Hi there, I run a people search in clay and opted to execute a work email waterfall process as well. However, I've observed that the waterfall process doesn't consistently function as expected. There are over 30 leads on which all the 5 emails provider have run and have the same valid email address. So that means for one valid email address are spent 5 times more credits, also on those cases the work email column is still empty.
Hey, thanks for reaching out! Taking a look but to clarify, this may be happening:
The waterfall gives you the option to add a validation provider to check results from each email finder. When the validation tool determines the email is invalid, the next email finder tries to find a new one, but if it finds the same email as the previous column it won't try to validate it again because it already tried to validate once before, and this could result in no final output because none were valid.We automatically hide validation columns helping keep tables clean and that's why it's not evident at first glance. Let me know if you still have any questions though!
When the first provider provides a valid email address, the next provider should not run because the condition is not met. However, if you check the lost this didn't happened many times.
Hey Ervin - just checked, I believe this is actually working correctly. Check out this video here: https://www.loom.com/share/74663f595b4748f8be325dde7e2302a7
Walked through whats happening on your table^
Thanks, I understand the reason behind it but if the cost is 2 credits per verified email on this way we are paying 12 credits for an invalid email address π€
Yep thatβs why we donβt re-run the email validator. Weβre also going to work in refunds if the same email was found, but its going to take some time to create on a technical side
Yep, adding a smaller note on top of what Matthew mentioned. The data comes from those providers and they charge Clay when they find something, that's why it uses credits, we have no control over what these providers return.