Hi I’ve just joined from Owl Labs in Boston, MA! Looking to find net new prospective accounts not already in our (very large) SFDC database to send my team after. If anyone has any good introductory videos to this end please send them my way (and yes, I’m working through the 101 vids on the website)
Hey. I think this course from Clay University would be helpful for you! https://www.clay.com/university/course/crm-enrichment Finding companies and people is free in Clay, and the enrichment of these profiles is what costs credits/money. You can probably use Lookup functions in Clay to check against existing data and enrich the leads/accounts which are new.
Thank you! I watched all the 101 videos which these look to mirror but I’m going to go through these as well. The kicker is that I have 800k+ accounts in my SFDC and in order to do the best look up against that to eliminate then I need to enrich the Find Companies table with websites so Clay can use that to compare to our SFDC instance. So, (I think) I’m spending credits to add the website when I likely already have the website in our CRM and don’t need that info. From what I’ve seen, the starting columns for a find companies table are:
Name
Description
Primary Industry
Size
Type
Location
Country
LinkedIn URL
So the lack of domain or website doesn’t let me look up efficiently. I have no clue if I’m making any sense or not.
then enrich for domain where the website column is null?
So I have domain and website on all my 200,000+ accounts in SFDC (800k I realized is total with contacts and leads) but I can upload 50k of my sfdc records into Clay to run against their database in order to find net new accounts. To be fair, our domains are probably trash/ not normalized we’d have to work on that. I like the idea of only enriching those without a website but again the upper limit for our plan on imprting SFDC records means this would be really slow going and I can’t picture how to parse out my SFDC accounts into 4-5 non-overlapping cohorts to upload and then compare against clay’s database
FWIW I did a basic find companies and started with 40 accounts, looked up against my SFDC and 8 remained as not having a look up record aka net new. So not horrible but not great
You might have to do some high level segmenting of the 800k accounts prior to clay, then. What other data points are you working with at the SFDC account level? Could you segment by industry verticale, geography, company size, or even rep ownership in SFDC first?
Yeah I could def parse down but geo/ rep ownership to start or something like that. Our ICP/ TAM is really any company who has hybrid meetings so it’s a good problem to have but it’s so incredibly broad which is why our account base is so enormous. For the Claude example you mean export all account records to csv, upload to claude (we have gemini but presumably the same) and say give me 4 cohorts of 50k or less each and focus on any accounts over 200 employees or something along those lines? You’ve had success with this before?
Yep, exactly. You could probably also define an even more narrow ICP based on your historical closed/won or SQLs on top of the demographic and firmograohic segmenting and ask Gemini to parse all your records for that. When I say more narrow ICP I mean you could put in a csv of accounts that have converted and ask it to come up with commonalities and apply those parameters to your full data set to identify similar accounts. Ocean.io does this exact thing but idk if you want more tooling.
Let your data engineers develop a quick RAG application sourcing from SFDC to produce internally enriched dataset and integrate this App with Clay API to take the enrichment to the next level. SFDC --> Vector Database --> Embedding Model (all-MiniLM-L6-v2 (lightweight, company-data optimized) --> RAG Application (LangChain _ Saleforce API) --> Clay API