Personalized printing is the process of combining a database, images, and text into a page layout that is then printed, typically on a color digital printing press. Personalized printing differs from static printing in that any or all components on the page can be unique; static printing simply repeats the same thing over and over. It also differs from mass mailings that simply have a recipient's address printed on the cover of a magazine, for example, with a low-quality, high-speed inkjet printer.
Personalized printing has been around for decades, but it was never very sophisticated—only text could be varied, and there wasn't much finesse in the application of variability. At best, a letter could be "personalized" with a salutation and a name. With the advent of color digital presses, sophisticated software, and powerful computers, each page can be finely tailored to each recipient.
Personalized printing technology isn't limited to personalizing direct mail, the same technology is used to print prepaid phone cards, lottery tickets, and many other items that require unique information on each piece. Another example of pushing the technology is apparel tags: typical clothing manufacturers turn over 60 percent of their product line per year. A company in Oregon, Oswego Print, receives apparel tag orders electronically through a web site, prints the tags, and ships them to an offshore clothing plant. The variability exists in the assorted graphic logos, barcodes, SKU (stock keeping unit) numbers, and prices. These are all pulled from a database when the tags are printed. This solves the problem of having to print tags, usually either far too many or not quite enough, well in advance of the clothing run to make sure the tags are ready to attach to the clothing before it is shipped out to buyers. By using variable-data printing and a digital color press, the most economical number of tags can be in the clothing factory within a week.
The Database Is The Key
Before any personalized printing can occur, a database needs to be created and filled with all of the information that is desired for the print run. It needn't be terribly sophisticated, a simple customer list with name, address, and purchase history is all one needs. Collection of this data can be a time-consuming and costly endeavor, especially in the push model of marketing. Push marketing literally pushes information at the consumer. It is incumbent upon the marketer to make sure that the information sent to the customer is correct and timely. Nearly everyone has received junk mail with a wrong spelling, duplicates for nonexistent household members, even credit cards in the name of family pets. Many large databases are rife with errors, duplicates, and outdated information.
A growing means of collecting data for current and especially new customers is with a well-designed web page. The web has created an enormous increase in pull rather than push marketing. In pull marketing, the customer goes to the company for information. Prior to the proliferation of corporate web sites, customers had the telephone as their usual means of contacting a company for information, which can be a frustrating experience given the propensity for many companies to install phone systems designed to prevent callers from talking to a live person as much as possible. A customer can go to an intelligently designed web site and do two things: browse for information or request additional information to be delivered via email or by post.
There are a few successful examples of well-designed web sites that let the customer make a number of choices from a range of products, specify options and colors, and then request a personalized brochure to be mailed for further consideration. An example is the Subaru of America web site, which presents visitors with a range of car models and lets them select one or more for more information. The visitor can specify equipment levels for a given model, such as a nicer trim level, an automatic transmission, a compact disc changer, and so forth, and can also choose from the available colors for that model. Once the "virtual" Subaru is built, the visitor can then enter his or her name, address and telephone number. This information is gathered and processed and submitted in batches to a Xeikon digital color press. A few days later, the customer receives a personalized brochure of the car, complete with a photo of the specific model, trim level, and color, along with a breakdown of the cost of options, the total suggested price of the car, and the name, address, and phone number of the closest Subaru dealer.
Another example of pull marketing is being done by Whirlpool, the appliance manufacturer. A customer calls a toll-free telephone number, answers a few questions, and within a few days a full-color brochure specific to that customer's request, such as refrigerators, arrives in the mail.3 Pull marketing is the easiest means of collecting prospective customer data; the customer is obviously interested in the product and is willing to provide personal data in order to receive more information about a product. Think about the difference to the customer's perspective, instead of being harassed by telemarketers or bombarded with poorly targeted junk mail, he or she will appreciate the availability of a company's product information to be reviewed at leisure.
The most important component of a successful direct-marketing variable-data printing (VDP) run is a carefully qualified database of people who are to receive the printed material. A high-end VDP run, especially on good paper stock, is not an inexpensive proposition, so it is in the best interest of the producer to have a database carefully culled of inappropriate recipients. Database selection and tuning is difficult and time-consuming, but is as essential to a direct marketing campaign using VDP as to conventional bulk mail or printed materials. A campaign for luxury goods will certainly fail if mailed to residents in low-income areas.
A personalized print run is not the best means for reaching new customers. In fact, it costs five to 10 times more to acquire new customers than it does to market to existing customers. This is the great strength of personalized printing, all of the data is already in place, or should be, for a highly personalized print run targeted at existing customers. It is this technique that can produce returns of three to 15 times higher4 than traditional direct mail response, which is typically two percent.
Software Manipulation Required
An issue in preparing a personalized print run exists with the software used to produce these runs. Data in a database must be exported to a plain text, tab-delimited file from the client's database. The client supplies this file, any images, and the final page layout to the shop that will print the job. The plain text file is imported into the particular program used to generate the personalized pages. These programs offer data handling tools called "rules."
A rule is a directive to the program that takes the form of "if
A question occurs in the production phase: Who is responsible for creating these rules? At what point in the process is responsibility for accuracy handed off from the client to the printer? Rules cannot be specified in the data file; the programs are expensive and often protected with a hardware key, so it is impractical for the client to perform this step. This leaves an operator at the print shop responsible for correctly following the client's written instructions for rules-based processing unless the client does it in the print shop. Additionally, it leaves the client responsible for writing up exact instructions for the print run. Both methods are impractical. This is a relatively new business, so new ways of conducting it will need to develop so that data accuracy is not left in the hands of the wrong person. The programs available for producing VDP currently are not designed for use in commercial printing; they are targeted at in-house printing operations of a large organization, which might maintain a digital color press as a part of its reprographics operation.
The upshot is this: Today a customer cannot walk into a commercial print shop that does not understand the customer's specific marketing needs, hand over a disk containing the database, page layout and variable images, and expect to pick up the finished print job a few days later. The production process requires the participation of the person most familiar with the data, the marketing program, and the desired result. A random press operator cannot know these things.
What About The Rip?
It is the ability to perform rules-based selection and placement of text elements, such as items from a database or separate text files on disk, and graphic elements, such as electronic color scans or vector-based artwork, into a page layout that differentiates VDP from simple mail-merge operations that one can easily do with even the meanest word processor. Most VDP processing systems work within QuarkXPress as a Quark Xtension. A copy of QuarkXPress and a Power Macintosh computer system will also be a required component of the VDP workflow. At present, VDP production applications only run under Mac OS. A QuarkXPress Xtension is often only part of the complete package. Some vendors offer specialized RIPs (Raster Image Processors) in packaged systems that are designed specifically for VDP imaging.
The biggest challenge in VDP production is the fact that a variable print run is composed of unique pages. A normal digital print run is RIP-one, print many, analogous to a conventional press, where each sheet is exactly the same as all the others. VDP produces pages that vary in content in degrees of very small to completely different. In a PostScript workflow that is not optimized for VDP, this means a complete rasterization of each page before it can be sent to the imaging engine. In the case of established imaging technologies, such as an imagesetter, a platesetter, a color proofer, or a prosaic office laser printer, driving the engine at its full rated speed is not really the prime concern. In VDP production, driving the target print engine at its full speed is critical to not only productivity, but to the actual process itself, particularly for digital web presses such as the Xeikon. A web cannot be stopped and started, blank areas appear in the web if the print engine does not receive data for a particular segment. Sheet-fed digital presses like the Indigo E-print can certainly accommodate stop-start operation but at great expense to production efficiency. RIPs designed for the first generation of 4-color digital presses, for the most part, cannot drive the press at full speed when processing variable pages. Consider that a letter-sized page (8-1/2 by 11") rasterized as a process-color print job at 600 printer spots per inch will occupy 67 megabytes. This is an enormous amount of information to store and process, and most RIPs just can't process this much data fast enough to feed a modern digital color press.
Within two years, this processing obstacle should be eliminated to a large extent. Indigo has already overhauled the RIP for its E-print digital press that it claims is now capable of driving the current press at full engine speed when imaging variable information. Scitex, Agfa, Varis, TR Systems, Xerox and others all incorporate proprietary extensions to PostScript to pre-rasterize static images and later merge these with variable data at print time, which really speeds up the process. This method of forms caching is the primary reason why VDP has been possible at all. A bottleneck still exists in the process of rasterizing variable images and in transferring rasterized data to the print engine. Future enhancements to RIPs will alleviate these problems.
Auditing The Job In Process
Auditing is an important feature that should become standard on all variable-data/digital press systems. At present, only Agfa offers an important feature on the ChromaPress and in the Intellistream RIP. The NexPress, a joint venture of Kodak and Heidelberg, reportedly will offer a real-time auditing feature on the press and in the RIP, so if a problem occurs and a piece is not printed, the RIP will automatically print a duplicate.
These issues are simply part of the evolution of the personalized printing process. It's a new business, and it has some growing pains that must be dealt with before it can become common practice.
David Broudy is a graduate student at the Rochester Institute of Technology who expects to receive his Master of Science degree in May 1999. He is the co-author with Frank J. Romano of Personalized and Database Printing: The Complete Guide, a book due out next month. Broudy has also written several textbooks for graphic arts software applications such as QuarkXPress, Photoshop, and FreeHand.
Article reprinted with permission of author. Copyright 1998 by David S. Broudy. For more information, email the author.
Personalized Printing References
www.chromapress.com
Another necessary component for variable-data imaging is a software program that integrates textual and graphic variable data from a library into a page layout based upon a set of data-handling definitions or rules. These rules can be as simple as a static list of page elements or as complex as a set of if-then-else constructs.
The lack of auditing features can be a major drawback to variable printing, especially for sensitive or confidential material. Most desktop VDP applications and digital color presses cannot perform auditing functions during a press run. Auditing functions ensure that every page sent to the press is actually printed. If there is a web break, the press runs out of toner, or if another malfunction occurs during printing, there is no way to guarantee that the press run is resumed uninterrupted after the malfunction is corrected without expensive and tedious manual comparison of each piece to a proof. Trying to determine which 25 of a 10,000-piece run are missing would be a nearly impossible task. High-speed systems that are used to print utility bills or bank statements, for example, typically print optical marks on the edge of the paper and use an Optical Mark Reader (OMR) to verify that each piece is accounted for. If a piece is missing, it is identified and the operator is alerted so the missing piece can be rerun.
For more information on developments in this ever-changing field, check the following web sites:
www.indigonet.com
www.printers.ibm.com
www.scitex.com
www.seyboldpublications.com
www.xeikon.be