Needed: A Leap of Faith

TWENTY YEARS AGO, IT WAS a radical idea that scared people:

“You’ll mail my list net-net so my rental income will go down.”

“You’ll use my data and I won’t get paid.”

“My house file will disappear into a big database that anyone can use.”

Building private databases for business-to-business mailers was new and required a leap of faith.

But like everything in the list industry, building B-to-B databases was dependent on ethical business people trusting one another to make a bargain and stick to it. The fact is, every time we ship names we’re relying on the integrity of others.

Twenty years later, databases are still working, more efficiently than ever. List owners are making money. Data security is protected.

Now it’s time to use this prospecting model for consumer mailers. Right now we have the ability to increase the cost-efficiency of consumer direct marketing, but it will take a change of thinking in the list community.

They say “junk mail” is the mail people don’t want. But consumers want to receive offers relevant to their interests. They want us to know who they are — in the mail, on the phone and online. And they will expect more tomorrow than today.

Right now we have data at our fingertips that could improve response and customer lifetime value. We gather the data every time we do a merge/purge, then we throw it away.

Ethical use of that data can mean more successful prospecting — which then means larger house files, larger universes to rent, and more use. This in turn would mean more income for list owners.

Yet we’re tied to a merge/purge prospecting model that hasn’t changed much in 30 years. Bits of database thinking have crept in — some out in the open like modeling, credit screening and reuse; others relegated to semi-secrecy, like “super dupes.”

A private promotional database is not a cooperative file like Abacus, I-Behavior or the Wiland database (although those files could be part of a private database). It’s built just for the customer, working with an experienced list broker. Only the customer can access the data, for numerous offers if needed.

It starts with a single merge/purge and grows, accumulating useful data for improved prospecting. The greatest benefit is to mailers who prospect frequently and/or have multiple offers.

HOW DOES IT WORK?

First, everything is out in the open. List owners are told what may be done with their data. And nothing is done with data if the list owner doesn’t approve.

A private prospecting database starts as a merge/purge. Pricing is negotiated with each list owner, and the mail plan is constructed. All the lists on the plan are ordered, merged, and perhaps enhanced with demographic and other data. Net names are mailed as usual.

But now the history and output of that merge is retained as an information base. Responses are matched back to the base, including blind “white mail” responses from online buyers, so the mailer knows where these names came from and lists are properly credited.

Models are built to identify the characteristics of responders, such as original list source, number of multiple hits, nature of multi-hits (that is, the lists used to achieve them), and demographics.

With the next merge, new names are ordered and added to the base. More multis are created, identifying additional source data for each household — for several individuals in a household. (Who’s the best prospect for each offer, the man or woman in the home?)

A new mail plan is created that reflects traditional list-response data and what’s been learned from analyzing previous responders.

Names are pulled for the next mailing, some of which were freshly ordered, others from the first build which are now effectively “reuses.” List owners are paid based on the pricing each agreed to.

How are list owners paid? The best option is fractional allocation — every time a name is selected, each list owner who contributed the name is credited. Each multiple appearance counts as a fractional use. So a name appearing on five sources is credited as a one-fifth use to each list owner, and each owner is paid per its own agreed pricing.

Pricing is one of the factors in the model, so an expensive list might be replaced by a less costly source, as in today’s merge/purge environment.

Results from mailing number two are matched back to the base and analyzed. New models are created, with additional criteria such as offer details, which might be different.

This database approach is particularly appealing for mailers with multiple offers. For example, catalogers with several titles which probably rent many of the same lists for different books. A database environment allows a mailer to identify prospects for multiple offers and send to their names again and again.

Using a database, mailers actually can measure the impact of multiple mailings to a household. Frequency of contact becomes an aspect of the model to determine not just the impact of frequent contact but also the affect of contact when sending out several offers.

Eventually, lifetime value data becomes the prime criterion for modeling a database. Prospects are selected who are most likely to respond not only to an initial offer but to convert to long-term customers. Models can identify best prospects for multiproduct or multichannel buying, higher dollar sales or frequency of purchases.

Why should mailers test a private prospecting database? If 20 years of B-to-B success is any indication, it will mean more effective prospecting. Good prospecting is the first step in the cycle of direct marketing success for everyone. Mailing efficiently means DMers can send out greater volumes, rent more names, build larger house files, have more names to rent to others — on and on, full circle.

WHAT’S IN IT FOR THE MAILER?

1. Net-net pricing, based on negotiations with each list owner.

2. Enhancements paid on a royalty basis when used.

3. Modeling on front-end performance and LTV to choose prospects for initial response and conversion.

4. Ability to select by list and/or by model criteria across entire database.

5. Reactivation, to identify the best prospects in a lapsed file.

6. Match-backs, to know where Web “white mail” comes from and what mailings they received.

7. Contact strategy. Use contact frequency as a factor in models.

8. Multi upon multi. Track each time a name or household appears.

9. Male/female per household. Identify multiple names in household and select best responder by gender.

WHAT’S IN IT FOR THE LIST OWNER?

1. Reuse of older names means new income.

2. New list usage due to more extensive modeling.

3. The ability to negotiate pricing just as be-fore.

LARRY MAY is CEO of Direct Media Inc., Greenwich, CT.