Deduping improves the ROI of a mailing by removing duplicate records from the mailing list. But despite the simplicity of the concept, the process of deduping can be quite complex.Choose-Match-Criteria-1024x512

There are two key types of rules you need to decide upon for your deduping process. The first one is what level of deduping best meets your organization’s needs. Do you want to send separate pieces of mail to two different individuals at the same address? How about two different families at the same address? The way you answer these questions will help you determine which of the following consumer-focused matching levels would be best for your organization:

  • Individual: Avoid sending duplicate mailings to a particular individual by flagging records where the first name, last name and address all match.
  • Household: If you don’t want to send more than one mailing to a particular household, this type of matching flags as duplicates records that share the same last name and address.
  • Resident: If you want to avoid sending more than one mailing to a particular address, use resident matching to flag all records with the same address.

Because people’s names get on lists in a variety of ways, misspellings, nicknames and abbreviations are rampant. While the data hygiene process does a good job of standardizing addresses, first and last names can vary widely. For example, is “John Martin Cook, Jr.” the same person as “Johnny Cook”? How about “Martin Cook” or “John Cooke”?

So the second major decision you need to make is how tight you want your match criteria to be. Fortunately, matching software knows how to take common nicknames into account, and will mark “Bob” as equivalent to “Robert”. But after nicknames are taken into account, imagine giving each set of two records a score between 1 and 100 based on how similar the two names are. If every character of the two records is identical, it would score a 100; if no character is the same, it gets a 0. You can then choose which of the following levels of matching you want to use based upon the two records’ matching score:

  • Exact: 100
  • Tight: 90 – 99
  • Medium: 75 – 89
  • Loose: 60 – 74

To make things even more complicated, you can set different levels of matching for different elements of the two records. For example, if you want to err on the side of not dropping individuals who have donated before from your list, even if they live at the same address, you might require the name to be a tight match, the address a medium match, and the zip code to be exact in order for two records to be flagged as duplicates.

As you choose the appropriate matching level and match criteria for two lists, think through the consequences of eliminating a record inappropriately or of not eliminating a record that is a match. For example, if you want to make sure that major donors don’t get mailed a low-dollar solicitation, you might choose to match at the household level using loose match criteria. On the other hand, if you want to mail a solicitation to as large a list as possible and not much harm would be done if you mailed another solicitation to someone who gave $25 last month, you might choose to match your recent, low-dollar donor file to the solicitation list using tight, individual match criteria.

The many options for match criteria and levels of matching give you a great deal of control over how to dedupe your lists. However, if you find the plethora of options confusing and would prefer to rely on experts who’ve been helping companies like yours manage their lists for decades, please call John Bell at (310) 372-9010 or tell us when would be a good time for us to contact you.

Want more than a cookie cutter approach? MMI Direct dives deep into your data
to determine how best to optimize your list and maximize your direct mail ROI.