Data Quality Management for Membership Organisations & Associations

Polishing the crown jewels - data quality management

For any customer centric association and for membership organisations specifically, keeping an accurate, up to date and clean collection of membership data is an essential part in being able to provide the kind of customer experience that a member expects. But keeping clean data, de-duplicated and up to date is no easy task unless it is planned for and built into the day to day management of membership information.

Membership data comes by various sources, from:

• Completed membership forms

• Website form submissions

• Particulars telephoned or emailed through

• Business cards collected at events

• Faxes received

• And many other channels

So owning a centralised store for core membership data rapidly turns into an important requirement for any membership organisation. There is nothing more embarrassing or harmful to a members idea of an organisations perception of them, than being wrongly communicated to by two areas of the same organisation who hold different views of a members details. Accordingly having a clean, precise “single view of the truth” is crucial.

So how do we go about obtaining clean data and keeping it clean? The foundation of any data management policy is a periodic bulk data cleanse exercise that takes all important data - at an organisation and/or individual point - and de-duplicates the data against itself and then cleans important data fields against established key reference information. For example:

1. Validates addresses against the post code data file

2. Validates that phone numbers are in the right format and have the right std code for the location provided

3. Validates that email addresses are in the right format and have valid domain names associated to them

This periodic data cleanse will give a clean base from which to build going onward.

Keeping the data up to date and clean on an continuous basis must then be developed into business procedures and technology solutions as data is updated by membership management staff and/or by the members themselves. Giving data validation facilities at the point of data entry, whilst not impacting in a damaging fashion to the data entry experience, ensures that the information added matches data reference points. For instance checking the address lines are right for a particular post code, the std code for the phone number is correct, and the email address has a valid domain name. All of these “lookup” facilities are easily built into a data entry user interface - for both membership employees and members themselves - and add immense value to making sure that clean information is entered into a software solution.

As soon as the basic data has been validated and cleansed and the record is entered to the database, it is essential to make certain that the record being entered isn’t already present in the database and, hence, is being duplicated. Data de-duplicate algorithms should be configured appropriately for your particular required level of de-duplication checks to make sure that you can discover the majority of duplicates but, at the same time, do not negatively impact on the process of entering data into the software solution. For example, you will need to determine which combination of data fields should be used for a duplicate check - in most cases, a combination of first name, last name, post code, first line of address, and email address are employed.

Within the combination, you’ll want to determine the degree of “fuzziness” that you will use to match information - to illustrate, is “Dave” and “David” exactly the same, is a small mis-spelling of a first line of address (a common thing in web form completion) satisfactory and will it be able to be used for de-duplication purposes? A competent data management specialist will be able to direct you on the essentials for such algorithms and will be able to install them for your particular situation. These algorithms should be reviewed repeatedly to make sure that they are meeting your requirements.

The matter of when, and by whom, data de-duplication should be carried out should be addressed. For membership employees, should it be upon entering and saving information within the software solution (i.e. at the point of data entry) or should it be a separate offline activity (or should it be a mixture of both)?

For members entering or modifying their data on line then it is highly unlikely that you will ask the member to de-duplicate their own data - there are vast data protection and privacy implications of permitting them to de-dupe themselves. Consequently where and by whom, is such type of data de-duplication to be completed? Yet again, your data management policy should address these things and should be periodically reviewed to ensure that it is functioning from a data quality viewpoint and a customer experience viewpoint.

As we have seen from just the short outlines above, there's often a fair amount to think about with regards to a data management strategy. It is vital that any data management policy is created on realistic, pragmatic customer and member engagement strategies and becomes a natural part of the customer/member engagement procedures. If it becomes an artificial “add on” to such processes, then it won’t happen and your data will promptly decay and become worse than useless. A clear, concise, common sense data management strategy should be documented and adopted across the company. It also needs to be reviewed on a frequent basis as a vital part of the general customer/membership management review process. Only after that can it provide an invaluable platform for the continuous quality management of membership data, providing the foundation for the development and implementation of all product and service offerings that your association wants and needs to provide to its membership base in the future.

Comments? Questions? Email Here

© HowtoAdvice.com

How to Advice .com