P&G Heads for the Database Abyss
“Database marketing is officially sexy,” gushed researcher Julie Katz on April 11, 2007 in Forrester’s Marketing Blog in response to the announcement that Procter & Gamble’s Elva Lewis has conned her management into testing the database marketing waters.
“The boat’s setting sail, and we have one foot on the dock and one on the boat,” Lewis told AdAge.com’s Matthew Creamer. “If you listen to A.G. Lafley or Jim Stengel, they’re all talking about the declining return on investment in TV. This trend tells us that we should go to one-to-one marketing. We just haven’t put our money where our mouth is.”
If you listen to Denny Hatch, Ms. Lewis, you’ll bail out of this cockamamie scheme right now.
Package-Goods Nuttiness in the Early 1990s
Back when my wife, Peggy, and I were running WHO’S MAILING WHAT! we used to collect, analyze and log into our archive roughly 2,000 mailings a month sent to us from correspondents around the country. We actually handled more than that, but many were duplicates.
At the beginning of the last decade, we began receiving a new breed of direct mail—from package-goods people who not only were offering cents-off coupons but also were planning to track the redemptions back to both the mailing effort and to the individual household.
The object: to build a giant database of nickel-and-dime purchases by households.
When these things first started arriving, we dutifully logged them in, but paid scant attention, assuming it was the work of crackpots who did not know their arithmetic and figured direct mail would be a neat way to build a brand.
It turns out I was partly right. They were not crackpots, but rather were young, MBA, brand manager tigers who did not know their arithmetic and the greedy agencies that aided and abetted them.
And they all went bust.
Elva Lewis’s Pipe-dream
Lewis, who has been with P&G for 21 years, told AdAge.com, “We want solid relationships with 40 million to 60 million households. Right now, we have decent relationships with 10 million.”
Working with Targetbase, an Omnicom agency, Lewis has segmented her market into nine or 10 groups that she calls “cohorts”—new moms and baby boomers, for example. To new moms, she wants to promote Pampers and Tide.
“Never try to sell two things at once,” said the late direct marketing guru, Dick Benson.
“Our long-in-the-tooth marketers think that TV is a cheap way to get mass reach and that targeted marketing is an expensive way to get smaller reach,” Lewis told AdAge.com. “We haven’t been entirely successful at breaking through their thinking. We can use targeted scale and let the fixed costs of direct marketing—list buys, acquisition, response modeling, mail and fulfillment—be spread across 10 brands. It doesn’t work for P&G otherwise.”
P&G has 86 brands, mostly low-cost household and beauty items. Among them: Duracell, Crest, Iams, Folgers, Pampers, Tide, Pantene, Bounty, Cascade, Swiffer, Charmin, Olay, Prilosec, Pepto-Bismol and Pur Water Filters.
Does brand loyalty exist for Duracell batteries during a half-price sale on Energizer? Will a new mom buy Tide if a same-size container of Era on the shelf above is a buck-and-a-half cheaper? Will she ignore the low price on Scott’s in order to stock her kitchen with Bounty towels and her bathroom with Charmin?
Maybe she will be loyal to Crest and Pantene—items she puts in her mouth and her hair. And her Pur pitcher can only take Pur filters. But the other stuff?
When she buys Pampers, on the back of the cash register tape will be a hefty discount coupon for Huggies.
Chances are most decisions in the package-goods world are price-driven with loyalty out the window.
The Arithmetic of Direct Mail
Here’s a sample of direct mail arithmetic—a quick, down-’n’-dirty way to figure out what is needed to make a profit. Obviously, the numbers change based on the selling price and the cost of goods sold. For our purposes here, let’s say shipping revenue and cost of shipping is a wash.
|Revenue per order:||$40.00|
|Cost of goods sold:||$8.00|
|Reserve for returns (15%):||$6.00|
|General & Admin (15%):||$6.00|
(includes credit card processing)
|Allowable Cost Per Order:||$14.00|
If your mailing costs $450/M, then $450 ÷ 14 = 32. Thus you need a 3.2% response for breakeven.’’ Not ideal, but doable.
If you are doing an off-the-page space ad and paying $1,500, then $1,500 ÷ 14 = 107 orders needed to break even.
Okay, so Ms. Consumer buys a $6.00 container of Tide at the supermarket and uses her $1.00-off coupon she got in the mail from Elva Lewis at P&G.
Run the numbers.
Let’s say P&G normally would get $4.00 of that $6.00, but with the $1.00 savings coupon, P&G’s net is $3.00.
Insert $3.00 into the spot for revenue above and $.50 for cost of goods sold. Forget the reserve for returns, but include the 15% each for G&A and profit. Your allowable cost per order is $1.60.
The result: Breakeven is 281 containers of Tide or a 28.1% response.
According to the Friday Report on March 22, 1991, a great redemption percentage on direct-mailed coupons is 50 orders per thousand or 5% (which roughly is double that of a free-standing insert or FSI).
Multiply 50 orders by $3.00 per order revenue and the gross is $150 per thousand, which represents a loss of $300 for every thousand mailed.
This model simply cannot translate to anything near profitability—ever.
Add to this the expense of maintaining this household on a database and tracking the buying patterns on penny-ante items, and there is no way in the world for this system to work.
The Mail-0rder Package-goods Pioneers of the Early ‘90s
At the end of this story are illustrations of the mailing pieces we received at the time. They are the products of epic lunacy in direct marketing:
* Quaker Direct. With a budget of $18 million, Quaker mailed 54 million envelopes containing cents-off coupons for 14 products ranging from Aunt Jemima Syrup and Cycle Dog Food to Gatorade and Quaker Oats. Each coupon was bar-coded with an individual household ID so the purchases could be data-entered and tracked. “At this point the database moves from a survey-driven data file to a sophisticated transaction-driven file,” wrote Mollie Neal in Direct Marketing. “Quaker executives call this ‘household management,’ in which they are able to organize detailed information by family, which enables them to communicate on a one-to-one basis in the future.” The whole thing bombed and Quaker closed it down in early 1991. Quaker’s Dan Strunk left for reasons never revealed to the press.
* Kraft. The company mailed out a quarterly magazine, Good Beginnings, filled with recipes, ads and “over $4 in money-saving coupons” for Brown Rice, Tang, Sanka, Postum and Post Bran Flakes. It did not last long.
* Select & Save. (US Patent #5822735) An insert program mailed in a bright red, poly/plastic envelope filled with all the usual suspects—insurance, Sears portraits, Kodak Film and Disney books, plus cents-off coupons for Bayer Aspirin, Lipton Rice & Sauce and Good Seasons Salad dressing, each with a unique household ID. Sending out a collection of mail-order offers and cents-off coupons is okay, but trying to build a database from the redemptions is just not cost-effective.
* Reward America. Citicorp’s attempt to build a database of supermarket shoppers and keep track of what they bought. Fred R. Bleakley in The Wall Street Journal on April 3, 1991 reported this six-year caper cost $200 million and generated $20 million in revenue before it finally was axed.
* Air Miles. (“Buy Smart, Fly Free.”) The brainchild of Michael Miles (sic), the program promised frequent flier miles on virtually everything. The illustration at the end of this story shows how a muncher could redeem the UPC codes from four mini-bags of potato chips for one frequent flier mile. Air Miles did some very flashy, expensive mailings, but the idea that people would spend time to acquire a single air mile was preposterous. The business obviously failed.
*Ponds Lotion, Dove Soap, Ore Ida (Heinz), and General Foods Viennese Chocolate Coffee. All of these were very elaborate mailings containing cents-off coupons for single products. Ponds and General Foods included samples, while the Dove Soap mailing contained litmus paper to use for testing your skin.
*Schweppes. Schweppes sent out a 6˝ x 9˝ envelope containing a single, 55-cent discount coupon with a Bermuda Sweepstakes on the back. Drop the numbers in the above formula and look at a major marketing catastrophe.
Also received were mailings from Marlboro and Camel cigarettes, which made more sense. It was estimated at the time that if a tobacco company could get a smoker to try a carton-and-a-half of a new brand, that person would become a customer.
Carnation and Gerber sent mailings to new mothers offering cents-off coupons for baby formula. Like cigarette mailings, if the customer (in this case, the baby) thrived on the product, brand loyalty would be created.
So, now Elva Lewis wants to take P&G swimming in these pirana-infested waters.
Advice to Ms. Lewis from this long-in-the-tooth marketer: Have one year’s living expenses in the bank and keep your résumé up-to-date.