Enjoy the benefits of fast turn-around, flexibility, and the ability to generate timely reports from this database approach.
JSI’s merge-purge processing is designed to put rented and/or traded files in a common format and eliminate duplicate names and addresses. However, it does much more than this. It handles large numbers of different data sources arriving at different times, supports complex designs for the output of records to test/control panels, and has many optional “bells and whistles” that we have built to resolve data-quality problems, to eliminate poor quality records, and to improve targeting.
The JSI Database Approach
As we receive the rented or traded lists that will be merged to create your output file, we build a database specific to your job requirements. This allows an output file to be built in stages as rented lists arrive. Fast turn around, flexibility, and the ability to generate timely reports are supported by this approach.
Our work can be completed quickly because your lists are processed as they arrive.
Daily Status Reports keep you updated on outstanding lists and provide statistics on duplication rates and name surviving, allowing real time decisions to achieve your desired goal.
Because your rented lists are processed as they arrive, most of our work is already complete by the time the last of your rented sources arrives.
When you have ordered a large number of lists for a mailing, things don't always go as planned. Our database approach means instant information that allows you to act immediately if a list goes missing or is not "as ordered". As lists arrive, we can tell you how many names are in the "current" output file so that you can react and keep to your mailing timetable. If you fall behind on your timetable because lists do not arrive on time or because you have insufficient output, we can create initial output followed by a “make up” panel.
Because we build a database, we keep you up-to-date on "net" (merged-purged) names available for your output file and we do this automatically as your rental lists arrive.
Once the output file of merged-purged names has been shipped, we provide complete reports on sources of names, net names mailed by source, overlap between lists prior to the purge of duplicates, sources of names for each panel, and the impact of any "kill" lists used.
Control Over Matching
Our matching algorithms rely on decades of experience in writing software to handle Canadian names and addresses. They match records despite most spelling errors and other inconsistencies in both surnames and addresses. You can control the "closeness" of matching by choosing from varying matching procedures. Digits from address lines can be included in calculating matches or they can be left out. One or two letters from a first name can be included or left out. Urban records can be treated differently from rural. For example, you can choose a "coarse" match using surname and address lines for urban records, and a "fine" match for rural records. A fine match uses one or two letters from the first name as well as surname and address information.
And we will advise you on the matching option that we think is best suited to your job given your objectives, and the nature and quality of the lists that you are using.
We offer a wide range of optional processing steps that can be “turned on” during merge-purge processing. Among these are: Address Correction, National Change of Address (NCOA™), Gender Assignment, Name Parsing, elimination of addresses that are “too long” to fit output limits, reduction of files to one record per household, and elimination of records addressed to corporations. For a more extensive list of the optional processing steps offered in conjunction with merge purge processing, see Bells and Whistles.
List Priorities and Control Over Selection
You may want some lists (for example, your house list or your house-lapsed records) to have priority as sources of names in your output files. Or you may want names that are common to more than one list to be randomly selected from the list sources. We can select using one of these principles or a combination of the two. For example, you can designate your house list as a priority source and otherwise require random selection of names that are common to more than one of the remaining lists.
We can also enable “kill lists” on a priority basis. For example, we can organize your main house file to kill against all your rented/traded lists, but not against a file of “lapsed” house records that you want to mail in the output from your merge-purge.
We can attach source codes to the records that you mail and divide your mailing into panels that you define. We can also preserve your mailing database and once your campaign is over and you have sales or donations from it, we can match these records back to the original mailing file and report to you on the rented or traded lists from which these returns came.
Support for Tests and Controls
Our methods for allocating records to test and control panels ensure the equivalent record composition in test and control panels and support reliable measurement of test/control differences.
NCOA is a Canada Post trademark. Postal Code is an official mark of Canada Post Corporation.