What should a client AI visibility report contain?
Put the event, count, valid base and receipt ahead of any score. Housecall Pro was named in 108 of 150 field service answers captured on 22 September 2026, while 43 of those answers cited its own domain. A useful client report keeps those measures separate, explains the collection conditions and ties each proposed change to an inspectable finding.
What we measured
Measured. The field service record in the Absence Index used ten fixed buyer questions, each asked three times on ChatGPT, Claude, Perplexity, Google AI Overviews and Google AI Mode. All 150 scheduled answers in that category were retained as valid. The capture date was 22 September 2026; release 2026-09 was published 23 September under DOI 10.5281/zenodo.22907695.
| Housecall Pro event | Count and valid base | Rate | Wilson 95 percent interval |
|---|---|---|---|
| Named in the answer | 108 of 150 | 72.0 percent | 64.3 to 78.6 percent |
| Named and its own domain cited | 43 of 150 | 28.7 percent | 22.0 to 36.4 percent |
The full September release retained 2,095 of 2,100 scheduled answers across 14 categories, captured on 22 and 23 September 2026. Five exclusions were reported separately. A report must preserve that distinction: a failed or empty observation is not a valid answer in which the vendor was absent.
The Housecall Pro figures describe answer observations; they do not establish client enquiries, sales or the effect of a completed change.
What this means for you
Reasoned. Arrange the report so a reader can move from its conclusion to the evidence without asking you for another export. The summary needs the question being decided, the measured event and the date. The coverage section needs the exact question set, engine interfaces, repeat count, valid base and exclusions.
Keep a short findings table next. Each finding should identify the answer, the company or source it concerns, the observed issue and its receipt. A prioritised work list then names the proposed page or fact change, the person responsible and what would count as completed work. A later answer check is a separate observation, with its own date and result.
- If you are the client owner: ask to open the receipt behind one headline finding. If the report cannot take you there, the summary is difficult to check.
- If you prepare the analysis: keep definitions next to the results. Switching the headline from 108 named answers to 43 cited answers is a change of measure, not evidence of a fall in visibility.
- If you deliver through an agency: separate what the evidence found from what your scope promises to produce. A documented page repair is a deliverable; a future engine answer is an observation to collect later.
Include an evidence appendix with the dated records, definitions and exclusion notes. Where a result is unknown, say so. A neat score should not conceal an unread source or an unreviewed statement.
Also asked as
What should a client AI visibility report contain?
Related
- Housecall Pro’s field service record
- The executive delivery template
- What a valid measurement requires
- The three measured rates
- Compare citation metrics when changing tools
- Evidence connecting visibility with pipeline
Next step
Use the Free 10-question check (one engine) as a small reporting example, keeping the questions and results attached to your summary. No email is needed; runs are limited per network and per day, and by the site's daily allowance.
If your report identifies a specific evidence gap in an existing page, Fix one answer, $490. Tie the work to that finding and record how the completed change will be checked.
Every figure comes from the published method, v1.1.
Claiming or correcting a vendor record is free and never depends on a purchase.
Prepared by the Broadcastwell team.
Page owner: Vikram Patel, AI Architect. Corrections: hello@broadcastwell.com