"High quality data" is a term sought after by most Direct Marketing companies or any company seeking to increase their business opportunities through CRM or data marketing initiatives. Many companies underestimate the importance of standardised and validated address data, which often results in failed direct marketing campaigns due to the inability to maximise the potential of their data. Often the "nixie" rate (Return To Sender mail) is used as a benchmark, but it is not necessarily accurate as not allnixes are returned and RTS mail is not returned equally across all categories (Box vs. Street, Rural vs. urban, etc.).

At least 50% of companies adopting a CRM strategy are unaware of the extent of data quality problems in their environment. The other 50% realise that poor data quality may impact on their profitability, but generally do little about it due to lack of perceived importance.

For any business relying on customer contact to expand their business, using incomplete, inaccurate and undeliverable address data will result in poor responses and a loss in potential income. The accuracy of the data on customer databases is therefore an important ingredient for any marketing campaign.

Is the accuracy of your data burdening or boosting your marketing programme?
How do you determine your primary customer segments? A good basic technique is by using a simple RFM assessment model (Recency, Frequency, Monetary value) you will be able to broadly identify your most profitable customers. These variables should allow you to segment your different market segments into primary, secondary and tertiary market segments, based on your own evaluation of your most profitable customers. Other measurable transactional variables you may wish to apply would be income, amount spent over a certain period, regular paying customers, etc., which will strengthen your matrix.

The next step is to measure your primary market segment against your known valid address data.
But how do you measure the quality of your data? A few companies offer data assessment or data audit facilities, which will provide you with a reasonably accurate measurement of your data hygiene. Ideally, if you applied an address scoring methodology to your address data, it would provide an accurate, fixed matrix against which to measure your most profitable customer segment. Although PAMSS is a certain standard of address measurement, it is by no means a sound address quality scoring system as it basically only matches suburb and suburb code, ignoring possible inaccuracies in the remaining address fields. On average, 50% of data passed through any PAMSS validation system fails many stringent address standardisation rules, but often gets delivered due to "local knowledge". Generally the nixies will be much higher in this 50% category.

A standard data health or hygiene check will identify which addresses are valid and mailable by matching addresses to a known address standard, but will not be as accurate as a scoring system. Applying basic information such as RTS mail, address confirmation, response and purchase information from your last mailing campaign will also create a quick perception of the percentage of your address data accuracy.

Create a customer measurement matrix to determine if you are getting to the right customers.
Separate your address 'score' (actual score - not PAMSS passes) and market segments into two separate columns. Then cross-tabulate your most valuable customer segment to the standardised and validated address sector and apply a score to the result. This will give a clear outline of which segments of your most
profitable customers fall into 'mailable', 'unmailable' or 'unknown' categories, providing a reasonably accurate account of what percentage of your most profitable customers are in fact contactable.

Through this exercise, you will clearly see whether you are in fact contacting the right market segments. If your scoring per segment in the primary market equals your score in the tertiary market, you are heading for a potential problem.

Advanced basics - How do we start achieving true data quality?
Ideally, you should apply an address scoring methodology to all your address data to realise the true quality of your address data. By applying an address scoring system to your data, you may be surprised at what percentage of your data is valid and usable. By assigning address-scoring variables to your database, you can facilitate a wealth of new information through your existing data. Unfortunately, very few companies in South Africa offer true data scoring methodologies.

What are the benefits of address scoring?
When one looks at the variables associate with scored addresses, the benefits become self-explanatory. There are a number of benefits that one might not immediately relate to the classification of address information into categories, such as:

	Validation and mail delivery - Determine what percentage of mail is likely not to reach the desired destination before it gets mailed (as opposed to what is eligible for postal discounts).
	Prevent inaccurate/incomplete address capture at time of data capture.
	Address categorisation - The ability to categorise addresses by unit, street, dwelling type, etc.
	Cost saving - Focus on improving the address data only of your most profitable customers.
	Identify deliverable addresses - Reduce the RTS (returned or Nixie mail) component of your mailings.
	Data accuracy - Differentiate between valid, invalid and corrupt data.
	Added value in address data - A basis for creating accurate consolidation, de-duplication and household information.

Remember, the secret of successful direct marketing is the quality of your address data. If you understanding the variables you are dealing with and how to integrate them with your marketing strategy, there is no doubt that you will significantly increase the effectiveness and profitability of your campaigns.

This is confirmed by Pricewaterhouse Coopers' Global Data Management Survey 2001 that has found that companies reported problems from poor data quality, including extra costs, failures to bill or collect receivables and lost sales. Therefore, it pays for companies to examine their data quality; what it is, the role it plays in creating customer intelligence and how it contributes to a healthy bottom line.

Data quality pays - build you customer intelligence on a sound foundation, for it will create customer satisfaction and loyalty and will gain the full ROI from CRM initiates.

Winnifred Knight and Luisa Mazinter, email winn@themarketingsite.com or phone 082 575 9922

Back to www.bizland.co.za