Discover all customers of a given company by scanning websites, case studies, review sites, press, social media, job postings, and more. Use when you need competitive intelligence on who a company sells to.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Find all customers of a company by scanning multiple public data sources. Produces a deduplicated report with confidence scoring.
Find all customers of Datadog
Who are Notion's customers? Use deep mode.
| Input | Required | Default | Description |
|---|---|---|---|
| Company name | Yes | — | The company to research |
| Website URL | No | Auto-detected | The company's website URL |
| Depth | No | standard | quick, standard, or deep |
Ask the user for:
mkdir -p customer-discovery-[company-slug]
Collect all results into a running list. For each customer found, record:
1. Website logo wall
Run the scrape_website_logos.py script:
python3 skills/capabilities/customer-discovery/scripts/scrape_website_logos.py \
--url "[company-url]" --output json
Parse the JSON output and add each result to the customer list.
2. Case studies page
Use WebFetch on the company's case studies page (try /case-studies, /customers, /resources/case-studies). Extract customer names from page headings and content.
3. G2/Capterra reviews
If the review-site-scraper skill is available, use it to find reviewer companies:
python3 skills/capabilities/review-site-scraper/scripts/scrape_reviews.py \
--platform g2 --url "[g2-product-url]" --max-reviews 50 --output json
First, WebSearch for the company's G2 page: site:g2.com "[company]". Extract reviewer company names from review author info.
4. Web search for press
WebSearch these queries and extract customer mentions from results:
"[company]" customer OR "case study" OR partnership"[company]" "we use" OR "switched to" OR "chose"5. Company blog posts
WebSearch: site:[company-domain] customer OR "case study" OR partnership OR "customer story"
6. Wayback Machine logos
Run the scrape_wayback_logos.py script:
python3 skills/capabilities/customer-discovery/scripts/scrape_wayback_logos.py \
--url "[company-url]" --output json
Logos marked still_present: false are especially interesting — they indicate former customers.
7. Founder/exec LinkedIn posts
WebSearch: site:linkedin.com "[company]" customer OR "excited to announce" OR "welcome"
8. Twitter/X mentions
WebSearch: site:twitter.com "[company]" "we use" OR "just switched to" OR "loving"
9. Reddit/HN mentions
WebSearch these queries:
site:reddit.com "we use [company]" OR "[company] customer"site:news.ycombinator.com "[company]" customer OR user10. Job postings
WebSearch: "experience with [company]" site:linkedin.com/jobs OR site:greenhouse.io OR site:lever.co
Companies requiring experience with the product are likely customers.
11. YouTube testimonials
WebSearch: site:youtube.com "[company]" customer OR testimonial OR review
12. SEC filings
WebSearch: site:sec.gov "[company]" — Look for mentions in 10-K and 10-Q filings.
13. Podcast transcripts
WebSearch: "[company]" podcast customer OR transcript OR interview
14. GitHub usage signals
WebSearch: site:github.com "[company-package-name]" in dependency files, package.json, requirements.txt, etc.
15. Integration directories
WebFetch marketplace pages where the company lists integrations:
16. BuiltWith detection
python3 skills/capabilities/customer-discovery/scripts/search_builtwith.py \
--technology "[company-slug]" --max-results 50 --output json
17. Crunchbase
WebSearch: site:crunchbase.com "[company]" customers OR partners
Merge results by company name using fuzzy matching:
Apply these rules:
High confidence:
Medium confidence:
Low confidence:
Create two output files:
customer-discovery-[company]/report.md:
# Customer Discovery: [Company Name]
**Date:** YYYY-MM-DD
**Depth:** quick | standard | deep
**Total customers found:** N
## High Confidence (N)
| Customer | Source | Evidence |
|----------|--------|----------|
| Shopify | Case study | [link] |
| ... | ... | ... |
## Medium Confidence (N)
| Customer | Source | Evidence |
|----------|--------|----------|
| ... | ... | ... |
## Low Confidence (N)
| Customer | Source | Evidence |
|----------|--------|----------|
| ... | ... | ... |
## Sources Scanned
- Website logo wall: [url] — N customers found
- G2 reviews: N reviews analyzed — N companies identified
- Wayback Machine: N snapshots checked — N logos found (N removed)
- Web search: N queries — N mentions
- ...
## Methodology
This report was generated using the customer-discovery skill, which scans
public data sources to identify companies that use [Company Name]. Confidence
levels reflect the strength and directness of the evidence found.
customer-discovery-[company]/customers.csv:
CSV with columns: company_name,confidence,source_type,evidence_url,notes
Write the CSV using a code block or Python script.
| Script | Purpose | Key flags |
|---|---|---|
scrape_website_logos.py | Extract logos from current website | --url, --output json|summary |
scrape_wayback_logos.py | Find historical logos via Wayback Machine | --url, --paths, --output json|summary |
search_builtwith.py | BuiltWith technology detection (deep mode) | --technology, --max-results, --output json|summary |
All scripts require requests: pip3 install requests
External skill scripts (use if available):
skills/capabilities/review-site-scraper/scripts/scrape_reviews.py — G2/Capterra/Trustpilot reviews (requires Apify token)skills/capabilities/linkedin-post-research/scripts/search_posts.py — LinkedIn post search (requires Apify token)--api-key flag); free scraping is used by default.まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Research supported advertising promises from a specific buyer's context, current product facts, customer language and observed ads. Turn them into static, video or copy ideas; match video ideas to actual recipes and save a product-scoped angle bank for write-video-ad-script. Product-led discovery can use verified facts without reviews. Reference persistence and reach are observations, not proof of performance.
日本語の概要は準備中です。原文の説明を表示しています。
Analyze ad campaign performance data (Google, Meta, LinkedIn) to identify what's working, what's wasting budget, and specific cut/scale/test recommendations. Runs statistical analysis, funnel diagnostics, and multi-channel budget reallocation with specific dollar-amount shift recommendations and scenario modeling.
日本語の概要は準備中です。原文の説明を表示しています。
For paid lead-gen and participant-recruitment ads, replaces vanity CPA with true CAC per qualified lead by joining ad-platform data with downstream funnel events, surfaces tracking gaps, and classifies every creative into Scale / Keep / Investigate / Cut.
日本語の概要は準備中です。原文の説明を表示しています。
The intake stage of a Meta ad harness. Researches the brand first, states what it assumes, and asks only what it cannot infer and what would change money, claims or legal exposure if guessed wrong — at most four questions a turn, in plain business language. Writes the brand's stable facts (ads/brand.md), the campaign brief, and a log of what was asked vs assumed. Use it when someone wants to run Meta ads and the brand's ad facts are not yet recorded.
日本語の概要は準備中です。原文の説明を表示しています。
Analyze the message match between your ads and landing pages. Checks if the promise in the ad copy carries through to the landing page headline, body, and CTA. Flags disconnects that kill conversion rates. Works with Google, Meta, and LinkedIn ads.
日本語の概要は準備中です。原文の説明を表示しています。
Turn a brand's facts, a campaign brief and the product into a one-minute Meta ad strategy with its reasoning — goal, audience, objective and placements, budget split, how many ads and how different they must be (generate 3x so the user can choose), the test frame, the success metric, and safety rules for the user to accept. Decides what launch-meta-ad-campaign then executes, and re-reads the strategy after a deep check to show which assumption changed. Use it after intake and before any creative is generated or any campaign is built.
日本語の概要は準備中です。原文の説明を表示しています。