I work in the non-profit sector and use the Salesforce data cleaning tool to remove duplicates. However, duplicates often occur because donors use aliases (for example, different names or email addresses) and some choose to remain anonymous. What data cleaning strategies do you use to handle these challenges? Our team is small, and we have around 5,000 to 10,000 records to clean.
Feel free to ask me anything to provide better context of my situation!
's process tips for preventing dupes is the best first answer here.
With that said, it's a really common problem, and it sounds like you need tools to clean up an existing mess. Apsona has been my go-to for years, but note that if you don't already have it they changed to a bundled app pricing model at the beginning of 2025 and the entry price is higher now (note that Apsona says their new licensing model provides better overall value).
Apsona also does not automatically handle NPSP household account merges gracefully. When the last contact is merged out of a household it leaves an Anonymous Household record behind rather than automatically merging the household account along with the contact. Some other duplicate cleanup products handle this better, I know that Duplicate Check has functionality in place for it.
Speaking of merging accounts properly, to make sure everyday merges handle that properly make sure you're using the
NPSP potential duplicates component that was introduced in Spring '24on your Lightning Pages instead of the standard Salesforce component.
When performing a bulk duplicate cleanup, I've often used Apsona's feature which allows for uploading a CSV list of record ID pairs. Because of the account merge consideration, I typically do duplicate analysis outside the system. Power Query for Power BI Desktop and regular ol' Excel are my tools of choice for the analysis. I use field concatenate operations to find duplicate contact matches, then isolate the contact IDs and their account IDs. Merge the list of accounts first, then the list of contacts.
For analyzing duplicate contacts, I work through a series of match criteria starting with the most reliable and working down to the least reliable. Each pass needs to be analyzed for accuracy, and a determination made as to whether it's a good list to bulk merge. Be careful and remember that it's way easier to merge duplicates than un-merge a false positive. Standard criteria for analyzing contact dupes:
- Exact name + email + phone
- First name + email
- First name + phone
- Full name match
- Email match
I also like to run an email and phone cleanup script in advance as a value add for emails and a way to ensure that phones are in a standardized format to facilitate matching.
