review response count and its rate. Not computed, and it is not a missing-data problem: with a real zero reviews published in the period there is no population to count, and a rate over a zero denominator is not a number. Action on your side: none — this resolves itself the first month the business receives a review. If it stays blocked: the row keeps this wording; it will not become 0 or 100%.
- Item 4 — the Growth measurement set. No source exists in this plan and none was supplied.
call.tracked_count is not obtainable unless the business already pays for call tracking; such subscriptions are excluded from both plans. Action on your side: if the business already logs inquiries, an aggregate monthly count by channel and by stated source is enough — aggregate counts only, no patient-level file, ever. If it stays blocked: the section stays empty, the derived figures stay uncomputed, and no figure is estimated in their place. - Item 5 — any prior-period figure for the period-based metrics. There is no September value, because September holds only a single-day snapshot taken during the free check and a handful of page reads. A single day is not a month. Action on your side: none — this closes by itself as complete months accumulate. If it stays blocked: the previous-period column keeps saying so; it will never be back-filled with a plausible number.
- Item 6 — holiday hours. Not shown on either property, so we cannot report a match or a mismatch. Action on your side: decide a holiday-hours policy and publish it, or state that there is none. If it stays blocked: the field stays “not shown on the profile on [date]” and no assumption is made about whether the business is open.
Standing gates, unchanged, with their dates:
Listing API access for a BravData-owned project — standing. Analytics access or a vendor aggregate export — standing. Practice-management vendor export of aggregates — standing. Any comparison to another business, any benchmark, any local-pack position — not obtainable: no source provides this today, and we are not going to estimate it. What search will show next month — structural: nobody knows, and we will not pretend to.
What we recommend next (all three)
Three items, each with the evidence it rests on. The list is never padded to a quota, and anything we could not evidence was dropped.
The invented report’s three recommendations with the evidence each rests on, its work estimate and whether it needs the client’s approval.| # | Recommendation | Evidence it rests on | Work estimate | Needs your approval? |
|---|
| 1 | Fix or remove the “Book online” link on the Services page, and have the business’s web person confirm which booking route is real. | The free check observed the link opening a page reading “404 — Page not found” on 2026-09-17, reproduced 2026-09-18, and the link was still present and still failing when we re-read the page on 2026-10-26. | No work for us — the change is the business’s (or its booking vendor’s). We will re-test it and record the result in the next report’s §3 at no extra charge. | Yes — the change is yours to make |
| 2 | Publish one page on “same-day sick visits” that states the business’s own hours, how a patient asks for a same-day visit, and what the published response expectation is. | The check’s discoverability pass; and the AI sample of 2026-10-23 answered that it could not confirm which Anytown practices offer same-day visits. | Up to 2 staff hours (including one revision) | Yes — publish only after you approve the copy |
| 3 | Start logging inquiries as aggregate counts by channel and by stated source, so the inquiry section can exist and be measured. | That section is empty this period and §5, item 4 explains why; the dictionary holds no inquiry metric we can obtain without a client log. | Up to 1 staff hour buys our work only: a one-page logging template and one review of the first week’s entries. The logging itself is yours at your front desk, and none of your time is charged. | Yes — the logging itself happens at your front desk |
Work estimate for the next cycle:
Recommendations 2 and 3 together come to up to 3 staff hours; recommendation 1 needs no work from us. Nothing is committed until you approve it on the call.
Decisions needed from you (all five)
The five invented decisions the report asks for, with the owner, the date and what happens if the decision waits.| Decision | Owner | By when | What happens if it waits |
|---|
| 1. Fix or remove the “Book online” link, and tell us which booking route is real | D. Placeholder, practice owner (DEMO) | 2026-11-13 | The link keeps failing for anyone who tries it; the row stays a live defect and we will keep reporting it as observed |
| 2. Grant listing manager access or send the October performance export | R. Placeholder, practice manager (DEMO) | 2026-11-13 | The impressions, views and actions rows stay blank next month; we will not substitute an estimate, and we will not ask again without a decision |
| 3. Grant analytics access or have the web vendor export October sessions and form submissions | R. Placeholder (DEMO), with the web vendor | 2026-11-13 | We cannot say whether the first-screen change produced anything measurable, and the next report will say exactly that |
| 4. Decide a holiday-hours policy and publish it, or state that there is none | D. Placeholder, practice owner (DEMO) | 2026-11-13 | The holiday field stays undeterminable and every future check keeps carrying that limit |
| 5. Approve the next cycle’s task list (recommendations 2 and 3 = up to 3 staff hours of work in November) | R. Placeholder, practice manager (DEMO) | on the call, 2026-11-10 | Nothing starts; unapproved work is never begun, and nothing carries over from one month to the next unless we agree otherwise in writing |
Closing note on the second example
No patient-level, clinical, personal or identifying data appears anywhere in this report, its tables or its appendix — the most granular thing it holds is an aggregate count for one location for one month (and this month, mostly blanks). A blank is not a zero, and a real zero is not a blank: the report tells you which one you are looking at, every time.