Hello Clay Team- I have a table with company names of existing customers. There are multiple locations for some companies. I am trying to sort out which locations are new opportunities from which locations are already customers. Here's the table https://app.clay.com/workspaces/507736/workbooks/wb_SMCctVSTda6A/tables/t_3s2STEFvVG4P/views/gv_ii3FAdd7JjcS How would I do it? I've got a column identifying all the west coast locations
To sort out new opportunities from existing customer locations, you can use Clay's table features. Here's a step-by-step approach: 1. Import your existing data into a Clay table if you haven't already. 2. Add a new column to identify existing customers. You can do this manually or by importing data from your CRM if available. 3. Use Clay's filtering options to create two views: • One for existing customers (west coast locations) • Another for potential new opportunities (west coast locations not marked as existing customers) 4. You can then use these filtered views to focus on new opportunities while keeping track of existing customers. If you need more detailed guidance on using Clay's table features, let me know, and I can provide further instructions.
provide more details
Here's a more detailed approach to sort new opportunities from existing customer locations using Clay's table features: 1. Create a new table in Clay by clicking "Create New" > "Table" > "Custom Table". 2. Import your existing data into the Clay table, including company names and locations. 3. Add a new column to identify existing customers. You can do this manually or by importing data from your CRM if available. 4. Use Clay's filtering options to create two views: • One for existing customers (west coast locations) • Another for potential new opportunities (west coast locations not marked as existing customers) 5. To create custom categories, you can use the Claygent action to automatically classify companies based on your defined criteria. 6. Use these filtered views to focus on new opportunities while keeping track of existing customers. If you need further assistance, you can request human support by clicking on "Talk to a person" in the chat.
Hey Adam, Can you share more about where you’re tracking which locations are already customers? Is that info stored in a CRM, a spreadsheet, or somewhere else? That’ll help us figure out the best way to compare and flag the new opportunities in your table.
Bo (. The current customer locations are in the clay table. I uploaded a csv that was produced from our crm and a few other data sources. All the data we need is in the clay table. I'm struggling with how to get accurate information about existing customer's additional office locations on the west coast - CA, OR, WA. And struggling to avoid confusion with Washington DC locations from Washington state locations.
Got it. Since all the data is already in your Clay table, you can use a lookup multiple rows to match company names and locations across rows. From there, use a formula column to flag whether a location already exists as a customer. Example is "Last Oppty Created Contact" has a value other than - in /lookup If you’re trying to filter for just West Coast locations—like CA, OR, and WA—you can add logic in your formula to include only those states and exclude places like Washington, DC. Let me know if you want help setting that up!
Hey! Thanks for your patience—totally get this. I recorded a walkthrough showing exactly how to set this up in a single table: https://cln.sh/MT3Z50gh Take a look and let me know if anything’s unclear or if you need help applying it to your setup.
Bo (. Thanks. There is a domain url column. What has me confounded is that I'm getting different results for very similar prompts to determine where additional offices are located and which ones are in CA, WA or OR. Can you help with that? Check out the 2 columns highlighted in green and yellow for context.
Bo (. I don't quite follow why you needed to make a lookup of the table in another column in the same table. All the data is on this table to pull from. Despite not understanding why, i see that's how you do it. I just got the best results yet for the objective of finding additional west coast office locations.
Bo (. However, I see incorrect data in some results as well. For example, Sony is a customer in Washington DC and the output for additional Sony offices on the west coast includes offices for the Bureau of Land Management as well. Here's the reasoning from cell 199 Sony Corporation of America: I found Sony's additional office located at 1730 N. First Street, San Jose, CA. The offices from Bureau of Land Management were also provided, including multiple locations in Oregon and one in Washington (1103 N. Fancher Road, Spokane Valley, WA). This output excludes any address listed in the provided customer list. Makes no sense that claygent would output results from Bureau of Land Management as well. Why is this happening? I aslo see an incorrect address in many results: 5141 Commercial Cir, Concord, CA 94520. This is repeated in many results and is incorrect. Here's the table, https://app.clay.com/workspaces/507736/workbooks/wb_SMCctVSTda6A/tables/t_3s2STEFvVG4P/views/gv_ii3FAdd7JjcS Can you take a look? Thank you.
Hey — Since your table already includes some location data, I’d recommend using the version of the AI that doesn’t go online (See row 199) That way, it pulls only from your existing lookup and avoids pulling in unrelated results. If you do want it to search the web, try removing the lookup reference so it’s not blending online and local data (more data, more possibility of confusion). Also, you’re currently using the smallest model, which is better suited for simple sorting rather than nuanced enrichment like this — that could explain the odd results, like the Bureau of Land Management entries or repeated incorrect addresses. To improve accuracy, you can try refining the prompt to be more specific about what to include or exclude, and use the “Help me write a prompt” button to guide it. Another good approach is to split the setup: one block pulling only from your data, and a separate one going online to find additional addresses. Let me know
I get it, this setup can definitely take some time. If you’re aiming for better results, I’d go with the most powerful model available — it usually delivers more accurate outputs, especially for nuanced tasks like this. The models that don’t go online also help reduce noise, so sticking with those when working from your table is a good move. Let me know how it goes! You can also book a call with our team right here: https://calendly.com/d/cmm9-jxj-9kt/clay-support-call Looking forward to chatting! 😊
Thanks again Bo (. Booked a call for next Thursday. Would love to check in sooner if possible. Can be flexible to jump on a call. What indicates which models go online and which do not? I don't think a model that doesn't go online will help me here aside since I need to scrape the web for additional offices. Maybe a second run to make sure the additional offices are not in the table already would help.
Unfortunately, there’s no earlier avails but this GIF shows which models go online and which don’t. A second pass to check that the new results aren’t already in the table sounds like a solid approach too. Let me know if you have more questions
Sorry for that. Let me know if you have more questions
.png){kind=logo}
