De-Duplication

How Does It Work ?

Well, ADMAR have been enhancing and managing data since 1990.

Our highly skilled and experienced Bureau Team have de-duplicated thousands of files from simple stand-alone requests right through to complex multi file projects.

We have worked for small niche clients with a few hundred records up to blue-chip companies with millions of records. We offer the same high level of planning, professionalism and customer care to every project regardless of the size.

WARNING

Regardless of how thorough the process is, or how cleaver our routines are, we can not stress enough that De-Duplication is more of an art than a science. For example you may tell us to assume that Dr J. Smith is the same as Mr John Smith, but what if the second record refers to Dr Jayne Smith - this may only occur in 1% of cases, but it could occur.

The main problem with DeDuplication is that whatever rule you decide to apply, one record will always be the exception. You can not therefore expect any process to be 100% accurate.

At the start of the exercise we will work closely with you to understand exactly what level of match is required to be deemed a duplicate, for example is

Mr A Jones the same as Anthony Jones ? or is
Shellmex (UK) Ltd the same as Shell Ltd.
Can we use phonetics so that Miss Smith and
        Mrs Smythe are matched ?

We also need to understand the purpose of the project, for example are you looking to mail just for one contact per company, if so which criteria should we use to decide which should be the unique records, or do you want to be left with a file containing all distinct individuals within the household.

Using in-house bespoke routines we run your file through a number of different processes before we even start the de-duplication process. The Processes Are:

	Give each record a unique file and record number
	Standardise Addresses and Enhance if Necessary (where possible)
	Split the Individual or Company Name into elements and standardize
	Identify duplicates according to client specifications using standard and phonetic searches in multiple waves
	Add a field to the duplicate record to show which record it was matched against
	If requested Transfer data between records
	Collate the cleaned data and export as instructed
	Return to client with a data report