A Legacy of….Expedience?

This column follows a discussion on the Outputlinks group page on LinkedIn. I invite you to contribute your thoughts to the conversation by clicking here.

By Scott Gerschwer, OutputLinks

The French mathematician Henri Poincaré once said, "There are no solved problems. There are only more-or-less solved problems." This principle applies to high volume transaction output (HVTO), where a problem may be resolved in the most expedient way possible, in a manner that might be described as “more-or-less solved” with little action taken to solve the core of problem as long as the overall effects of that problem are mitigated.  

A case in point is the use of legacy print streams. What are “legacy” print streams? The term applies to the data streams that generate high volume customer output such as bills, statements, invoices and policy documents. As print-and-finishing operations grew more sophisticated during the nineties, continuing to the present with alternative delivery channels such as web and mobile, organizations needed to re-tool their applications for these purposes. But the applications, as written, were not designed for these purposes.

The industry came up with a (relatively) quick fix: software solutions that modify native print streams to take full advantage of capabilities as they developed---capabilities such as additional fonts, printer control commands, dynamic formats, color and electronic delivery---without altering the underlying “legacy” data. At this point, legacy print streams have gotten so old that the IT professionals who built the applications and best knew the code have long since retired.

According to Kodak’s Pat McGrew, co-author (with William McDaniel) of Wrestling Legacy Data to the Web, "legacy" data includes line data, line data conditioned with Xerox or IBM machine codes, line data prepared with pagedefs/formdefs, mixed mode AFP often generated by a custom-written bill or policy-generation program, as well as AFP in its many variations, Metacode, LCDS, Interpress, Postscript Level 1 or 2, early PCL—created by CSF, Documerge, XICS, CompuSet COBOL embedded with DCF macros...and so on.  

HVTO created with legacy print streams comprises as much as 90 percent of all applications, according to research. Despite the great strides made in document composition tools over the past eight to 10 years, the industry is still reliant on outdated print streams, however modified. It doesn’t appear that this will change any time soon. Research on the number of Print 09 attendees whose primary focus was finding a new document composition solution revealed none, zero. Not one person attended Print 09 in order to obtain a new document composition engine.

It’s somewhat troubling that a segment of the Print industry that is trying to keep up with alternative channels is still reliant on code written in the ‘80s and ‘90s.

It’s also surprising for a number of reasons, although one in particular stands out: that the largest acquisitions in the high volume print/mail industry have been document composition vendors. This includes, in no particular order, Group 1, XMPie, Sefas Innovations, Exstream and Document Sciences. Purchasers of these companies include (again in no particular order) EMC, Xerox, Pitney Bowes, Hewlett-Packard and the French Postal Service.

None of the purchasing companies are known to make acquisition decisions based on wishful thinking. They must have shared some notion that a trend was developing that would make document composition a hot item. And yet the industry continues to press on using outmoded print streams constructed in proprietary languages a long time ago. 

The reason for this is, according to the experts that have weighed in on this matter on the OutputLinks discussion board on LinkedIn, that only about five out of one hundred applications actually need the bells and whistles of the latest document composition software. Most of what gets printed does just fine getting printed the way it is, at least from a business perspective. The operational perspective is another matter, as we will see.

Scott Baker, VP of Business Development, GMC Software, recalled a panel discussion for the New Technology track at 2007 Xplor University, attended by approximately 55 enterprise accounts. During the final session, PC McGrew, the chairperson, conducted an informal "raise your hands" survey of the attendees regarding print streams. When asked who was still running AFP, almost everyone raised their hands; there were ten or so that were running Metacode workflows. When asked about PostScript, only eight raised their hands, and when PPML was mentioned no one raised their hands. Ditto for PDF. 

Ms. McGrew asked who in the audience was willing to change their output protocols to accommodate a move from mono to color. No one raised a hand. The message was pretty clear: "if it ain't broken, don’t fix it."

But this has to change. Print is expensive, considered by some to be out of date, obsolete, uncompetitive and hopelessly antiquated; how can it compete with new media channels if the tools and software that feed into it are so old and obsolete?

The HVTO industry has responded to the trend toward multi-channel delivery. It understands that consumers like to receive documents on handheld mobile devices, which underscores the need for functionality that supports not only high-volume print, but also real-time interactive and on-demand applications---from a single platform. The industry has adapted to change. Yet for the vast majority of applications the legacy print streams remain in place.

Because legacy solutions continue to work hard for business owners, and converting structured applications to a new platform is difficult, time-consuming, expensive---and even risky---the legacy continues.

Admittedly, the risks are daunting: bills and statements are business-critical documents. Interrupt the flow of these documents and you may halt or slow incoming revenue streams. Changing the look of a utility bill to include promotional messages may lead recipients to misconstrue it as a piece of junk mail. And that’s bad for business.

My colleague Brett Dashwood wrote this on LinkedIn, “Regardless of the format, if the data driving an application already contains everything that it needs to drive the existing communication requirements, then why replace a system that's working?”

There may be no earthly reason, at least from a business perspective. Post-composition and/or transform software breathe new life into old print output with minor modifications, the addition of production integrity markings, concatenating with newer data streams, and even adding color and white space management for simple Transpromo-type messages.

From a business perspective, the process works just fine as is. The cost and IT resources needed for an application migration are high barriers: It can take months and tens of thousands of dollars to migrate a single application. Multiply that by the number of applications processed by an organization and we’re talking about years and millions of dollars.

And, ironically, the chance for a clean migration may already have passed: companies are reluctant to mess with source applications because many of the guys who knew the old code have retired. The IT department doesn’t want to touch these applications for fear that they’ll be bogged down in a quagmire of proprietary code and ancient architectures.

Scott Draeger, product manager for HP Exstream Dialogue, put it like this: “If you can use what you have to build what you need you can execute on your communication strategy for a long time without upgrading, it spares the CIO from having to ask the CFO for the funds for a massive data migration/unification/re-imagination effort.”

But is this mind-set sound, long-term, strategic thinking? On LinkedIn, Scott Bannor,  midwest regional manager at Elixir Technologies, writes that change will occur only when the executive suites realize that document production is a critically strategic operation and take ownership of the associated technology and processes.

I agree: For too long now, NO ONE has “owned” the billing process. A bill is the bastard child of several stove-piped business units. GMC software’s Scott Baker pointed out a number of instances where, due to numerous external fixes to the data accomplished through programming logic applied at the post-composition stage, rather than upstream, the only place where the data/documents are truly accurate is in the formatted print spool. That can’t be a “best practice” for anything, can it?

So what will ultimately be the driver of change?

Scott Baker suggested that new regulations, the availability of lower cost color output, the emergence of high-speed color continuous feed production printers that are capable of supporting high-volume transaction production, the emergence of transpromo as a viable application, and the need for multi-channel electronic distribution are all driving change. Efficiency, process automation, zero defect production, and postal optimization solutions are also hot areas of interest right now.

Mike Lambert took the operational perspective. He pointed out that patching applications and concatenating multiple streams to create a single application is inefficient. And, like driving a car without ever changing the oil, eventually the inefficiency of the engine will become problematic. Paying for licenses and maintenance for obsolete software makes little business sense. Maintenance and support for legacy systems won’t continue forever; in fact, several vendors have already announced deadlines for maintaining legacy products. 

Scott Draeger put it like this: someday the fatal failure of some old-but-critical piece of equipment or system is going to force you to spend your maintenance dollars on eBay.

One driver of change is the desire to bypass IT altogether by taking control of and manage the document content without their involvement. The sales strategy of at least one document composition vendor is to go directly to marketing or the C-suite and sell them on the idea that the IT department, considered by many to be a barrier for change, is not needed to implement their software. Other vendors mentioned in this article are selling total solutions: for example, GMC Software has added a portal builder and is building analytics. Dialogue has long marketed a campaign tracking solution and, as part of the HP Software Universe, has hooks into many useful applications.

Lowering the technology barrier is a good selling point. Sefas Innovations promotes a way to gradually fix the legacy problems, optimize the processes, equipment and infrastructure, and migrate to a more robust multi-channel platform—all using the same engine. But every vendor has their own positioning strategy. Organizations need to evaluate each of them to determine the best fit for their particular business strategy.    

The modernization of the HVTO industry appears to be on hold right now. When this will change is anyone’s guess. The experts agree that legacy data will likely be with us for a long time. But true business leaders can tolerate inefficiency for only so long; few visionary leaders, once educated, can stomach the status quo. 

Change is coming.

I invite readers to add their input into the spirited discussion at the OutputLinks LinkedIn page.