Data Cleaning Part 1- Deceased and Goneaways

24th September 2016

If you are sending out Direct Mail then you will be using data, collections of names and addresses of customers or prospects that you want to reach with printed communications. Depending on where you obtain your data it may be accurate or inaccurate, it may be considered recent and fresh or old and stale.

Personal details don’t stay the same, people may move house, get married or change their name, refuse permission to mail them certain item, or they may die. There are several ways of data cleansing and today in Part 1 we are going to start by talking about removing deceased and gone away records from your data.

There is no point in sending a mailing item to an individual’s previous address or to a deceased person, and people are often upset or unimpressed when they receive mail with an inaccurate name; those people who have requested not to be mailed may complain to your company or to an official body if you continue to mail them direct mail items.

Industry Standard Files

There are several standard industry files that your data can be matched against to remove deceased or gone away records; deceased files include:

  • TBR – The Bereavement Register is a file of deceased individuals who have been registered by their relatives.
  • Mortascreen – A file of deceased individuals gathered from both verified and un-verified sources.

Gone away files include:

  • GAS – The Gone-Away Suppression file is made up of confirmed gone-aways.
  • NCOA Suppress – The National Change of Address (NCOA) is a dataset that is sourced from the Royal Mail’s redirection service that keeps track of customers who move house. There are over 23 million records in the NCOA® Update file, and it is growing by 1.2 million every year.

Check out our in depth look on Industry Suppressions for more detailed information on industry suppression files.

Person or Household level Matching

Matching records against stop files can be done at different levels with the most common being household and person, the level you choose may depend on your requirements for a specific mailing or the maintenance of a database or list.

If you want to clean up a database of your customers for example, then you would match the files at a person level, at this level we try to identify individuals and remove the actual person at the actual address from your database.

For more sensitive mailings you will want to remove the deceased and gone away records, but you may also want to remove people who live in the same household as the deceased so you do not cause upset to those families; this is household level matching.

Temporary or Permanent Matching

The type of matches that you pay for are classed as temporary or permanent matches. Temporary matching means you can drop records from a single mailing that match the relevant stop records. Normally your data bureau will perform the matching and drop those records without telling you which records have been dropped; if you choose to do another mailing in the future with the same data you will have to run the matching again. Temporary matching s the cheaper option if you’re sending a one off mailing.

Permanent matching is more expensive but your data bureau will send you the records that have been stopped so you can update your mailing list or database for future mailings meaning you won’t have to pay for those stops again.

Industry Suppression files are updated regularly and if you are mailing a list you have previously cleaned it may need cleaning again with up to date stop data.

Saving Money

Matching may sound expensive, but remember that although you are paying for the cleaning, a match is cheaper than postage. If you have opted for permanent matching and practice regular maintenance of your list then it will be cheaper in the long run than paying for postage to people who won’t receive your mailing as intended.

Next Time

In part 2 we’ll look at other data cleaning techniques including de-duplication, word processing, customer suppression files, and PAF correction.