2005 Apr 02
Advanced Personalization
David M. Raab
DM Review
April, 2005

Telecommunications executives talk about “the last mile”: the challenge of connecting to customers through wires controlled by often-hostile local phone companies. In fact, many new technologies face a similar problem. They are stalled because downstream systems cannot deliver the benefits the technology is supposed to provide.

Marketers have heard the story often. The sales automation system failed because the salespeople wouldn’t use it. The email inquiry system failed because we couldn’t respond quickly enough. The campaign failed because we couldn’t distribute the leads. The predictive model failed because we couldn’t apply scores in production. The segmentation system failed because we couldn’t manage so many variations.

Such problems are probably inevitable. Execution systems are built to do today’s work and what will clearly be needed tomorrow. Additional capabilities may never be used but impose real costs. This means that building them is often not a sound business decision. The result is that when radical new opportunities appear, the execution systems must be rebuilt to take advantage of them. (As the previous examples suggestion, execution systems include people and processes as well as technology. Training and organizational issues are often much larger obstacles than the computer systems.)

Rebuilding these systems takes time, but–if the opportunity is clearly worth the cost–it does eventually happen. Of course, it’s hard to prove the value of an opportunity you can’t execute, but not impossible. It’s why people do pilots and write white papers and publish case studies. The point is the inability to execute is a temporary situation which ends as new execution systems are built to accommodate the new requirements.

Marketers are now seeing such a transition in the realm of personalization. Sophisticated segmentation and campaign management systems have for years made it possible to assign highly tailored messages to individuals, both for outbound campaigns and in response to interactions. But delivering those messages has been a problem, particularly once you get media such as Web pages and call center scripts. Because the outputs in both those cases are electronic, they need not worry about many physical constraints.

Printing, on the other hand, is quite physical. Traditional printing reproduces images from fixed plates and is impossible to personalize. But methods to generate variable images have advanced in recent decades so that personalization is now affordable in many situations. Print personalization techniques range from filling in blanks on preprinted forms to generating individual full-color documents from unique digital images.

As physical capabilities have expanded, the constraint on personalization has shifted to the intermediary systems that control how jobs are set up and executed. These accept the lists of documents with personalization instructions and translate them into specific formats that printers and other devices can execute. For example, a system might receive a file of customer names and message codes, and convert these to a stream of print images for a laser printer.

Such systems were originally different for each printer. Setting up a job required programming skills, and setting up each run took more effort to align forms, load customer data, load standard messages and images, and set other parameters. Although users could generate personalized output, high set-up time and costs meant they still worked with a limited number of large-volume batch jobs. The amount of variation and ability to make fast changes were severely limited.

Newer systems, from vendors including ClickTactics, Exstream and Trialogue, break through several of these barriers. They can combine inputs from multiple sources, such as Web, campaign management and billing systems, to consolidate messages aimed at the same person. They provide graphical job set-up and design tools that allow non-technical users to make changes for themselves. They maintain repositories of reusable text and graphic elements, making it easier to create new outputs and enforce corporate standards. They include workflow capabilities to streamline review and approval processes. They control multiple output devices, allowing a single job to direct printers, email, Web pages, and other channels as appropriate for each customer. Some can set up one document so it can be automatically rendered in different formats. They can control other media such as selective inserters that place preprinted materials in different envelopes.

The ability to consolidate inputs from different sources is especially significant. It means marketers can add or discard individual marketing programs without worrying about set-up details, minimum batch sizes, or flooding the customer’s mailbox. This frees the marketer to develop customer management strategies of almost any complexity and still be confident they can be executed. Similarly, the ability to direct outputs to different media without worrying about the details lets marketers focus on identifying the most effective techniques.

These systems have their drawbacks. Consolidation sets up the personalization system as a sort of super campaign manager that chooses among the recommendations of the existing marketing systems. This requires some complicated rules to prioritize messages for the same customer. Some systems add complexity by creating a new customer database with its own history of messages sent and responses for each customer. Consolidation also raises an obvious organizational issue, since no marketing manager wants her message to be the one that’s overridden. But this final issue is not insurmountable: companies unready to manage such conflicts can keep some streams separate.

The ability to render one document in different formats has similar effects. It pulls control away from the individual channel systems (printers, Web sites, call centers, etc.) and gives it to the central personalization system. Operators of the channel systems may resist losing their power as gatekeepers.

Relying on one system for such a broad range of functions also raises the familiar question of whether specialized tools might actually perform particular functions much better. Actually, the answer to this question is quite clear: yes, the specialized systems will be better. But it’s the wrong question to ask.

The relevant questions are whether the personalization system can work in conjunction with the specialized systems–which in most cases they can–and whether any resulting capabilities are sufficient for the purpose at hand. After all, the purpose of the centralized personalization system is to remove the execution barriers created by the existing systems. Slightly less sophisticated campaigns or less elaborate documents are acceptable prices to finally reap the benefits of refined segmentation, individualized offers and quick reaction that have long been promised but rarely achieved.

* * *

David M. Raab is a Principal at Raab Associates Inc., a consultancy specializing in marketing technology and analytics. He can be reached at draab@raabassociates.com.

Leave a Reply

You must be logged in to post a comment.