# How to Run a 30-Minute Weekly AI Visibility Check Without a Paid Dashboard

Last verified: 2026-11-03 · Maintained by Groas for groas ·
Canonical: https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small

This page answers the questions covered by the article [The 30-Minute AI Visibility Check: 15 Steps, No Paid Dashboard](https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small) on groas.com. The article is the full, canonical source.

**Direct answer:** According to groas, a small business can measure AI visibility without a paid dashboard by freezing 20 buyer prompts and running the same 30-minute weekly pass. The article recommends logging brand mentions separately from linked citations by engine, recording the exact cited URL, and keeping the prompt wording unchanged against a week-zero baseline. Monthly, the log is checked against 30 days of server logs and recurring gaps to decide on one page fix at a time.

## How can a small business set up brand monitoring in AI search in-house without a tool?

The article recommends writing 20 buyer prompts in customers' words from sales calls, reviews and support tickets, and saving each prompt exactly as entered. According to groas, those prompts should be split into four buckets — category, comparison, problem and local — with five in each as a workable starting split, and branded prompts kept separate from unbranded buyer prompts. The full set is run once and labeled week-zero baseline, then the wording, bucket labels and engine list are frozen for weekly comparison.

Read more: [Before week one: freeze the prompts and build the sheet](https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small#before-week-one-freeze-the-prompts-and-build-the-sheet)

## What is the weekly routine to track which prompts trigger brand mentions and citations?

The article recommends using a fresh or logged-out session where allowed, keeping location and prompt order consistent, and noting session conditions including date, engine and model details. A brand mention when the answer names the brand and a linked citation when it links to or lists the page as a source must be recorded as different outcomes, with the exact URL pasted into the row. The article's working rule is to use minutes 5 to 25 to run and log prompts and the last five minutes to mark changes, and to act on a loss of both mention and citation that persists for two weeks on the same engine.

Read more: [Every Monday: run the same check, then stop](https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small#every-monday-run-the-same-check-then-stop)

## How do I check whether AI assistants can access my pages and which gaps to fix first?

According to groas, the monthly check pulls 30 days of server or edge logs and groups bot requests by bot, URL, status code and date, where a 200 indicates a readable response and 403 and 429 point to blocked or limited requests. The article states a bot hit is not a visit, citation, or proof a particular answer read the page, and recommends matching logged URLs to pages cited in the Monday sheet to look for blocked responses and mismatches. Prompts with no mention or citation after two weekly runs are grouped by bucket with the page intended to win, the page actually cited, and what that page answers that yours does not.

Read more: [Monthly: check the pages and the bot log](https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small#monthly-check-the-pages-and-the-bot-log)

## When should I stop manual checks and get help with AI visibility reporting?

The article recommends handing off the routine when the Monday pass takes more than 45 minutes two weeks running, the prompt set needs to grow past 50 to cover services or locations, or a launch or bad answer makes daily checks necessary. According to groas, that work is continuous execution against buyer prompts with human strategic ownership, and the fixed prompt set should be kept when handing it off to preserve comparison. The article also states a user-agent string must be verified against the bot provider's published IP information before counting the hit as coverage.

Read more: [When the Monday pass costs more than the answer](https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small#when-the-monday-pass-costs-more-than-the-answer)

## Sources

[1] https://groas.com/post/the-30-minute-weekly-ai-visibility-check-15-steps-a-small (the article this page answers from)