Database Marketing - The Camel in the Dark
by George Duncan
After years of working with marketers building (or not building) databases, and having read numerous books and articles on the subject, it's clear to me today that defining database marketing is the classic example of a committee describing a camel by each touching a different part in the dark. Indeed, the authors of one hefty reference baldly state in their introduction, "There is no universally accepted definition of database marketing." So there.
Nevertheless, any attempt to employ direct marketing techniques absent an understanding of the role of the database is likely doomed to failure. For one thing, the very definition of direct marketing (faithfully inscribed in each monthly issue of Direct Marketing magazine) includes the word database:
"Direct Marketing is an interactive system of marketing that uses one or more advertising media to effect a measurable response and/or transaction at any location, with this activity stored on database."
That last phrase was added some years ago as computerized databases took customer information out of the shoe box and put it online. Remember the shoe box? Back when I was growing into a 38 short portly in New York, I was a regular customer at the late lamented men's clothing store, Rogers Peet & Company. The salesman there had a 3x5 card in a box with my name, phone# and size written on it. When the fall shipment came in, he called me (!) to alert me to the fact that he now had some 7 or 8 crisp new suits in my rather difficult size, and suggesting I stop by before they were sold. I went.
That box of 3x5 cards was the salesman's database. He had a record of when I was in last (recency), how often I shopped at RPs (frequency), and the price range of my last purchases (monetary). Recency, Frequency and Monetary (RFM) still forms the backbone of many database marketing systems -- although there is some back and forth among DB gurus about RFM's continued relevance and value amid today's more sophisticated statistical analysis methods.
More recently, I received a self-mailer from a business forms and stationary supplier from whom I had purchased some imprinted labels. "Dear Mr. Duncan," the card said. "It has been 18 months since you purchased your imprinted, pressure-sensitive labels (Item# 00043) and your supply may be running low. Just check the replacement quantity you prefer on the card below, detach, and drop it in the mail. Your new supply of labels will arrive within two weeks." Same thing as the suit salesman, except now the cards are on a hard disk and they can automatically spit out a personalized letter on a timely basis. Automated sales followup is just one use of a marketing database -- but it's a good one!
Some Fad!
In a posting to an online database marketing mailing list to which I subscribe, another subscriber wrote: "I see DB marketing becoming yet another fad only surviving through the proselytizations of soapbox opportunists eager to get in on the action of training and seminar circuits. What long-term value does it give us to know what colour underpants our customers wear or when their pet canary has its birthday. I guess my point is that we can't get carried away with gathering information just for the sake of gathering information."
Well, I agree with the last part -- as would most database architects, since gathering and storing information is time consuming and costly. But Victoria's Secret might love to know what colour underpants their customers wear, both for product research purposes, and in order to give frequent purchasers of colored undies a special offer next time they introduce a hot new flavor. And if I were Mr. Hartz Mountain (or Mr. Local Pet Shop), I'd surely make a note of your pet canary's birthday if I could capture that data. Can you imagine a birthday card arriving for your beloved canary suggesting you stop by and pick up an appropriate birdy gift? Tweet! Tweet!
As for the "fad" rap, tell that to the folks at Land Rover, Ltd., who used their marketing database to identify 4,000 Range Rover owners to invite to a special marketing event that eventually sold 1,000 new vehicles at $52,000 each. Yes, a $52 million winner (and a DMA Gold Mailbox award to boot) for an investment of about $150,000. Some fad!
Today's marketing case histories are filled with examples of successes similar to the Land Rover extravaganza for companies of all kinds pouring big bucks into their databases and into the marketing promotions they slice, dice, and spin off from the data -- from the airlines' highly popular "Frequent Flyer" programs to Sears' mail order database that helps the catalog and retail giant profitably target nearly two billion direct mail pieces a year to the company's 24 million shop-at-home customers.
But how about the little guy? The marketer who doesn't mail 2 billion pieces a year, but more like 2 to 20 thousand?
A Database For the Rest of Us
Now that we've touched a few of the humps, let's go back the front of the camel and see if we can identify what kind of animal this is. The diagram in Figure 1 is snapshot of a simple lead generating process using a database. If you're in direct product sales, think of your first sale as a lead, because if you depend for your profit on turning a one-time buyer into a regular customer, that's what it is.

Figure 1
In the Fig.1 graphic, your product/offer is seen at a trade show, a magazine ad, a direct mail piece, picked up through "word of mouth" or prompted by a news story on your company or product. (Add trade magazine bingo card, tv ad, the Internet, etc.).
You may wish to verify the inquiry, determine the level of qualification before entering it into the database, or passing it to the sales force.
Once entered, the database may pop out a label or a personalized letter to accompany the product, the special report, the demo, or whatever you must fulfill.
From there, the TSR/salesman or marketing folks access the database and follow up to continue the dialog (in long sales cycle products) or upsell the customer and enter the results of that contact in the database. OR, the lead may go directly to a sales team who acts on it and enters that result in the database.
In the case of a sale, the info goes to customer service for support or for whatever other reason they may need a record, and the info again is entered in the database.
By now you've no doubt noticed that whatever happens, the activity is entered into the database. (No-sales can be evaluated, coded, and entered in the database for a different kind of followup in say, six months or a year, depending.)
What It Is
If we can't define a database, we can at least identify its key functions. More accurately, we can describe the functions of a Database Management System (DBMS) which today is not a box of index cards, but a sophisticated database software program residing on a client/server platform, a minicomputer or a mainframe.
In their book, Business-to-Business Direct Marketing, World Class database consultants Tracy Emerick and Bernard Goldberg spell out the primary ways in which a database is used:
- Database provides names of customers or prospects.
The original name and address list is an important part of the database, but all of the activity relating to each of the names on the list is also essential.
- Database is a vehicle for storing and measuring responses.
Once you reach the targets you identified and they respond to your offer, the database should provide the ability to track each contact and each response.
- Database is a vehicle for storing and measuring purchases.
Now that you know if a prospect has responded to your direct marketing program, you want to track whether these respondents actually become buyers."The back end conversion of respondents to orders is the key measurement in evaluating the success of the direct marketing program," state authors Emerick and Goldberg. "By tracking the actual respondents who purchase, you can establish the cost per order and the average order size."
- Database is a vehicle for continuing communication to the prospects, respondents and customers.
The database should allow you to have a sustained and complete ability to contact the initial list of prospects, the group that responded, and the group that became customers.
In the business stationary example described earlier, the company had stored my name and address on their database, what I purchased and when, and how much I paid. They also coded me as a customer and programmed in the resell date. They input appropriate copy to their database and programmed it to print and mail the self-mailer offer for a new supply of labels on the preset date. Then they went to lunch.
The key point of this description is that, unlike some others I've seen that seem to relate mostly to mega-mailers, it applies to any mailer, regardless of size. The four functions outlined above will work equally well for my local bird seed retailer and the Sears mail order buyer file.
The Family Jewels
Basically, what you put in your database are the family jewels. All of your contacts and the results of those contacts with all of your customers. Finally, the power of data processing is bringing us back to the days of the guy with the box of 3x5 cards who could call me personally and tell me he had some suits in my size! Databasers Do It Digitally.
According to a survey conducted by DIRECT magazine (2/96), the following types of customer data were maintained on the database by 50% or more of the companies surveyed:
- Names of prospects
- Length of time he/she has been a customer
- Number of purchases annually
- Dollar value of purchases (Monetary)
- Recency (date of most recent purchase)
- Frequency (how often he/she buys)
- Source of original lead or contact
- Age/date of birth (didn't mention canaries)
- Other purchase influencers at same address (critical for many types of long sales cycle B-T-B products)
- Rentals of customer's name
- Sociodemographic info by survey
- Sociodemographic info by overlay (See reference to demographic overlays in my October, '96 column, Finding Your Way Through The Mailing List Swamp.)
- SIC code (also helpful to B-T-B marketers.)
- Promotional history
- Company info by overlay
- Company info by survey
- Track nonresponders
- Track database usage by category
The survey also reported that 40.5% of respondents use their database to cross-sell, 54.1% use it to prospect, and 54.7% use it reactivate customers.
Suspect to Prospect to Customer: The Database Triple Play
Some years ago, a travel wholesaler in Boston obtained a list of several million "senior citizens." Older Americans are known to travel abroad in greater numbers than their younger -- and presumably more tied down -- sons and daughters. This list, therefore, represented several million Suspects. People the company had reason to believe -- mostly their age -- might be interested in their various travel programs.
In order to determine exactly who on that list were so inclined, I was asked to create a lead generating direct mail piece. The criteria were quite specific. They wanted only people (1) 50 years plus, (2) who traveled overseas, not just domestically. My further charge was to produce (3) "highly qualified respondents."
Those who responded to the dm package then became Prospects. By requesting the travel-related information we offered, they had raised their hands, so to speak, and identified themselves as folks who met the company's stringent travel requirements. Actually, because of the structure of the offer, they were quite well qualified prospects (see details below).
Respondents were mailed the travel brochures they requested and those who actually booked a trip became Customers. This company offers a percentage discount credit to any customer towards his or her next booking. When a customer booked a second trip, he/she became a Client.
Of the company's regular clients, those who passed along the company's travel information to family and friends, or invited friends to accompany them on a future trip, or agreed to have their favorable comments used in the firm's promotional literature became Advocates -- the final link in the database marketing chain.
Clearly, the data collected in this company's database enables them to move people through the pipeline from suspect to advocate in a highly efficient and effective manner. They boast a very high rate of repeat business among their clientele, and since keeping a customer costs a mere fraction of the initial cost of acquiring one, most of that added revenue goes straight to the bottom line.
Slicing and Dicing for Fun and Profit
Most direct marketers are familiar with such P&L metrics as Breakeven Analysis, Gross Response, Net Response, Cost-per- Inquiry, Cost-per-Order, Contribution to Overhead, Return on Investment (ROI) and other basic measurements that define financial success or failure.
Less familiar, and much less often computed, is Lifetime Value of a Customer. That is, how much revenue can you expect from a particular customer from acquisition until the relationship terminates. It's an important number because it tells you how much you can afford to spend to acquire a customer, how much to budget for subsequent marketing activities, how much data you should collect on your database, and it provides vital guidance for many other critical decisions.
The formula will vary with the circumstances, but it's a matter of computing total marketing expenditures and matching that against total revenues per customer, including present value of future profits.
Database technology helps improve the bottom line in several other ways as well, especially for larger mailers. Predictive modeling is often a major reason for the care and feeding of a database, because it provides a way to find additional profitable customers with some reliability.
Say, for example, my travel agency client wanted to set up a predictive model of those "clients" I mentioned -- folks who booked two or more overseas tours. They would start with a demographic analysis of those customers, possibly including purchase behavior ("Hey, guess what! 92% of them bought jogging suits at Sears in the last 12 months!"), and other factors.
To that might be added a number cruncher's stew of lifestyle data, predictive variables and so on and out would come a computerized model against which to run all new prospect lists. The "hits" would be culled out and added to the "suspect" portion of the database.
Among the statistical tools at the modeler's disposal are such exotica as:
The point that needs to be made about these modeling techniques is that are expensive. If you're not mailing around 200 million pieces a year, you probably can't afford them. It takes heavy usage of modeled databases to recover the costs of the technology. For the rest of us, RFM works just fine.
ACT! - And Start Reaping the Database Goldmine
Another way to understand database management is to study the programs that are designed for it.
You don't have to be mega-mailer to put your customer information on popular contact managers like ACT! and Goldmine. Even dentists have office management systems (one I worked with locally I named "Dental DOS"), that kick out a personalized postcard reminder when your pearly whites are due for cleaning. My veterinarian's contact system even gets my cats' names right when it prints out their shot appointment letters.
After the contact manager types of database program come PC based database programs, "flat file," and "relational." Clearly, relational databases are more powerful and flexible and will do more of the heavy lifting for larger mailers. But flat files work fine for smaller mailers.
Some PC based DBMS include Microsoft Access, dBase, Lotus Approach and Paradox. Mainframe and client-server programs include FoxPro, Oracle and others. Note that these are databases. There is also a slew of mailing list management programs out there like Arc Tangent; catalog programs like Mail Order Wizard; and postal related programs like Group 1. You may well want to consider any of these programs, but be sure to understand the differences.
Whatever system you use, even if you have to start with the proverbial shoe box for now, start capturing those names, and include as much data as you can afford to collect and manage. Be sure to allow for expansion because changing systems in mid- stream is painful and costly. And whatever system you use, be sure it's ""Y2K" compatible. That is, it won't revert to 1900 at the turn of the century and lose all your data.
As I've said before in these columns, "true junk mail is an offer sent to the wrong person." A database system can help you target the right person, the first time.
And when your database is up and running, here are some ideas for what you can do with it ...
CONSUMER
(In addition to seasonal sale announcements, special events, promotions, etc.)
- Send timely reminders of needed services: doctor/dentist appointment, oil change/tuneup, "your imprinted labels will be depleted soon, order now with the attached reply form..."
- Send a card or letter on the anniversary of a purchase.
- Send customers' kids birthday cards, if you can get that on your database.
- Periodic surveys.
- Invite customers/prospects to a product demonstration or educational seminar. (Free to customers, slight fee for prospects.)
- Using careful segmentation, send:
-- Price changes, product changes, policy changes.
-- New product announcements. ("Preview" for customers)
-- Product samples to customers -- or OFFERS for product samples for prospects with qualifying information. - Contests and sweepstakes opportunities. (Supports mail order sales and in-store traffic building programs.)
- Frequent buyer programs. (But YOU keep the paperwork. Don't expect the customer to do it.)
- Newsletter (especially as part of Frequent Buyer Program).
- Annual Report.
NOTE: In addition, ALWAYS include such items in the products you ship to customers along with discount certificates, etc. designed to spur repeat business.
BUSINESS-TO-BUSINESS
Always include in ads - AND PRESS RELEASES - an offer for something other than "more information." Items you offer through space or mail should be of intrinsic value, tangible, and carefully targeted to the status of the recipient: Suspect, Prospect, Customer, Client, Advocate.
- Special Report or "White Paper" on technology or industry dynamics related to your product. Give it a title and offer it by name.
- Case studies of successful implementation of your product. (Get permission early and have these on file.) "7 Ways xyz Widgets have saved companies like yours more than $x million!" Note the specificity of "7 Ways" and "$x".
- Profiles of successful client companies. "Meet the pros who are making a difference in astrophysics today!" (Can be any level, CEOs, engineers, etc.)
- Test results of product performance - trade publication reviews and articles on such performance.
- Conduct an annual industry survey on some vital issue (exec salaries, corporate marketing practices, product or corporate performance benchmarks, research expenditures) and publish the results, mail to customers.
- Send press releases on new product announcements to customer/prospect segments. Include info for requesting product brochures.
- Reprints of articles on your company or products.
- Reprints of your ad campaign w/ note. (In case you missed our ads when they ran in xyz and abc, We're sure you'd want to see them.)
- Contests (customers only). Integrate w/ sales incentive contest for sales reps.
- Invitation to informational (product or industry) seminar.
- Product video (This can be integrated into all of the above mailings and offers, when properly targeted.)
- Newsletter.
- Annual Report.
(Notes on travel agency mailing described above):
I started with a booklet the company had pretty much lying around the office which I re-titled and teased on the envelope, "Go abroad in safety and comfort with 'Going Abroad: 101 Tips for Mature Travelers'- FREE to Travelers 50 and over!" That put the overseas travel and age requirements right up front.
Inside, I further qualified the respondents by asking them to answer four pointed questions about their age (again) and travel habits. They could also check off 3 destinations that interested them and receive free brochures on those locations.
The package was personalized on the envelope, letter and response device, both to impress older recipients, and to eliminate their need to write out any documents.
Resources
Database development, marketing and computer technology are highly complex areas. I strongly recommend reading a variety of books and magazine articles on these topics, and talking with a number of database consultants and service vendors before venturing into the database underbrush. It's a jungle in there!
Following are some places to start.
Direct Marketing Magazine, 1-800-229-6700
Target Marketing Magazine, www.targetonline.com
Marketing Tools, 607-273-6343, www.marketingtools.com
Business to Business Direct Marketing, by Bernie Goldberg and Tracy Emrick, Direct Marketing Publishers, Inc., 1304 University Drive, Yardley, PA 19067 215-321-3068
Also check the Direct Marketing Bookstore at www.netplaza.com.
Finally, there's the annual conference and exposition of the National Center for Database Marketing, 800-927-5007 or by E-mail at NCDMinfo@cowlesbiz.com. One trip and you'll be databased for life.