AI Search Visibility Tracker

Four workflows covering the full AI visibility lifecycle: discover which prompts matter → find citation opportunities → audit your site → track over time.

All workflows use apify/google-search-scraper for AI search. Workflow C also uses apify/website-content-crawler.

Recommended flow: Run Workflow A to discover prompts → Workflow B to find citation opportunities → Workflow C to audit your site → Workflow D to track everything on a schedule.

Workflow A — Competitor Prompt Discovery

Goal: Find which queries surface a competitor in AI search answers, so you know which prompts are worth monitoring.

Inputs to collect

Workflow

1. Generate 15–30 candidate queries from seed keywords using these templates:

   • best [topic], [topic] tools, how to [topic], [topic] for [use case]
   • [topic] vs [competitor brand], [competitor brand] alternative
   • [topic] API, [topic] pricing, [topic] tutorial

2. Run apify/google-search-scraper for each candidate query. For each result, extract:

   • aiOverview.sources[], aiMode.sources[], chatGptAnswer.sources[], perplexityAnswer.sources[], copilotAnswer.sources[], geminiAnswer.sources[]
   • Also check answer_text / aiOverview.text for competitor brand name mentions (word-boundary match: \bBrand\b)

3. For each (query, source) pair where the competitor domain or brand appears: record a hit.

4. Output a prompt-major table sorted by total hit count descending:

| Query | ChatGPT | Perplexity | AI Overviews | AI Mode | Copilot | Gemini | Total |
|-------|---------|------------|--------------|---------|---------|--------|-------|
| "best web scraping API" | ✓ | ✓ | — | ✓ | — | ✓ | 4 |
| "how to scrape Google" | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | 6 |

5. Deliver the top-N queries (default 10) as a ready-to-paste list for Workflow D's config.json prompts.

Workflow B — Citation Opportunity Finder

Goal: For a target topic, identify which domains and content types AI engines most often cite — revealing where to publish or pitch content.

Inputs to collect

Workflow

1. Run apify/google-search-scraper for each seed query across selected AI sources.

2. Collect every URL from sources[] across all results. Normalise to registrable domain (blog.example.com → example.com).

3. Aggregate:

   • By domain: count citations, list which AI sources cite it, list which queries triggered it
   • By content type: infer from URL path patterns (docs → /docs/, /reference/; blog → /blog/; news → known news domains)

4. Rank by total citation count. Output:

Top-cited domains for "web scraping" (42 queries × 6 sources):
| Domain | Citations | AI Sources | Inferred type |
|--------|-----------|------------|--------------|
| docs.apify.com | 38 | ChatGPT, Perplexity, AI Mode | Documentation |
| scraperapi.com/blog | 21 | AI Overviews, Gemini | Long-form blog |

5. If deep-crawl enabled: run apify/website-content-crawler on the top-3 cited URLs per domain. From the markdown output, extract:

   • First heading that directly answers the query
   • Presence of code blocks in first 500 words
   • Word count
   • Whether an H2/H3 contains the exact query phrase

6. Summarise patterns: "AI engines in this topic prefer [long-form docs / short direct-answer posts]. Typical cited page: [word count range], [has/lacks direct-answer H2], [has/lacks code example above the fold]."

Workflow C — GEO Website Audit

Goal: Check whether a specific website's content is structured for AI citation; compare it against what AI engines actually cite for your target prompts.

Inputs to collect

Workflow

1. Run apify/google-search-scraper for each target prompt. For each (prompt × source) record whether your registrable domain appears in sources[].

2. For prompts where your domain is not cited: identify the top-cited competitor URL for that prompt.

3. Run apify/website-content-crawler on:

   • Your most relevant page(s) for each un-cited prompt
   • The top-cited competitor page for each un-cited prompt

4. For each un-cited prompt, produce a gap card:

Prompt: "how to scrape Google search results"
Your page: apify.com/blog/scraping-google  →  NOT cited on ChatGPT, Perplexity, AI Mode
Top-cited: docs.brightdata.com/scraping/google (cited 5/6 sources)

Structural gaps:
  ✗ Your page: answer buried after 900 words, no direct-answer H2
  ✓ Competitor: H2 "How to scrape Google in 3 steps" at word 120 + code block at word 180

Recommended actions (priority order):
  1. Add H2 that mirrors the query phrase within first 300 words
  2. Move code example above the fold
  3. Add a "Quick answer" summary box at the top

5. Deliver: per-prompt gap cards + a consolidated action table ranked by expected impact.

Workflow D — Recurring Visibility Tracker

Goal: Snapshot brand citations and mentions across all six AI surfaces on a recurring schedule and track changes over time.

Prerequisites

(No need to check upfront)

• APIFY_TOKEN saved in a .env file next to config.json (the runner auto-loads it).
• Python 3.9+ on PATH. pip3 install requests (only third-party dependency); pip3 install tldextract recommended for accurate registrable-domain matching on multi-part TLDs.
• For automated daily runs: macOS / Linux with launchd or cron available (the installer handles both). Windows users get printed schtasks instructions.

Steps

Copy this checklist and track progress:

Task Progress:
- [ ] Step 1: Load or collect the seven required inputs
- [ ] Step 2: Confirm AI sources and cadence
- [ ] Step 3: Write config.json + .env, then install the OS schedule
- [ ] Step 4: Run a snapshot now so the user sees the first report
- [ ] Step 5: Deliver the history report (diff vs. all prior runs)

Step 1: Load or Collect the Seven Required Inputs

If config.json exists in the user's working directory, load it and skip to Step 4 unless the user asks to reconfigure. On first run, ask all seven anchors before any Actor call:

After those seven, ask optional follow-ups: countryCode, languageCode, location (UULE), preferred run hour (default 09:00 local).

Then one verbosity question -- save as config.json:include_full_answers:

• on_demand (default): report shows short quoted snippets around each surface-form match. Full LLM answers live in the named KV store; user can ask later.
• always: report embeds the full LLM answer verbatim whenever any entity is mentioned. Useful for one prompt; gets unwieldy at 5+ prompts.

Step 2: Confirm AI Sources and Cadence

Echo back the user's seven choices in a single paragraph for confirmation. If the user toggles sources, update the in-memory config before writing.

Step 3: Write config.json + .env, Then Install the OS Schedule

Create the working directory layout next to where the user wants reports to land:

working-dir/
  config.json     # copied from the skill's config.example.json, edited with collected values
  .env            # APIFY_TOKEN=apify_api_xxx   (chmod 600)

cp ${CLAUDE_PLUGIN_ROOT}/reference/scripts/config.example.json ./config.json
# then edit with the collected values, save

echo 'APIFY_TOKEN=your_token_here' > ./.env
chmod 600 ./.env

Then install the OS schedule:

bash ${CLAUDE_PLUGIN_ROOT}/reference/scripts/install_cron.sh --cadence daily --hour 9

Cron expression mapping:

Step 4: Run a Snapshot Now

macOS:

launchctl kickstart "gui/$(id -u)/com.apify.ai-visibility-tracker"
tail -f ~/Library/Logs/ai-visibility-tracker.log

Linux / generic:

python3 ${CLAUDE_PLUGIN_ROOT}/reference/scripts/run_snapshot.py --config ./config.json

Both paths:

1. Call apify/google-search-scraper with the configured prompts + AI-source toggles.
2. Parse each (prompt x source) cell for citations and brand/competitor mentions.
3. Append rows to the named Apify Dataset (schema in reference/output-schema.md).
4. Write the raw item to the named KV store (long-term archive).
5. Compute the history vs. all prior runs.
6. Write reports/snapshot-<ISO-date>.md next to config.json.

Step 5: Deliver the History Report

Open reports/snapshot-<ISO-date>.md and surface the top findings in chat. Lead with:

• First-ever citations / mentions today -- entity x source combinations crossing the threshold for the first time.
• Drops -- entity was cited in the latest prior run but isn't today.
• For every cited entity, the exact matched URL(s).

Output format

Snapshot summaries are entity-major: one Markdown table per tracked entity (brand, then competitors), one row per AI source, columns Source | Cited | Mentioned | SoV% | Matched URLs | History.

Actors

Pricing changes; check the pricing tab before quoting numbers. Disable unused AI sources to reduce cost.

Telemetry. Workflow D runs through reference/scripts/run_snapshot.py, which already sets User-Agent: apify-awesome-skills/ai-search-visibility-tracker-... on every Apify API call. When calling these Actors ad-hoc in Workflows A-C, pass the matching flags so usage is attributed: --user-agent apify-awesome-skills/apify-ai-search-visibility-tracker and --json (use --format json for apify datasets get-items), and append 2>/dev/null. Example:

apify actors call apify/google-search-scraper \
  --user-agent apify-awesome-skills/apify-ai-search-visibility-tracker \
  --json 2>/dev/null

Quality Rules

• Non-interactive. No stdin reads in run_snapshot.py -- launchd / cron has no stdin.
• Word-boundary brand matching (\bbrand\b, case-insensitive). See reference/citation-matching.md.
• Registrable-domain citation matching (blog.apify.com counts as apify.com). See reference/citation-matching.md.
• Never skip a row. If an AI source returns nothing, write a row with cited: false, mentioned: false, answer_text: "[no answer returned]".
• Every row carries the Apify run ID so any finding can be reverified.

Error Handling

APIFY_TOKEN not found -- Tell the user to put it in .env next to config.json (echo 'APIFY_TOKEN=...' > .env && chmod 600 .env). Token at https://console.apify.com/account/integrations. config.json not found -- Run Step 3 first to create it from the template. Dataset name not set -- Ask the user for a name; the runner will create the dataset on first append. Actor run FAILED -- Print the console link from the runner output and ask the user to inspect it. AI source returned no answer -- The row is still written with [no answer returned]. Not an error. website-content-crawler returns no markdown -- Page may be JS-heavy; try with useBrowserCrawler: true. Schedule not firing -- See reference/scheduling.md troubleshooting section. No previous run to diff against -- First run only. The report renders the snapshot without a history section.