October 1996
Database Maintenance Planning: House-Breaking the Monster
by Kenneth M. Culpepper
Direct Marketing Magazine, October 1996, pp.44-49
This article is part three of a series on database maintenance planning. Database maintenance planning is a disciplined process that integrates entire corporations and organizations for the purpose of efficient delivery of direct marketing mail pieces. It is an ongoing process that produces perpetual financial benefits of waste avoidance and customer recovery revenue.
The articles cite examples from a case study of an actual organization's research, implementation, organizational impact, financial summary and results of database maintenance planning. It is a retailer, direct marketer, cataloger, publisher, wholesaler and sponsor of conferences and events. It has three different transactional system databases along with several other stand-alone databases that combine for a total of approximately 3.3 million consumer customers.
House-Breaking the Monster?
The previous articles on database maintenance planning, described what the Database Maintenance Plan (DMP) does and how critical research about the organization-wide database(s) and mailing are assembled to justify its implementation. This article will expedite translation of the research into the business planning and explanations of the maintenance processes, their flow and implementation.
The monster is the DMP. To most multi-business corporations, integration of any major business activity or anything that attempts to change a once successful way of doing business, will be seen as a monster. This is the major reason why the business planning approach of organization-wide database maintenance is critical. House-breaking, is the flow, explanations and business planning of the database maintenance and customer recovery processes, as well as the implementation of the DMP.
After all research has been gathered regarding organization-wide mailing and the database sample(s), the translation of this information into a business plan that proposes DMP implementation begins. (Research results from the case study's database maintenance processes are located in Figure 1.) SEE FIGURE 1 The DMP should be able to communicate the plan to all areas of the organization. Therefore, for its justification, it must be able to propose itself to different types of people throughout the organization in their corporate language.
The DMP should first document what the estimated database maintenance results were from the customer base sample; then translate these results into what the financial benefit will be to the organization, if implemented. Figure 2 SEE FIGURE 2 is an example of how the case study illustrated its sample's database maintenance results into a financial benefit to upper management. The DMP conservatively used only 75% of the previous year's mail volume to estimate what the first full year of its implementation would accrue, and from this point on acts on the assumption that the sample analyses results are representative of the entire customer base.
Database Maintenance Processes
Database maintenance is best described as two different processes: 1) cleaning and maintenance and 2) customer recovery. Since the DMP measures financial impact by waste avoidance and customer recovery, both need to be separated and presented as two different database maintenance processes to the organization. (See Figure 3.) SEE FIGURE 3
The cleaning and maintenance process is the running of the database(s) through actual data hygiene processes. These are done by outside vendors that receive an organization's customer and address files; then run these against the most recent correct customer by address files provided to them through several sources. In this process, the DMP is measuring the customers who are currently being mailed to, and are not receiving their mail pieces from the organization. In other words, analyses of how much waste that will be avoided.
The customer recovery process is most of the same cleaning steps, however, in this process, the DMP is measuring the customers who are not currently receiving their mail piece, but will begin to receive their mail piece, now that an address correction has been made. In other words, analyses of how many customers that will be recovered.
Therefore, implementing a DMP consists of two database maintenance processes that are working simultaneously. The database maintenance processes and other related acronyms are listed in a glossary in Figure 5. SEE FIGURE 4 & 5
Cleaning and Maintenance Processes, Flow and Implementation
Once the research has been translated to illustrate the financial benefit of the DMP, the explanation of the database maintenance processes, their flow and implementation should be communicated to upper management. Figure 4 is a spreadsheet example used by the case study to explain the database maintenance processes and their flow.
The following explanation of the cleaning and database maintenance processes from the case study references the top flow chart (Figure 3) and the top half of the spreadsheet
(Figure 4) to help better understand the cleaning processes and the order in which they take place.
1. The database (2.4M individual names and addresses) presently consists of 48% deficient addresses. Twenty-one percent of the customer base is non-deliverable, and another 27% is being mailed inefficiently.
2. When the customer base was run through standardization and verification processes (Group 1/in-house), 42% of it was corrected. At this point, 93% (2,232,000) of the customer base are deliverable to the address. Seven percent of the customer base cannot be corrected by Group 1 or other United States Post Office (USPS) licensed software.
3. The customer base (2,232,000) was run through COA and LACS to calculate the extent to which it contains changed addresses and convertible rural route style address information. It was found that 12% (267,840) registered a match of COA, and a 3% (66,960) match on Nixies. This means that at least 15% of the customer base mailings are not being delivered by the USPS. Since the majority of the case study's mailings are second or third class, these mail pieces do not get forwarded to it's customers. This means that approximately 79% of the customer base (to this point) is considered reliable enough to deliver the mail pieces to the intended customers.
4. After COA, the customer base (now 1,897,200) went through a merge/purge process (de- dupe) that identified a 2% (37,944) hit of duplicates. Since a duplicate means that multiple mail pieces are being sent to the same customer, now the present customer base (to this point) is only 77% reliable enough to deliver only one mail piece to the intended customers.
5. The last step in the cleaning process showed the results of the house cleaning. This process showed that the customer base (now 1,859,256) is comprised of 21% (390,444) multiple customers at the same address. This now means that if the mailer wants only one mail piece per household, to go to the intended customer, then only 61% of the present customer base is reliable enough to do this.
The percentages of deficient addresses in the customer base do not subtract evenly because: 1) hygiene steps are derived from the completing of a previous step in the filtration process, 2) percentages are subtracted from the remaining percentage instead of the entire database percentage, and 3) to reduce the margin of error from duplicating the counting of the same customer addresses.
Address Element Correction
The customer recovery processes begin with an organization determining whether to send the customer base through Address Element Correction (AEC). AEC is a service provided by the USPS that corrects the address elements from the records that in-house software cannot. The customer address records that AEC corrects, will then be placed back into the customer base to go through the remaining database maintenance processes.
If in-house postal software exists within the organization, run the customer base through it first to determine which customer address records will pass through standardization and verification. Figure 3 illustrates that standardization and verification is the first process that a customer base goes through during cleaning and maintenance, and shows the flow of customer recovery analyses also starts with AEC. Separate those records that passed from the rejects; then place the rejects on tape or disk to be sent to the USPS for AEC. The USPS will explain the acceptable format and other data extracting specifics. This is important because your customer address records that did not pass through standardization and verification, may be able to be corrected by AEC.
If in-house postal software does not exist within the organization, send it to the vendor of choice and have them send the rejects from standardization and verification to the USPS.
The customer address records that are correctable, should then be sent back to the vendor and added back to the customer base. Use judgement to determine whether the rejects from standardization and verification are significant enough (number of) to send to AEC at this time. Time may be more of an issue than a few corrected records. A good time frame estimate for sending and receiving data back from AEC is 30 days, depending upon the total number of rejects.
Customer Recovery Processes, Flow and Implementation
The following explanation of the customer recovery processes from the case study references the bottom flow chart (Figure 3) and the bottom half of the spreadsheet (Figure 4) to help better understand the cleaning processes and the order in which they take place.
1. Seven percent of the case study's customer base addresses cannot be corrected by standardization and verification (Group 1/in-house). This 7% can be sent to the USPS for AEC. Seven percent of the customer base (2,400,000 customers) is 168,000 customers. The USPS standards estimate that 50% of the deficient addresses will be corrected by AEC. This means that out of 168,000 deficient customer addresses, approximately 84,000 of them will be corrected.
2. The corrected data will then run through Group 1 again (other 93%), for standardization and verification; then sent to the vendor for COA. Approximately 2,316,000 individual names and addresses from constituent will be sent to the vendor for COA.
3. It was also found that approximately 12% of the case study's customer base is comprised of COA hits. This means that approximately 277,920 customers have moved and are not receiving the mailings. However, COA also shows that approximately 95% of the 277,920 are recoverable because they have left a forwarding address. This means that a total of approximately 264,024 of the case study's established customers are recoverable by AEC and COA.
4. From the sample research, it is assumed that approximately 2% (5,281) of these recovered customers will be duplicates, therefore the estimated number of recovered customers is 258,743.
The final piece to customer recovery is a service provided by the USPS called Address Change Service (ACS). This service would perpetually return mail pieces back to an organization when a customer has moved, and returned with their forwarded or corrected address, if available.
Conclusion
The case study has presented the DMP in several different corporate languages. It used flow charts, financial benefits and technical explanations. This is important because each business and organization is different. There are different aspects of corporate culture to deal with in all of them. Because of the many different types of individuals that are throughout an organization, the DMP must be a document that can speak the language of any type of individual who needs to understand the purpose of it. Only through the database maintenance planning approach, can a multi-business organization integrate its database hygiene efforts.
The next article will focus on the scheduling of DMP implementation, financial forecasts and perpetually scheduled database hygiene updates.
Ken Culpepper is president of Integrated Marketing Solutions, Inc., a
knowledge-base marketing firm that
integrates tactical marketing strategy with management of multiple contacts
of businesses to their customers, prospects and channel customers. IMS incorporates
marketing business planning, long-term corporate ROI strategies, and marries
knowledge-based marketing with e-commerce strategy and systems. IMS has
offices in Atlanta, GA (770) 390-9199 and Nashville, TN (615) 782-0461.
(Web: migmar.com/ims)