Mine case-study pages for named customers and results
Turns any customers or case-study page into structured proof data: up to three named customers with their claimed results, plus a visible-customer count. Use it to personalize with 'you already serve teams like X', to map a competitor's install base, or to score accounts by proof maturity.
You are a competitive-intelligence researcher running inside an enrichment pipeline. Your job is to mine a company's case-study or customers page and extract who they serve and what results they claim — the raw material for 'you work with companies like X' personalization and for competitive displacement plays.
Work from the case-study, customers, or testimonials page you are given (a URL to browse or scraped page content).
Extract up to 3 case studies. For each, capture the named customer and the headline result exactly as claimed. Rules:
- Only include customers named on the page. Logos without an attached story count as a name with no result.
- Quote results as stated ('cut onboarding time 40%'), never rounded, never embellished.
- Skip anonymous case studies ('a leading fintech') entirely.
- Prefer case studies with numbers over ones with vague praise.
OUTPUT CONTRACT — follow exactly:
Return a single JSON object:
{"case_studies": [{"customer": "<exact company name>", "result": "<claimed result verbatim or paraphrased under 15 words, empty string if none stated>", "industry_hint": "<one or two words, empty string if unclear>"}], "customer_count_visible": <integer — total named customers or logos visible on the page>}
- Return ONLY the JSON. No markdown fences, no commentary.
- If the page is unreachable or names zero customers, return exactly: {"case_studies": [], "customer_count_visible": 0}
- Never invent customer names or results. If unsure whether a name is a customer or a partner, exclude it.How to use it
- 1
Copy the prompt into Claude, ChatGPT, or any LLM — or into a Clay Claygent column.
- 2
In Clay, map the case-study page URL as the input; find it first with a Claygent step that locates '/customers' or '/case-studies' on the domain.
- 3
Run a 15-row sample and verify every extracted customer name against the live page — the exclusion rules should leave zero invented names.
- 4
Split the JSON into columns and merge the strongest named customer into your opener or your competitive battlecard.
Best practices
Run this on your competitors' sites quarterly — their newest case studies tell you which segment they're winning and where to counter-position.
When personalizing with a prospect's own case study, reference the result, not the customer name — 'the 40% onboarding claim on your site' reads as deeper research.
Treat customer_count_visible as a maturity score: pages with 20+ logos need a different angle than a startup showcasing its first two wins.
Example: what this looks like in practice
A founder selling implementation services runs this across 800 vertical-SaaS vendors. 510 rows return at least one named case study; the JSON gives her each vendor's flagship customer and claimed result. Her campaign opens: 'The Brightline story on your site claims a 6-week rollout — curious how many customers actually hit that without services help.' The line only merges when a result string exists, falling back to a customer-name variant otherwise. Reply rate lands at 6% and three replies start with 'ha, fair question'. She also flags 12 competitors' pages, mapping 31 named accounts for a displacement sequence.
Best fit
This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.
Apply for a Pilot Spot → →Frequently asked questions
Their own marketing gives most of it away: customers pages, case studies, testimonial sliders, and logo walls. A Claygent pointed at those pages with a strict extraction schema pulls named customers and claimed results at scale. It only captures what's public — which is exactly what you can safely reference in outreach anyway.
More personalization at scale (clay & claygents) prompts
Generate a cold email first line from scraped research data
The workhorse of AI-personalized outbound: converts whatever research your enrichment waterfall produced into one 12–22-word observation-led first line per row. The swap-test rule and fact-invention ban are what separate lines that read researched from the AI slop prospects now delete on sight.
Turn a Claygent website scrape into a personalized one-liner
Produces a sub-25-word, website-specific opening line for every row in a Clay table — no human in the loop. The output contract forces plain text only and a NOT_FOUND fallback, so unreachable sites and thin homepages return a filterable sentinel instead of a hallucinated compliment that ends up in a prospect's inbox.
Analyze a pricing page for sales-relevant intelligence
Extracts a structured pricing profile from any pricing page: model, tier count, exact displayed prices, free-offer type, and a one-line read on their sales motion. At scale it segments prospect lists by how companies charge — which predicts budget, buying process, and which of your angles will land.