Lists: Double Vision
Perhaps you are processing a merge/purge for an upcoming mailer. At the same time, you have an ongoing email prospecting campaign and telemarketing reactivation program running. Maybe it would be most effective to coordinate all three campaigns to avoid duplication of effort and even possible damage to your business reputation. Ever wonder how people view inconsistent or overlapping messages? I'd expect not very well.
For many marketers, this is a familiar story. Direct mail, email messaging and telemarketing each have a legitimate place in your marketing plans. Sooner or later, you find yourself searching for strategies to merge duplicates between online and offline data sources.
Searching for Duplicates
The first step is recognizing that duplicates must be linked between online and offline data. What does this mean? Finding email lists that also contain postal addresses or phone numbers can provide the common contact data points that allow duplicate identification. While not possible with every list, the trick is selecting data with a high population of email, telephone and postal information.
Apply the same rules to your internal customer and prospecting lists. Don't skimp on your Web forms, surveys, data acquisitions and exchanges, or even on simple key punching, by excluding email address, phone number or postal address fields from them. To complete internal files missing key components, search out a quality data provider who can append the missing information. As much as appending email addresses and phone numbers is a common practice, so is reverse appending the postal address based on the same email address or phone number.
After your lists are selected, running a merge/purge using only basic postal address match logic may not yield adequate results. A similar mistake would be using just an email address or a phone number to identify duplicates.
A better approach is to search for duplicates on any combination of the contact name plus email address, phone number or postal address. This way, any two email address matches with different postal addresses are merged. Another example duplicate match would be to match any records with the same contact name and phone number, but with different email addresses. Using this logic, each of the records in the chart (see the mediaplayer at right) are merged.
Choosing the Best Records
This raises some questions. When data are merged and different email addresses, phone numbers or postal addresses are found for the same person, which data should be retained?
- Should lists be prioritized based on costs?
- Could some lists have better email addresses while others have higher quality postal addresses?
- Does the last update date for the list, or even the record, make any difference?
- Should data quality filters be applied to postal addresses, email addresses and phone numbers?
The answers are all "yes." Setting up business rules for data retention involves many considerations that require careful planning—list costs, historical ROI, update frequencies, data compilation techniques, data maintenance, NCOA, eCOA, phone number verification and others. Due diligence in researching each data source with list brokers, managers, compilers and owners can pay large dividends.
Also, the converse is true and could provide equally important insight into data quality. If more than one data source provides the same email address, phone number or postal address, there is an increasing likelihood the data is valid. For readers who are merge/purge experts, this is an extension of calculating multi-buyers. Instead of counting the number of times an individual is sourced, count duplicate email addresses, phone numbers or postal addresses from different sources. Naturally, more is better.
Merge Before ... And After!
Removing duplicates, prioritizing lists, applying keycodes and validating data are all common merge/purge practices to carry out before the mailing. As the results come in, some marketers may turn their attention toward the next campaign. Others tend results by tabulating responses by keycodes, URLs or another response tracking. Smart marketers apply the same merge/purge strategies used before the campaign to process the data collected after campaign execution.
Campaign responses may come in the form of website visits, inbound telephone calls or email messages. Communication may cover the range of product inquiries, survey results, sales or even simple questions. Might capturing and merging these responses be beneficial? Could combining these campaign results with the output of your last merge/purge provide valuable insight? Once response data is merged with the output of your last merge/purge, a very linear, cause-and-effect analysis quickly calculates the response ratio per list source. Taking this a step further and using the strategy over time, it becomes possible to calculate response trends per list source, estimate the number of mailings required to generate the first response, and establish the lifetime value of customers.
Merging after a campaign is very similar to merging before. Whenever possible, capture the full contact name, email address, phone number and postal address for both completed sales and inquiries. Consider each as a full database activity. These responses may include a variety of data points such as a time stamp, general classification or categorization, a transaction amount, and an activity ID number. By merging each response with the output of your last merge/purge, you'll have access to a wealth of information.
Whether you are working through the merge/purge process weekly, monthly or a couple of times a year, the information learned need not be lost. For forward thinkers, using all of the above merge/purge strategies to build and maintain a private database may offer additional long-term insights. Not only can a private database provide a platform for evaluating list overlap and data quality, the number of new records per list can be calculated, response and sales data can be merged to produce accurate ROI calculations, and additional demographic or behavioral data may be appended.
Implementing a private database can also improve merge/purge efficiencies. Consider the possibilities if each merge weren't a one-off project. Files could be processed in real-time as they are received, eliminating the hurry-up-and-wait nature of merge/purges. Would you need to maintain separate internal customer, prospecting or other suppression lists? Probably not, because these could be maintained as part of the private database. And, reaching back to our opening scenario, coordination of upcoming mailers, email campaigns and telemarketing efforts is greatly simplified.
Do your research—this can't be emphasized enough. Make sure your merge/purge technology is up to the task. Take time to understand the strengths and weaknesses of each data source. Build a list prioritization and data retention plan. Carefully track and merge responses after each campaign. And, if a private database is in your future, look for the software interface, tools and back-end processing that best meet your long-term needs.
Eric Smith is the founder of database technologies company ListFusion and lead software engineer of DataTree online database software. He can be reached at email@example.com.