Hello, I am trying to use a fuzzy match to match to my salesforce records in another table. Is there a way to do this? The purpose is to be able to deduplicate against my existing salesforce leads/contacts. Thank you!
Yes, you can use the Lookup Records action to perform flexible matching against your Salesforce records. When using this action, you can toggle off the "Exact Match" option to perform a "contains" search, which acts as a fuzzy matching mechanism.Here's how to set it up:1. Choose your Salesforce object type (Lead, Contact, etc.) 2. Select the fields you want to match against (Email, Name, Company, etc.) 3. Enter your search value 4. Leave "Exact Match" turned off for fuzzy matchingThis approach is particularly useful for checking duplicates, managing exclusion lists, and verifying existing records before creating new ones. Sources: Salesforce integration overview | Clay University Salesforce Integration Guide
exactly how "fuzzy" is the contains function? Let's say that I have a clay record on Table A called Rebecca A. Smith. Would it recommend a potential match with Becky Smith on the other table? If not, is there a way to do this? I would also like to include other records in the query such as Company Name, Email, linkedIn URL, etc. The goal is to identify duplicates in an intelligent way, and perhaps AI could help with this somehow.
We help prevent duplicate data through a comprehensive approach:1. Use Clay as a central data warehouse to route and check all incoming data before it enters your systems. This acts as a gatekeeper to prevent duplicates from multiple sources. 2. Leverage AI formulas to create custom matching logic - these are credit-efficient tools that help you format, clean, and reshape your data exactly as needed. 3. Implement conditional runs to verify if contacts already exist before adding new ones. This ensures you only update truly new contacts while preventing duplicate entries.This combination of centralized data processing, AI-powered matching, and conditional verification helps maintain clean, duplicate-free data across your systems. Sources: Push Data To Your CRM [Clay 101] - Clay University Intro to CRM Data Enrichment [CRM Enrichment] - Clay University AI Formulas + Conditional Runs [CRM Enrichment] - Clay University
I'm delighted to hear that! Remember, I'm always here if you need further assistance with our app.
Hi Stephen S.! This thread was recently closed by our Support team. If you have a moment, please share your feedback:
Thank you so much for sharing your feedback Stephen S.!
Thanks! We've reopened this thread. You can continue to add more detail directly in this thread.
Hello, unfortunately, I have been unsuccessfully trying to identify duplicate records in salesforce based on new leads that I've developed with clay. Using the Lookup Record enrichment, I've tested out both on/off for "Exact Match". What I've found is that, in reality, it only finds exact matches. That means that if clay finds that someone's real colloquial name is "Andy" or "Becky" instead of Andrew or Rebecca, it won't match the record. I've tried to use multiple object fields and also tested with just one. The test that I used was an actual lead record named "Andrew Skipper". I tried changing his name to "Andy" and it didn't find him. I also tried tweaking his email slightly. Instead of "askipper@..." I changed to "andrew@..." and "andy@..." but it wouldn't identify. I also tried using exact match off for name, email, company, website and linkedin url; it came up with a bunch of records found, but they were all wrong. Am I doing something wrong here? Thank you!
This is likely happening because of the default search settings in the Lookup Record action. By default, the search performs an exact match comparison, which means it's looking for precise matches between your search terms and the Salesforce records.To find matches for slightly different variations, you can disable the "Exact Match" toggle in your Lookup Record settings. When exact match is turned off, the search will use a "contains" search instead, making it more flexible in finding partial or similar matches.Here's how to adjust this:1. Open your Lookup Record action 2. Look for the "Exact Match" toggle 3. Turn it off to enable contains search 4. Configure your search fields and valuesFor testing these changes, we recommend using a small sample of 10-20 test contacts first to verify the matching behavior works as expected. Sources: Salesforce integration overview | Clay University Salesforce Integration Guide Lookup Record with Salesforce FAQ: Is there a way I can test Salesforce enrichments?
Looks like you're still looking for help. We're escalating this over to our support team and someone should be in touch soon. In the meantime, feel free to include any other information that may be helpful in solving your issue!
Hey there Stephen thanks for reaching out, taking a look at your previous threads, the best way to approach searching for duplicates with our Lookup Rows integration is search for matches between unique identifiers such as LinkedIn Profile URL as every person will have URL that is unique to them only, regardless of what name they are using. For example, even if a person is listed as Rebecca Smith in location and then as Becky Smith in another, she will always have the same LinkedIn Profile URL as that will never change.
LuisArturo Thanks for your response. The challenges that I have are that A) linkedin url's can change, B) Not everyone has a linkedin URL. We could try company domain, but not everyone in my database has a website URL, nor do they own a website. I suppose most people either have one or the other, so I could first ensure that everyone has either a website or LI URL. However, some people have varying URL's for the same domain. I.e., the variability between duplicate records, for the same person, is rather complex. I was hoping there was a more elegant solution. E.g., I have had a developer use fuzzy matching that output A) the matched record; B) confidence score. Then, I could sort by the higher confidence score to focus on the higher probability dupes. I may have to look at creating a custom API for this, but I was hoping that you guys had something more out of the box that could at least be retrofit with your existing tools.
Hey! Yeah, this sounds pretty complex with all those URL variations. I'm curious - what were you using before for that fuzzy matching setup? If you share more about the API you used previously, we might be able to figure out a similar approach with Clay's tools! 🤔
Hey there - just wanted to check in here to see if you needed anything else! Feel free to reply back here if you do.
Sure, it wasn't an API (yet). It was a system on AWS The input was a standardized set of variables such as first name, website, LinkedIn URL, email, company name. The system would then compare with another input which was the Salesforce lead and contact set with the same available variables. It is a custom built Python script developed by our developer who's currently on vacation. There's currently a bug and he's not available to fix it, so I was trying to use clay to get around it. I would love to have this sort of duplicate matching feature in clay because then I would be able to do all of the data enrichment, deduplication, etc. and wouldn't need to mess with exporting and importing CSVs over multiple different programs in the midst of a workflow.
Hey there Stephen thank you for sharing the details on this with us. This definitely sounds like a very useful system to use in order to deal with these sorts of situations with individual's URLS. Will let the team know if there may something similar in the works for something like this.
Hey there - just wanted to check in here to see if you needed anything else! Feel free to reply back here if you do.
Hey there - just wanted to check in here to see if you needed anything else! Feel free to reply back here if you do.
We haven't heard back from you in a bit, so we're going to go ahead and close things out here - feel free to let us know if you still need something!
Hi Stephen S.! This thread was recently closed by our Support team. If you have a moment, please share your feedback:
.png&w=3840&q=75){kind=logo}
