Hey team, im using Zenrow to scrape website info, I want to extract 2 piece of info and put them into separated columns, 1 is for company links that contains "product" "services" or offering. 2 is for finding key words about their offering. What's the best way to do this
To extract company links and keywords about offerings using Zenrows in Clay, follow these steps: 1. In your Clay table, click 'Add enrichment' and search for 'Zenrows'. 2. Select the Zenrows action under 'Integrations'. 3. Set up the Zenrows integration with these inputs: • Company URL: The website you want to scrape • Autoparse: Enable this option • HTML Output Fields: Use this to specify the elements to extract • Render Javascript: Enable if needed • Premium Proxy and Anti-Bot: Enable for better results For extracting specific data: 1. Use CSS selectors in HTML Output Fields to target links and keywords. 2. Create two separate columns in your Clay table for links and keywords. 3. Use Clay's column extraction tools or AI formulas to refine the scraped data into the desired format. If you need more detailed guidance, please ask and I'll be happy to help further.
i do not see the option to "1. Use CSS selectors in HTML Output Fields to target links and keywords."
I apologize for the confusion. The HTML Output Fields option should be available when setting up a Zenrows scrape in Clay. Here's how to find and use it: 1. In your Clay table, click 'Add enrichment' and search for 'Zenrows'. 2. Select the Zenrows action under 'Integrations'. 3. Look for the "HTML Output Fields" option in the configuration settings. If you still don't see this option, there might be an issue with your view or the integration setup. In this case, please request human assistance by clicking on "Talk to a person" for further help troubleshooting the Zenrows integration in your Clay account.
okay that's done. I have the info scraped. Now what do I do
Now that you've scraped the data using Zenrows in Clay, you can refine it using Clay's column extraction tools or AI formulas. Here's how to separate the links and extract keywords: 1. Create two new columns in your Clay table: one for links and one for keywords. 2. Use Clay's column extraction tools to filter links containing "product", "services", or "offering" into the links column. 3. For keyword extraction, use Clay's AI formulas to analyze the scraped content and identify key terms related to the company's offerings. 4. If you need more specific guidance on using these tools, you can explore Clay's documentation or ask for further assistance. Remember, if you encounter any difficulties or need more detailed instructions, you can request human assistance by clicking on "Talk to a person" in your current conversation.
what's the best way to extract customer stories from websites
To extract customer stories or testimonials using Zenrows and Clay, follow these steps: 1. Set up a Zenrows scrape in Clay by clicking 'Add enrichment' and selecting Zenrows. 2. Configure the scrape with these settings: • Company URL: The website you want to scrape • Autoparse: Enable • HTML Output Fields: Use CSS selectors to target testimonial sections • Render Javascript: Enable • Premium Proxy and Anti-Bot: Enable for better results 3. After scraping, use Clay's column extraction tools or AI formulas to refine the data and isolate customer stories. 4. Create a new column in your Clay table for the extracted testimonials. This method allows you to efficiently collect customer stories from websites, even those with complex structures or protection against scraping. If you need more detailed guidance, please request human assistance by clicking on "Talk to a person" in this conversation.
what if it says Cell data size exceeds limit (8kB) if im trying to move the data into independent column? Should I use AI formula to analyze data while it's inside of the parent column?
If you're encountering a "Cell data size exceeds limit (8kB)" error when moving scraped data, using an AI formula is a great solution. Here's what you can do: 1. Instead of moving the entire data to a new column, keep it in the parent column. 2. Use an AI formula to analyze and extract the specific information you need. 3. Create a new column with the AI formula to process the data without exceeding the size limit. AI formulas are credit-free and can help you format, clean, and reshape your data exactly as needed. This approach allows you to work with large amounts of data efficiently while staying within Clay's column limits. If you need more detailed guidance on using AI formulas, request human assistance by clicking on "Talk to a person" in this conversation.
Hey team, here's the table: https://app.clay.com/workspaces/449716/workbooks/wb_pRAbcZiukf2B/tables/t_3RaUKZXsPqFk/views/gv_wRYNqRdjvs7U It's working fine but im looking for tips/advice for better practices. Im using GPT 4o mini in Claygent to analyze data extracted from Zenrows to decide on company primary offering. Result's solid but I wanted to see if there are better/more way to utilize this. Another question, can I somehow have AI go into the customer case studies pages? Zenrow extracts all links, how do I use formula to find ALL the testimonial/case study pages?
Hey, If everything’s running well with GPT-4o mini and Zenrows, I’d say you’re already on a solid setup—no need to over-optimize what’s working unless you’re hitting limits or noticing gaps. That said, if you want to go deeper, here are a couple of suggestions: 1. Merging Data Columns: You could experiment with combining extracted fields into a single prompt to give GPT more context, but only if it’s improving accuracy. 2. Finding Case Study/Testimonial Pages: You can use the “Find Sitemap URLs” enrichment to pull all structured links from the company’s website. • Then use a formula column to flag links that include keywords like "case-study", "testimonial", "customer-story", etc. • Example prompt-style formula: If /url contains “case-study” or “testimonial”, return the complete url For better accuracy, just do this with Claygent to search for and extract the most relevant case study pages directly. Let me know if you have more questions
.png){kind=logo}
