Keep it Clean: Address Standardization, Hygiene and Maintenance for Business Marketers

Business data is said to degrade at the rate of 3-5 percent per month. Whew. This means a third of your information on business buyers and prospects may be useless by the end of every year.

Consider these vivid examples from D&B about the volatility of U.S. business data:

  • A new business opens every minute
  • A new business files for bankruptcy every 8 minutes
  • A business closes every 3 minutes
  • A CEO changes every minute
  • A company name change occurs every 2 minutes

The business case for data hygiene is clear. Here are some annual percentages by data element, also from D&B:

  • In a year, 20.7% of the business postal addresses on your file will have changed. If your customer is a new business, the rate is higher, at 27.3%.
  • Phone numbers change at the rate of 18% overall, and 22.7 % among new businesses. No wonder your sales force is always complaining that your data is no good (although they probably use more colorful words, not suited for a professional publication like DIRECT).
  • Even the company name is unstable, changing at the rate of 12.4% among all companies and a whopping 36.4% among new businesses.

But it’s not only about data churn, it’s also about the complexity of the business customer record itself. Leaving aside important database elements like purchase history and promotion history, let’s just consider the contact information. In a typical B-to-B database, you need information about the company, the address elements like postal address, phone and fax, as well as the “firmographics,” like industry and size. But many companies have multiple places of business, known as “sites,” and you need to contact multiple people — buyers, influencers, product specifiers, purchasing agents, purchase decision makers — at each site. Now you see why B-to-B database administrators tend to be a very patient lot.

This goes way beyond “four-line,” which is our traditional mail-order view of a business address. In fact, the post office recognizes up to 8 lines, including the mailstop, the contact name, lines for two types of titles, company name, department name, and the regular 2-line address.

So it’s complicated, but it’s also a joy, for those who find the complexity a welcome challenge. Among those who take pleasure in wrestling with business data are David Knutson, customer database manager at Uline, a shipping supplies company outside Chicago, Bernice Grossman, President of the B-to-B database marketing consultancy, DMRS Group, and — I admit — myself.

The three of us decided to convene a workshop at the DMA’s catalog conference in Orlando in May, where interested parties could share ideas and experiences on B-to-B data issues. We were joined by more than 30 people, representing not only catalogs, but also database service companies, list brokers and multi-channel marketing companies. The group divided into four break-out sessions, addressing the following thorny topics:

  1. Address standardization and formatting
  2. File cleaning and updating
  3. Company classification and buyer role identification
  4. New customer data collection and verification, online and offline

With thanks to the Orlando workshop attendees for their energy and enthusiasm, I am delighted to report on the ideas generated, and to announce our intention to conduct a similar working session at the DMA Annual convention in Atlanta in October.

How to Standardize and Format Business Addresses

  • Refer to Publication 28 from the USPS for examples of field selection and layout of business addresses. It is available for free download as a pdf, from usps.gov. Search on Pub 28, and you’ll go right to it.
  • Find the Input Editing Standards document that was developed when your system was first created. It’s probably buried in a file somewhere in your office. Use the standards to train personnel who do key-entry. Update the document as needed, and keep it current.
  • Create a Standard Operating Procedure for data-entry, and train your key-entry personnel on it.
  • Add an open-ended field for “delivery instructions” to keep these instructions from getting into your name and address fields.
    • But remember: Open-ended fields are not to be used for selection or for later marketing efforts. They are only useful if someone is going to actually read and act on the content.
    • Typically these fields exist on the customer master, attached to the ship to record, if they are to be used for every order; or as a special message, within the order record, for one time use.
  • Include 3 types of address options for a customer: Ship to, Bill to, and Mail to.
  • Good address standardization and consistent formatting helps reduce duplicate accounts, allowing both internal and external dupe identification programs to work more effectively.

File Cleaning and Updating Techniques

  • Use data analytics to create exception reporting. Then have your less-busy call center shifts review the exceptions and clean them up, either by common sense or through outbound contact to verify.
  • Make sure your salespeople are compensated correctly. If they are commissioned on new orders key-entered, you are likely to end up with a lot of duplicate records.
  • Pay for postal returns (endorsements) and update your files with the returns. This should be done at least once a year, and ideally 2-4 times a year.
  • Learn the pros and cons of the different endorsement options and how they operate within both Standard and First Class Mail:
    • Address Change Service (ACS)
    • Address Service Requested,
    • Return Service Requested
    • Change Service Requested
    • Forwarding Service Requested
  • www.usps.com has a wealth of good information and tips on addressing options.

Company Classification and Identifying Buyer Roles

  • Don’t bother asking customers their SIC or NAICS. They don’t know it. Other options:
    • Provide an industry level classification option, such as Manufacturing, Distribution, Retail, etc.
    • Ask for a more complete description in an open-ended field.
  • If you really care about SIC, set up data appending with D&B or InfoUSA.
  • Number of employees can be just as important as SIC, and is often easier to obtain, especially if you provide check off boxes and ranges.
  • Buyer role is very subjective.
    • Use drop-down menus to gather buyer role info AND include an open-ended field so they can write in if their role does not appear.
    • Keep in mind that open-ended fields will not allow the data within to be used for selection, but they can be used to spot higher frequency titles, which you can then add to the functional drop-down menu.
  • Use separate fields for actual title and functional title. While this approachs is ideal, most companies have just one field allotted.
  • Invite Internet customers to self-classify whether they are business, consumer or government, so your name and address screens can be tailored to make sense.

How to Collect and Verify New Customer Data, Online and Offline

  • Ask for customer data more than once, and make it very easy for the customer to provide the information. Always ask for the information the same way — no matter where you ask for it.
  • Train new reps on your input standards. Give all reps refresher courses every 90 days. Send out frequent reminders and updates to the reps.
  • If your variable costs (COGS) is low, e.g., for software products, ship the product before verifying the address.
  • If you need to verify before shipping, use a service such as QAS, or Group 1, to identify suspect addresses and send them to customer service for review. Ideally, this process can be set up to be interactive so the customer service rep can make fixes on the spot.
  • Consider conducting a human review of all Internet-based orders.
  • Confirm email address by sending an order confirmation or a newsletter. If it bounces, then verify the address by outbound phone.
  • Improve email capture rates by requesting that the email be typed in twice.
  • QAS is offering a solution that lets the customer review a suspect address in real-time. It can either pop up a message saying something like “please look over your address and make sure it’s correct,” or it can identify the actual problem and offer valid alternate choices
  • Use input edits to either prompt the customer on problems, or to kick out orders for manual review.
  • Here are some characteristics that make an address suspect:
    • Zip does not agree with city/state
    • Address does not exist with Zip
    • Missing/wrong directional on the street
    • Building number does not exist within the street

If you’ve gotten this far in your reading, we declare you more than qualified to participate in our next workshop in Atlanta. Please join us.

Back to Articles and Columns »