CRM: Customer Relationship Management
meta-data: definitional data that provides information about or documentation of other data managed within an application or environment, basically data that controls data.
The purchase and installation of a CRM system may be an enterprise's most important initiatives, yet it's one that's difficult to get right. Spending on CRM initiatives is actually on the rise--Meta Group expect the market for CRM software to more than double from $20 billion in 2001 to $46 billion in 2003. At the same time, however, some studies indicate that more than half of CRM projects fail. A February 2002 report from Jupiter Media Metric indicates that U.S. businesses will spend $500,000 or more on CRM technology over the next 24 months. This is more than they will spend on other large-scale infrastructure initiatives.
With this huge increase in spending and investment in such enterprises the underlying question here is how will it affect your data and your clients. It is estimated that customer data degrades at about 2% per month meaning that up to a quarter of your data could be invalid or worthless in as little as a year. Data entry workers input poor or incorrect data because of time constraints, poor training or low pay. Duplicates are entered into the system because the initial system is setup in a manner that doesn’t allow for multiple contacts or multiple orders either within a company or across departments. This introduction of duplicate values presents a whole new set of issues when it comes time to mail and print. Extra postage may be incurred and print runs may exceed requirements for the mailing.
Customer and contact entries are frequently non-standardized and channeled across the entire enterprise wide area networks. Companies with multiple locations in various geographical regions may not be using the same system or even sharing their data with others across the enterprise.
In the end, the client suffers because you may not be able to contact them about a product, promotion, and sale or even get them the marketing material or collateral they require. Even worse, bills and invoices may not get sent or received creating accounting nightmares. This poor data quality can undermine the success of a CRM system preventing you from having a clear picture of who your clients are and how they behave.
Organizations can take the following steps to help ensure that the data entering the system is as pure and applicable as possible.
1. Setup data entry screens that prevent poor data from being entered. For example when the ZIP code is entered have the system pull in the applicable city and state automatically.
2. Setup individual fields for the name line components such as a field for pre name, first name, middle name, last name, etc. This aids in the process of text mining and duplicate detection later on.
3. Limit the people who have the ability or access to change certain key components of the customer information such as billing address or what services they prefer.
4. Attempt to configure your system to check for duplicate entries on combinations of key data such as ZIP code and address. If it finds a match have it either alert the person entering the data or return the similar values for their reference to determine if the entry should be completed.
5. Do NOT put irrelevant or inapplicable data in a field that is doesn’t belong in. For example, don’t put “Homeless” in an address line if you don’t have one for obvious reasons. Don’t put accounting information in mailing fields such as “Doesn’t pay bills on time” or “Owes $2390 right now.” This type of information can seep into a mailing and inadvertently get labeled.
6. TRAIN, TRAIN, TRAIN everyone who will be touching the data on the value and the importance of good clean data. This cannot be stressed enough. With a proactive approach to data cleansing you can prevent data bugs from haunting your system for years to come.
7. Setup systems to properly characterize certain types of clients and metadata that matter to your business models. A good system will allow key people entering data such as telemarketers or sales people complex enough comments to define the full extent of the relationship rather than forcing them to select from simplified categories. Within this idea organization and training are key. This allows for text mining to be done on the data at a later time.
8. Setup systems and processes that will periodically review the core data involved for anomalies and errors allowing a data team to go in and make necessary corrections.
9. Someone has to get buy-in and make everyone aware that this makes sense for the organization not the individual. None of this can be accomplished without executive or ownership sponsoring the initiative.
10. Departments, branch locations and all offices across the enterprise must share sacrosanct data. The more they share it the less they jeopardize the accuracy and validity of the aggregated data. In the end CRM is only as valid as the information put into it and maintained. The outcome of poor data means that your CRM will do more harm than good to your organization and you may be calling on your clients to find out what services or products they purchased from you last year.
11. Data fields, tables, and systems must be dynamic. No one knows what the future holds in terms of data quality needs. Remember to try and anticipate future demands by not “locking” your system structure down too tight. Setup systems or change current systems to track who did what and when and hold individuals accountable for incorrect information going into the system. It may be necessary to setup training sessions and do periodic evaluations of the data so that retraining can take place if necessary.
12. Make sure your system has a viable export and import routine. You must be able to get your data out, have it cleaned, analyzed, manipulated and brought back in without any degradation occurring. This also allows you to perform valuable statistical and empirical analysis along with data mining and text mining if necessary.
If you’re in the market to purchase a CRM System, don’t be tempted to buy the first one that comes along. Make sure the system fits your budget and the overall needs of your internal clients as well as your external sales force. Obtain a CRM system that allows connectivity to other systems and information sharing across multiple platforms, and your networks.
Tyler Anderson, IT Director, immedia