Join Tables with Fuzzy Matching

[davidlittle]

I have two databases that house similar information that I'm attempting to join/merge together.  The main join takes place between Company Name, and there is not always consistency with the company names, even within the same database.  I'm attempting to create a common Company Name between the two databases so that I can use Company Name (along with a few other fields) to join the two sources.  Here is an example:

Database 1

Company Name	Event Date	Notes	Contact
XYZ Retirement Communities	1/1/2019		Jane Doe
XYZ - Retirement Communities	7/1/2019	At Location Y	Jane Doe
XYZ Retirement Communities	12/1/2019		Jim Smith
123 LLC	1/1/2020

Database 2

Company Name	Event Date
XYZ Retirement Communities	1/1/2019
XYZ Retirement Community	7/1/2019
123 LLC	2/1/2020

Current Efforts

My original efforts were to join the tables together, then remove all duplicate names, leaving me with just a list of unique Company Names, and then use a fuzzy match.  What this left me was the following:

Company Name	Match Name
XYZ Retirement Communities	XYZ Retirement Community
XYZ - Retirement Communities	XYZ Retirement Communities
XYZ Retirement Community	XYZ Retirement Communities
123 LLC	123 LLC

What I'm seeing with my results is that I end up with multiple matches that don't always tie together.  In this case, my hope was that this match would result in the same Match Name for all 3 original Company Names, but instead I'm showing more than 1 Match Name.

Goal

Looking at Company Name, there are three separate versions of XYZ Retirement Communities.  All of these separate versions are the same company, and there should be a common name both to join the tables, and to ensure that within the individual tables, we have the same Company Name.  My thoughts were to create a new field for each table called "Common Company Name" that would be used to then join the databases.  M

[Bennel_Wilson · Accepted solution]

Hi @davidlittle ,

I tried a workflow based your query. Let me know if this works.

Regards,

Benn

P.S: If this solves your query, kindly mark it as a solution.