Variable printing in the Illinois Senate race

June 1, 2004

Johnson’s World by Steve Johnson

Although he didn't come out on top in the Illinois primary election this spring, Andy McKenna won, hands down, the prize for best use of variable data in a political campaign. There's a lesson to be learned from the digitally printed, variable-content postcards McKenna employed in the eleventh hour of his senatorial campaign — even if you shy away from political printing.

I've observed that variable digital printing for paying customers tends to run to two extremes. On one side are the oft quoted case studies of intensely complex ad campaigns with variable graphics, alternating versions and personalization that borders on the invasive. These generally are touted as utilizing Web-generated information from client or prospect inquires, or from sophisticated databases collected at point-of-purchase over a lengthy period.

The other extreme is the name/address/city/state/ZIP form of variable printing that is the most common iteration of the genre. It's relatively easy to do, because it is little more than address labeling. It can be performed with your run-of-the-mill wordprocessing software. It is also the least glamorous form of variable printing and is almost never covered by the tradepress pundits.

The difference is in the database

Analysis seeking to probe the difference between these two extremes often misses the mark. Most studies focus on the differences between the printing establishments that represent these two ends of the spectrum. The difference is not in the print shops. It isn't in the digital presses or even in the complexity of the software used to merge the variable and static data.

The real difference between the rare complex jobs and workaday mail addressing lies with something that we printers have no control over: the database. Although the pundits urge us to “educate” our clients and to help them cultivate and cleanse their databases, in truth, we are usually stuck with whatever the customer has to give us. And that is where the brilliance of McKenna's campaign postcard comes in.

I should stress that I have no affiliation with McKenna — I'm just one of millions of Illinois voters who received the card. My firm did not design or print the piece, nor do I know for certain who did. But the story is simple to deconstruct.

Because McKenna was running for office, voting records were the only database that mattered. Since he was seeking the Republican nomination, only registered Republicans were useful prospects. The Illinois voter rolls are nothing more than an extremely simple database containing name and mailing information, plus precinct, township and county. No amount of “consultative selling” on the part of the printer could improve that database. His campaign committee might have presumed, as his rivals did, that the database was good only for addressing envelopes, or at best printing a “Dear Mr. Johnson” letter.

Subtle personalization targets voters

Instead, McKenna opted for a fourcolor, digitally printed postcard chock full of variable information. The original mailing was an eye-catching 11 × 8½inches, with subsequent smaller postcards. The postcards were addressed and postal bar coded inline. Some postcards did indeed contain the obligatory “Dear Heidi & Stephen.”

It is worth noting that “Mr. and Mrs.” were not used, since voter rolls list only individuals. By using only first names, the mail merge was kept simple, and the confusion caused by multi-last-name households was avoided. And to the mailer's credit, I never received a duplicate piece, although two registered voters live under our roof.

Careful reading revealed repeated references to me, my wife, our hometown and our county. Unlike some demonstration mailings I've seen that garishly showcase the variable fields, these references were subtly placed. I counted six variable instances on the larger postcard. All of the cards used only our names or our village and county.

The printer in this case couldn't tell the client that his database was no good, because the database was as good as it was going to get. Nor could the printer suggest a drawn-out data-collection effort, since election day was looming only weeks away. Instead, the designer/printer/mailer worked with the tools at hand, but used them far more creatively than the typical variable-data effort.

The campaign was convincing, with each mailing resonating as though it had been individually tailored to the recipient. And in a way, it was.

The chance of ever scoring General Motors' next variably printed Cadillac-owners mail program is extremely slim for most of us. On the other hand, mailings such as I've just described are literally waiting to happen. If the concept of variable printing is ever going to succeed, it will be this sort of creativity that will make it work.