Prashant M.

I tried this. It got things complicated for me because I had to be careful with the conditional runs. If there are AI enrichments, your prompts need to be 'if this'....'then do this'...AI then doesnt give a good output If it is just 2 signals it should be ok but if you are tracking 5-6 signals in one table, best to split it. Most important - the mental overload of managing multiple complex signals.

Zapmail, Scaledmail, Premium Inbox

Use claygent to create a prompt to search for roles like director, VP of account management Put constraints saying - 'only search for roles as current experience' in the 'current company' + location/geography

Petar R. Starter plan should be ok unless you want to verify the emails on Clay itself. each verification costs 1 credit/row You can download the CSV without HTTP API, so starter should work

2. The normalization is inadequate. Sharing a note on normalization on how I use javascript + AI Normalizing using javascript This is going to save you lot of credits but it would require you to give specific instructions

The good part about javascript is it gives tight answers based on your prompt but it is not intelligent to take decisions on its own. Thats where AI comes in…but at a cost Normalizing using AI I restrict myself to Chagpt 4.1 nano at 0.1 credits per row and work AI through great prompts for good outputs. Chatgpt 4.1 nano, 4.1 mini ,4o mini, Claude 3 Haiku, Gemini 2.0 Flash all do the job really well. They cost below 0.5 credits per row. Use AI in content creation & manipulation mode only otherwise you would burn more credits.

Need to use make or n8n or zapier to do this. When you connect either of these to smartlead or heyreach, you will see inbuilt options of 'fetching campaign statistics'

Anas A.

If the above points are not the issue, then its a bug

that right, pls use 'contains' instead of equals

Claygent can do it unless the website has blocked scrapers