World-class direct marketing doesn’t happen by accident or overnight. Companies known for their excellence in one area of business – think L.L. Bean and customer service, Amazon.com and ease of use, The Sharper Image and merchandising strategy – spend years and buckets of money developing processes and technologies to advance their performance levels.

Of course, a large company can afford the investment high-level operations require. While medium- and small-size companies don’t have deep enough pockets to compete with large companies inch for inch, they are not barred from tailoring best practices to fit the size and scope of their organizations. Every company can assess its performance levels and take steps to move them forward.

The best practices that apply to direct marketing could fill a book, so this article takes a closer look at a few of the key operations that drive direct marketing performance: database hygiene, Web-supported customer service and predictive analysis.

BEST PRACTICES: DATABASE HYGIENE

The customer database is the source of your direct marketing strategy. To perform meaningful analysis (see the section in this article on predictive analysis), you must have faith in the accuracy of your data.

According to John Miglautsch, proprietor of Miglautsch Marketing, a database marketing consulting firm in Hartland, WI, three main principles guide sound database hygiene practices:

• Enter clean data;
• Keep reviewing your database for inaccuracies and clean them up accordingly; and
• Use your data to find out if it’s up-to-date.

Putting a finer point on these principles, Miglautsch shares his eight best practices for database hygiene.

1. People responsible for data entry must care about the accuracy of the data.

When the acronym DM meant direct mail, marketers too often employed $5 an hour temps who didn’t care, resulting in lots of sloppy data entry. Data quality did not improve much when the industry converted to real-time data entry via telephone operators keying in orders and handling customer service issues.

Now with the Internet, customers key in their own data. It’s not their job to care about data quality, they make things up, and they also can’t type.

Consider emphasizing the importance of accurate information to your Web customers. Let them know that their order cannot be processed unless their personal information is accurate. Show a confirmation page, and send confirmation e-mails at both order entry and shipment.

2. Correct addresses online.

Too many marketers are only using an outside service bureau to clean their list, checking the address quality weeks after the customer interaction. Getting the name and address information correction on the initial contact is far less costly than fixing it later.

Several inexpensive address-checking software options exist that can be integrated into your Web system. There is no excuse for capturing a non-mailable address.

3. Match customers online.

Many popular Web-ordering systems simply create a new customer identity for each and every order, leaving the job of linking and merging past customer history to the “main” system.

Consider using a cookie to identify customers revisiting your Web site. Display the user name prominently (“If you’re not John Miglautsch…”), so customers can see if they match their own previous identity.

4. Connect your order system and your Web site.

Make sure your data is consistent between your Web site and order processing system – part numbers, customer numbers, even processing dates have to be integrated. Too often, Web systems are cobbled together only well enough to feed data to the main system. Customers have no option to select or edit their personal information.

5. Create a club or login.

Help your channel-surfing customers reconnect with their own identity. Give them a reason – special offers, lower shipping, etc. – to register on your site. This makes connecting their Web identity to their main identity much simpler. Again, consider populating the site’s identity section with information from the main system so customers can see whether the stored information is correct.

6. Match against internal lists.

Don’t delete housefile names, even if they ask you never to mail them again. You need to keep them to use as an internal suppression file for those instances when one of them makes a purchase from your company or signs up for your e-mail newsletter: (I’m not suggesting that you shouldn’t mail them, but you need to flag their schizophrenia.) Keep the addresses of your fraud customers and match against them – no sense sending them more opportunities. Ask for e-mail addresses “in case there is a problem with your order.” E-mail is the least expensive way to tickle people into reconfirming their information and keeping your data clean and up-to-date.

7. Match outside lists and databases.

Ask your list broker about co-op databases in which hundreds of marketers pool their active buyers. It’s especially helpful in sifting through your old, inactive customers. They may not be buying from you, but they may be buying from someone – and you’ll know they’re not deceased.

If you rent a significant amount of names, you can create your own co-op database by flagging your house names that match outside lists. It’s considered unethical to note on which outside lists your housefile names appear, but it is not unethical to simply note that your individual records match an outside list.

There are a number of lists available solely for the purpose of keeping your list clean. The U.S. Postal Service’s National Change of Address (NCOA) file helps you keep up with your customers on the move. The Direct Marketing Association offers a list of people who do not want to be mailed (Mail Preference Service). There also are co-op fraud lists. And some marketers suppress prison addresses.

8. Use your data!

Pristine data combined with endless analysis will not yield anything of value. The best way to keep your data up-to-date is to generate active response and purchases.

Growth and profit is in the changing of your offer, not in cutting the waste from your list. Few companies shrink to greatness. Unless you can translate your data into variables into analysis into new offers, your company will not generate ROI on clean data. …Profitable customer contact pays for cleaning the data. The more money you make, the more often you can interact with your customers.

BEST PRACTICES: PREDICTIVE ANALYSIS

While analysis of individual direct marketing campaigns helps point you in the right direction for future efforts, nothing beats ongoing data reporting to uncover trends in key statistics, say Rhonda Drake and Perry Drake, proprietors of direct marketing consultancy Drake Direct in New York.

This type of analysis “allows managers to get a full picture of the health of the business, and change messaging and offer strategy to encourage growth,” they say.

Two primary best practices in developing trend reports lead to more specific criteria for monitoring customer trends. To predict your future direct marketing steps, you need to:

1. Define homogeneous groups for reporting.

Look at a variety of segments to uncover a shift in trends when one arises, say the Drakes. Some marketers segment based on RFM (recency, frequency, monetary value), and others on major affinity groups. In looking at your customer data based on a single segmentation strategy, however, you may miss a trend shift that could be identified if the data was segmented another way.

Drake Direct advises marketers to consider these additional segmentations:

• Demographic lifestage. Depending on the business, consumers migrating through lifestages may provide important communications opportunities for marketing. Segmenting current customers by lifestage and noting key events can be important for sizing up market opportunities.
  Key events to track: purchasing a new home; moving; marriages; births; and retirement.
  
  
• Marketing cohorts. Just as a consumer’s lifestage is important, his lifestage with a particular company or product group also is significant. Understanding the ebb and flow of the internal cancel rate of a business can be done by trending cohort groups (groups of customers acquired at the same time). In addition to grouping customers by their time of acquisition, the specifics of their initial response can provide a method of sub-segmenting.
  
  For example, customer behavior and quality varies by source. Sub-segmenting cohorts by their media source can provide further behavioral profiles.
  
  Key sources to segment: telemarketing; direct mail; Internet; broadcast; and space ads.
  
  Beyond media consumption, customer behavior is another segmentation variable more analogous among groups responding to similar offers.
  
  Key offers influencing behavior include: payment terms (cash with order vs. free trial, etc.); premium with order; sweepstakes or contest; shipping terms; price difference.
  
  
• Customer activity. Other segments to study include any that are used in the normal course of operation. Among them: product line; affinity; recency of activity; monetary segments; gift givers; lapsed customers; proprietary segmentation techniques.
  
  

2. Define key statistics for reporting.

In addition to defining which groups you’ll want to monitor, you’ll need to determine which key measurements to take. These measurements will be used to compare campaigns and business decisions over time.

According to Drake Direct: “These measurements will provide a means of highlighting improvements or degradations in customer performance, growth or shrinkage in desirable customer groups, and a means of comparing new sources of customers to existing channels.”

Key measurements to monitor include:

Orders

• average number of orders
• average dollar value
• number of orders in past 30, 60 and 90 days
• dollar amount of orders in past 30, 60 and 90 days
  
  Payment
• average paid per customer
• ratio or payments to order
• payments in past 30, 60 and 90 days
• proportion of payments on credit card
  
  Returns
• average returns per customer
• ratio of returns per customer
• returns in past 30, 60 and 90 days
  
  Activity Status
• time in days since last communication
• time in days since last order
• time in days since last payment
  
  Response
• ratio of orders to offers for cross-sell/retention
  
  Customer Service Inquiries
• questions
• change of address
• billing complaints
  
  

BEST PRACTICES: WEB-SUPPORTED CUSTOMER SERVICE

While the term CRM (customer relationship management) has fallen out of vogue as of late, the new era of customer-focused business continues on.

A lackluster economic environment has forced some companies to re-evaluate the ways they serve customers. Enter the Web, which has become an efficient channel for many marketing activities, in addition to customer service.

It’s true that the Web is a versatile tool for direct marketers, but it’s just a tool. Customer service programs can fail on the Web just as easily as they do in other channels.

Based on trial and error with more than a thousand clients, Greg Gianforte, CEO and founder of RightNow Technologies, a provider of customer service solutions located in Bozeman, MT, has developed a list of best practices for customer service on the Web. We’ve provided the top six best practices here; for the rest, visit www.targetmarketing-mag.com, and look for the section titled “Web Exclusive.”

1. Customer service representatives are also content drivers.

Your customer service representatives (CSRs) do more than answer phones and e-mail queries. They are the first to encounter customer service issues that impact your company’s reputation with customers and prospects. Often, a service issue with one customer becomes a pattern. Without the appropriate response processes in place, each successive query has to be handled as if it’s the first occurrence.

A more effective solution is to empower CSRs to author new knowledge-base content for the Web site when they handle new service events. (A knowledge base is a system of organizing queries and answers so that each input can be used to answer future similar queries without the need of human interaction.) Beyond improving self-service rates, it allocates CSRs’ time for queries that require personalized attention.

One common mistake: The knowledge-base items on the Web are not available to CSRs. Without this access, CSRs are not able to give the same, up-to-date answers to call-in customers that self-service customers receive.

2. Continuous improvement.

Once you’ve achieved your return on investment in your Web-enabled customer service program, resist the temptation to become complacent. For example, an initial self-service rate of 60 percent to 70 percent may seem satisfactory compared to the reduction in need for one-on-one time with live CSRs.

But RightNow Technologies has witnessed companies committed to continuous improvement of their online customer service program achieve 85-percent to 95-percent self-service rates.

And if you build in feedback mechanisms, the Web channel can provide you with customer comments about site content and traffic statistics to further improve your service levels.

3. Minimize clicks.

Your site should be designed so customers can access knowledge-base items as fast as possible. Clearly identify your customer service-related links on your home page – typically with buttons labeled “Customer Service,” “Need Help?” or something similar.

This main link should lead to a page of the top 10 to 20 customer service issues. RightNow Technologies has found that more than 50 percent of customer service questions are solved by this top issues page.

4. Customer service portals.

When service-related information is located in different areas of the Web site – product information on one screen, shipping information two links away – it becomes a hindrance to customers finding the information they need quickly. Once customers initiate a search for an answer within a service-related content area, says Gianforte, they should not have to leave it.

Instead of moving content from its current home, it’s best to create a customer service portal that links to all other content customers may need to answer their questions.

5. Empower your customers.

A large percentage of customers actually are sitting in front of their PCs when they call you on the phone, Gianforte notes. Rather than make them wait in a queue to receive an answer that’s already posted in your customer service area, use your “hold message” to tell customers about the online customer service area.

While customers are in the queue, they can look up the appropriate Web page and possibly answer their own question. Those who choose to wait for the CSR will still be helped, and it gives CSRs an opportunity to reinforce the self-service message by showing customers how to find the information they need faster.

6. One contact center database.

To provide comprehensive customer care, your representatives need to view all customer incidents from all channels in a common database. An integrated customer service database allows CSRs to be fully aware of all previous interactions with a customer relating to a particular incident. This multichannel view, says Gianforte, ensures that CSRs have all the information needed to quickly resolve problems that may already have involved several different communication sessions.

Also, says Gianforte, an integrated customer service database provides managers with a holistic view of the issues driving customer service activity across all channels. Besides being useful to customer service managers in adjusting customer service policies, it provides insight to marketing managers and sales managers, who also must be responsive to customer needs.