Sponsored Content WEBINAR: ELECTION 2008: E-mail Tips from a Political Insider   WEBINAR: Becoming the Illusionist: The Magic of Using "Brand" to Drive Response without Results Disappearing in the Process   WEBINAR: Score and Win Through Cooperation   TOOLS OF THE TRADE: Holiday's are prime-time for email marketing - maximize your ROI by getting to the inbox   ON DEMAND WEBINAR--MASTER CLASS: High Performance Online and Off-line Strategies for Large Scale Customer Acquisition   FREE WHITE PAPER - Direct Mail and the Internet Find any supplier you need - agencies, CRM, fulfillment, lists, e-commerce, paper, printers, telemarketing, and more. Featured Categories Lists and Data Telemarketing Database Marketing E-commerce Web Marketing Agency & Creative Services Print, Production & Paper Lists and Data Processing :: view all categories Get free access to more than 50,000 list data cards - one of the most comprehensive databases in the industry. >> Search Now DIRECT Buzz Blog Webinars Buyer's Guide Calculators JobZone ListFinder Promo Calendar Research Store E-Newsletters Print Magazine This Month in Direct Magazine Deal With It Direct had a full house for this year's list roundtable. Considering all the additional responsibilities on brokers' plates, that's impressive... See Full July Issue

DATA — WHERE TO GET it and how to maintain it — was the topic of Direct's 2006 business-to-business roundtable.

With the help of our columnist Ruth P. Stevens, Direct conducted a virtual roundtable last month with three B-to-B data pros: Bernice Grossman, president of DMRS Group; Chris Purcell, BusinessWeek's director of new business acquisition; and Thomas C. Tweedie, director of consumer direct marketing and forecasting for Day-Timers Inc.

Stevens consults on customer acquisition and retention, and teaches marketing in the graduate program at Columbia Business School.

Grossman founded marketing and database consultancy DMRS in 1983. Clients have included Avis, Coca-Cola and Microsoft.

Purcell oversees strategic management of BusinessWeek's North American, Canadian and Latin American direct-to-publisher acquisition programs.

Tweedie joined Day-Timers in 2002 and is responsible for the marketing of Day-Timers' direct business.

STEVENS: Everything in direct marketing stems from customer information and a good, clean, accessible marketing database. But in business-to-business, there are a lot of challenges in getting and keeping customer data. What are the chief sources of customer information to populate your database?

TWEEDIE: Data is collected at every contact point with our customers. Phone agents verify all address, phone and e-mail information on every call. E-mails are solicited and used to send order and shipment confirmations. We also ask for phone number and e-mail address on all mail pieces.

PURCELL: Information is collected directly from our customers via surveys, Web registrations and third-party overlays.

GROSSMAN: Customer data flows from the direct mail channel, Web sites, fax sites, and from in-bound phone and IVR as well as from field sales, trade shows and sponsored events. Customer data our clients collect includes ID number, title, name, company name, street address, city, state, ZIP, original source of entry, brand data, transaction data, demographic data, shipping data and billing data.

STEVENS: A lot of business marketers sell through independent third parties, like resellers and distributors. In such situations, it can be hard to get detailed information about end-user purchasing behavior. If this is part of your distribution picture, how do you handle the data implications?

TWEEDIE: We really do not have the ability to collect the end-user information. We do have our resellers' transaction data.

GROSSMAN: This is a major problem in B-to-B. For those products sold through distribution/wholesaler channels, the manufacturer doesn't actually know the end user. And it's been tried, particularly in the food service industry. Methods used to find this out include various types of contests where the end user's name and address must be submitted to the manufacturer via the distributor/wholesaler, usually on the invoice. This data can then be routed to the manufacturer's customer marketing database and tagged with the distributor/wholesaler info.

STEVENS: Where do you house your customer information? What platform, what interface? Is it managed in house or at a vendor?

TWEEDIE: We store all our customer information in house. We have customer-level data back to 1981 and transactional data back to 1990 for both the United States and Canada. The data is stored on a mainframe and brought down daily to SQL server databases where queries and analyses are done. Selections are made at the server level and then IT pulls the records to tape or for FTP to outside list houses and lettershops.

PURCELL: Our data is stored with a third-party data manager and we use a secure Web-based access system.

GROSSMAN: Most of our clients outsource their marketing database to one of several vendors. Some of the platforms used include E.piphany, Unica, Alterian, MarketZone Platinum and proprietary solutions.

STEVENS: Do you append third-party data? What elements, and where do you get them? Any thoughts on how this data adds value?

TWEEDIE: Day-Timers is unique in that our customer base is made up of consumer and business addresses. The data needed or used for each of these segments is different. We overlay our file every few years and then update new consumers since the last overlay. Consumer catalog buying habits and product categories bought as well as household data are used for campaign modeling and possible new product development. For our business addresses, SIC, [number of] employees and job titles are used for marketing, product segmentation and product development. Data comes from sources including Abacus, Direct Media and Next Action; and on the business side from Experian and other public and private prospecting databases.

PURCELL: We append title, company, SIC code, sales and revenue, and number of employees to help fill in gaps in our file and gain a better understanding of our business-address readers.

GROSSMAN: Many of our clients have data appended to their files. Third-party data from D&B, InfoUSA, Harte-Hanks, MarketModels and Experian are some of the sources. Sometimes, proprietary and/or dedicated vertical sources like trade magazines and private files are used as well. Data appended is usually things like SIC, occasionally NAICS, number of employees, headquarter or site location, sales volume, public/private, number of years in business, small office/home office codes and technology information. Firmographics broadens the picture of the customer and helps in segmentation and modeling.

STEVENS: Who has access to your marketing database? Any lessons learned in this area?

TWEEDIE: Marketing and marketing operations ‘owns’ and controls access to the actual data and its use, while IT is in charge of maintaining database integrity. All requests to alter any data or field must be approved by marketing. Any request to add data, use or export data must be approved by marketing. As the owner, marketing also authorizes who is able to access the database and at what level. With a serious concern for customer privacy, access to customer-level data, including payment information, is strictly controlled.

PURCELL: For security purposes, we limit access to customer data. Web access is enabled via an online account system. The McGraw-Hill Companies takes customer privacy very seriously and we constantly strive to improve how we handle customer data and communicate those policies to our customers.

GROSSMAN: The client and sometimes the agency have access. Usually access is password-protected and is limited to the individual's need or role. It's best to protect the level and detail of the data made available — not everyone needs to have access to all of it to do their job. Privacy and security must be a paramount consideration.

STEVENS: How do IT and marketing intersect when it comes to data management? Do the two groups get along? Which is ‘in charge’ of the marketing database?

PURCELL: We outsource the IT function and work with a vendor that can provide marketing insight to our team. It's an extremely positive relationship and works from a partnership perspective. But ultimately, marketing is held responsible for the database.

GROSSMAN: IT and marketing must be helped to intersect and to work together. The data usually originally comes from IT and then a set of rules and feeds are developed to allow the data to move, on an agreed-upon frequency, to the vendor. Once the marketing database data is designed, developed and implemented then the ‘in charge’ folks are those in marketing. A database should never be built in a marketing vacuum — IT always should be part of the task force.

STEVENS: Now let's look at data uses. What are the top applications for customer info? (Campaign selection, marketing intelligence, product development, etc.)

PURCELL: All of the above.

TWEEDIE: Customer segmentation and modeling for campaigns, new product development and niche product marketing, life-cycle marketing, and modeling for outside list selection for prospecting.

GROSSMAN: Query counts, data mining, data browsing, reporting, campaign selection and management, analysis and marketing intelligence.

STEVENS: B-to-B data degrades so fast. How do you keep your data clean?

PURCELL: Regular updates are mandatory. You always must weigh the expense against the updated data's value.

TWEEDIE: NCOA is done three times a year on the U.S. file and then followed up with ACS on the next catalog mailing. We also use Melissa Data's Address Object to verify and standardize our customer address information. Address Object is a development tool used to verify addresses, in real-time or batch processes, by comparing the address submitted against the U.S. Postal Service database of 145 million deliverable addresses. Thomas Roth, Web development manager at Day-Timers, says using Address Object has improved our mailings and our customer communications because now we can count on our address information. Day-Timers mails both catalogs and direct mail reminders to its customers. It cleans about 7,000 to 10,000 records per month. Roth says the most common address errors he finds are invalid city/state and ZIP code combinations and incorrectly spelled street names. Day-Timers validates all new and updated customer information it receives via phone and online. We verify thousands of addresses per week in batch updates and in real time.

GROSSMAN: In terms of postal data most vendors use First Logic, Group One, Data Flux and/or a combination of these and sometimes some proprietary software. In terms of tracking the movement of people within an organization and from one job to the next, I've not seen anyone do it, no less do it well.

STEVENS: Are you doing any modeling? What kinds of statistical approaches are or aren't working?

TWEEDIE: For modeling, we use transaction data, ZIP code, bad debt indicator and calculated variables. We also use four-year buyer data.

PURCELL: We're doing modeling for internal purposes only, looking at characteristics that allow us to identify what products and services customers might need.

GROSSMAN: My clients are testing regressions and CHAIDs. But only those clients who have significant amounts of customer data are doing this — just name and address and some light firmographics isn't worth it. Significant transaction data along with firmographics are showing success.

STEVENS: What's the biggest challenge you face in maintaining good, clean customer data?

TWEEDIE: Managing duplicate records. With many more orders coming over the Web without key codes or customer numbers, it's an ongoing process to avoid creating duplicate accounts. Duplicates cost money to mail and anger customers who get too many mail pieces. We run a nightly merge/purge of all ‘default’ uncoded orders besides our monthly process. The nightly merge/purge corrects customers who were improperly assigned default key codes. We believe customers are going on our Web site and getting a new customer number, even though they might already have an existing number on the mainframe. Since there's no history for a new number, they get one of the default key codes unless they specifically enter a key code on the site. The nightly merge/purge takes the orders for the day and bounces all the customers who received one of the default codes against the mainframe file. If a match is found, a new code is assigned based on the customer's promotional history.

GROSSMAN: The business and data is dynamic, so I'd say trying to keep track of where an individual goes — new jobs and new names — is the biggest challenge.