Hi Everyone I have a list of e-commerce companies but don’t have their LinkedIn page URLs. I tried using a Claygent prompt to extract Company linkedIn URL by scanning their websites, but it’s not very efficient. Is there a better way to find company LinkedIn pages and associated people data?
use the native clay "enrich company" action instead of claygent, way more efficient. just feed it the domain and it returns the linkedin url plus a bunch of other company info like size, industry, funding etc all in one shot.
I did try that and it wasnt doing a good job either.
yeah enrich company can miss sometimes especially for smaller ecom brands. the issue with claygent is usually the prompt isnt specific enough. what prompt are you using and what output are you getting back? share it and ill take a look, probably just needs a small tweak
Yeah possibly these companies are not on linkedin that's why I am unable to find their linkedin URL since the linkedin URL is not available, clay is unable to find the people data as well which is a big problem for me right now here is the prompt that I am using - #CONTEXT# You are an AI-powered web scraper that performs multi-step company research to locate and validate a company’s LinkedIn profile URL. You will use the input company identifier provided in the dynamic field to: (1) determine the correct company and its public website domain, (2) crawl and score relevant website pages to identify footer, contact, about, careers, and social sections, and (3) extract and validate the official LinkedIn company page URL. Use only publicly available, non-authenticated pages. Do not use LinkedIn browsing directly; infer and validate the LinkedIn URL from the company’s website and open web results. #OBJECTIVE# Find the official LinkedIn company page URL for the specified company, using the input field as the company reference, by visiting the company’s public website, prioritizing pages most likely to contain social links, and returning a single validated LinkedIn company URL. #INSTRUCTIONS# 1) Resolve company and domain - Interpret the input as the company reference. If it appears to be a domain, use it directly. If it appears to be a name, perform a web search to identify the company’s canonical public website domain. - If multiple companies match, prefer the one whose name and industry context are most consistent across homepage title, meta description, and about page. 2) Crawl and prioritize pages - Start from the homepage and extract all internal links up to 2 hops deep. - Prioritize and score pages by likelihood of containing social links: footer > about > contact > careers > blog > press > homepage header. - Within each page, score specific sections: footer social icons/links highest; then header; then body content mentions. 3) Extract LinkedIn company URL candidates - Extract any links containing linkedin.com/company or linkedin.com/showcase from high-scoring sections first, then from the rest of the page. - If none found on-site, run a web search query: "<company name> LinkedIn" and extract candidates that match linkedin.com/company or linkedin.com/showcase. - Ignore personal profile URLs (linkedin.com/in) and posts/activity URLs. 4) Validate candidates - Must start with https://www.linkedin.com/company/ or https://www.linkedin.com/showcase/ - Strip tracking parameters (e.g., ?trk=, utm_*). Normalize trailing slashes to a single trailing slash. - Cross-check the LinkedIn page name/slug against the company name and domain (allow minor variations, abbreviations, or hyphenation). Prefer candidates linked from the company’s own domain. - If multiple valid candidates exist, choose the one with strongest on-site evidence (footer link > about/contact pages > other pages > web search). 5) Output rules - Return a single field named linkedinCompanyUrl containing the final normalized URL. - If no valid company LinkedIn URL can be found, return linkedinCompanyUrl with value "" and an additional field reason set to a short explanation. - Do not return personal profile URLs. #EXAMPLES# Example successful output: { "linkedinCompanyUrl": "https://www.linkedin.com/company/acme-corp/" } Example no result output: { "linkedinCompanyUrl": "", "reason": "No linkedin.com/company or /showcase link found on site or web search" } #INPUTS# Company Reference:
your prompt is actually solid, the issue is these companies probably just dont have a linkedin presence at all so no prompt will find what doesnt exist. for people data without linkedin try running the domain through clay's "find contacts" with email waterfall instead, or use apollo/hunter directly with the domain. you'll get decision maker emails without needing linkedin url at all.
If it's small ecommerce companies, most of them don't have Linkedin pages... I've done ecom outreach in the past and Instagram is thee top social media platform they use
