Association for Postal Commerce
"Representing those who use or support the use of mail for Business Communication and Commerce"
"You will be able to enjoy only those postal rights you believe are worth defending."

A COMMENTARY ON IMB AND ONE COMPANY’S BUSINESS MODEL

The following is a postal perspective by Charley Howard, vice president of postology at Harte-Hanks. Howard serves as a member of the PostCom board and the Mailers Technical Advisory Committee. The views expressed, however, are solely the author’s. The PostCom Bulletin will accept comments from responsible spokespersons of alternativee views.

We are seven months from the likely date for the new postal price increases and the implementation of Full Service IMB [Intelligent Mail Barcode]. This may seem like plenty of time for the mailing industry to get ready, but there are numerous "open issues" on many levels that raise doubt that all will be well.

First, the specifications surrounding IMB are not complete, particularly the Customer/Supplier Agreements (CSAs) because of potentially huge impacts upon software companies. These are "local" agreements with the USPS as to when mail can be entered and be afforded desired delivery. They vary considerably and the software companies are anxious about being able to code all of these by May 2009 on top of all the other coding required for full service IMB. Arguably, Full Service IMB has more impact on software companies and mailers alike than postal reclassification of 1996. And if memory serves correctly, that was a disaster.

Second, there still are issues around some ink jet imaging of the IM barcode on mailpieces reliably produced at production speeds. These include some of the largest of the mail service providers. Considering this issue has been being addressed for the past two years, time is running short. In addition, there are many mailers who haven’t even yet tried the IMB for live production runs. Okay, Postnet is alive ‘til 2011, but there are reasons to convert to IMB as soon as possible. Also, a price differential is likely for Full Service starting around October 2009.

There is the issue of the barcoded pallet placard and its requirement to be on the outside of any stretch-wrap on the pallet. Manual operations should have no difficulty, but those employing robotics are having issues.

Then there are the "internal" issues within each of our organizations as to systems, production, quality control, customer facing issues and more.

These are all serious issues, but barring our software company’s preparedness, Harte-Hanks is comfortable with being ready. The reasons for this should give cause for thought from other service providers.

Harte-Hanks started to prepare for IMB nearly three years ago. This coincided with the Postal Service’s Intelligent Mail/Surface Visibility development effort and the Mailers Technological Advisory Committee’s (MTAC) involvement. It readily became apparent to Harte-Hanks that participating in IMB was going to require significant change and investment, and that our clients would require our readiness. It is important for organizations to thoroughly investigate all aspects of IMB and how they will impact their own companies and relationships with others.

Harte-Hanks provides mail tracking services and the requirements of what was to become Full Service IMB would require a complete rewrite of the solution set of this offering. We completely hashed out the issues, presented required funding requests, budgeted the internal development costs, and two years later implemented a completely new system. This system also integrates with our logistics system for feedback on drop ship entry scheduling requirement. This effort cost the organization in excess of $350,000 and took two years from beginning to end. If we started today – we would not be prepared.

Another important issue had to do with our automated warehouse effort. Being sort of "late to the game" with this, our development coincided with the idea of a USPS MTEL barcode on pallet placards. Our team met frequently to determine the best course of action and decided to utilize the postal-required barcode as our own as the uniqueness requirements were the same. This would mean there would be a single barcode on the pallet placard reducing scanning confusion. However, there were important considerations. For example, Harte-Hanks performs an internal co-palletization process. Several versions of a mailing going to the same postal geography are prepared separately, then merged in our consolidation warehouses. The issue was how to guarantee the proper uniqueness in the barcode required by the Postal Service. Our logistics operation handles this during the trailer planning phase by subordinating those pallets to be consolidated to "children" status and creating a "mother" pallet. This is back-fed to the data center to complete mailstreaming. In production, the children pallets get their pallet placards replaced once consolidated with the mother placard. The point here is that extensive effort went into assuring that Harte-Hanks would be ready on this issue and that it would fit our own internal needs. This was implemented in 2006.

This leads directly to a major challenge facing anyone intending to use Full Service IMB – complete uniqueness of all barcodes for at least 45 days. Managing uniqueness should not be taken lightly. It is one of the reasons a mailer would rather have their Mailer ID in the IM barcode versus using the Mailer ID of the mail owner. It is the only true way to assure uniqueness. You have to think through your entire production process and relationships with outside providers to determine the proper solution that best fits your needs. It is the reason we let our logistics function manage the co-palletization assignment process up front. The 45 days really translates to at least 60 days. This is due to the fact that "day" counting starts with the first postal scan, or mail entry. The actual unique barcode may have been created 10 to 20 days prior.

I believe Harte Hanks was, lucky in their choice of ink jet imaging equipment as we have not had any issues in producing the required IM barcodes at full production speeds. We started piloting IMB on Flat mailpieces with J.C. Penney mail in the summer of 2006 and the next summer started full piloting of Seamless Acceptance on selected J.C. Penney mailings of letter and flat-shaped pieces. Seamless Acceptance is akin to Full Service IMB on steroids. This was very demanding in that we were venturing into the "unknown." One example that proved fortuitous had to do with the PDR, piece record, file of Mail.dat. It just so happened the version of the software provided by our vendor did not yet support this file. We were vaguely familiar with the IDEAlliance’s ADIS file specification, which is XML based. The USPS allowed Harte-Hanks to use a limited version of this file specification. It proved very successful and literally led to the pushing forward of the development of IDEAlliance’s Mail.xml. These are huge files and one does not want to port them unnecessarily.

Another cooperative development is the final requirement of the pallet placard. Barcode size and x/y dimensions are important with these barcodes. Zeroes and null values effect the length of the barcode and hence the height. These were solved. The issue of placing the placard on the outside of stretch-wrapping was not an issue as we determined that our scanning from tow motors required this. We simply spray adhesive to the back of the pallet placards and adhere them on adjacent sides of the pallet after stretch-wrapping. As a result, we were already doing this prior to the pilot testing. One caveat though, has to do with postal wooden pallets. Stretch-wrap does not grip the wood nearly as well as with the postal plastic pallets, sometimes allowing the material to shift on the pallet while in transit. Operators will add additional stretch-wrap covering the pallet placard. We have instituted corrective procedures with regard to wood pallets that preclude the need to "add" additional stretch-wrap.

The tray label created a challenge for us, though, in that we determined we were early on creating unreadable barcodes through our sizing the barcode to fit the label. The 24/10 tray label is extremely tight for space. It turns out that to do this properly, you need a font that allows for the specific adjustment of the x and y dimensions otherwise the barcode will be useless. This is easily remedied.

One of the biggest challenges we faced, and others will face, has to do with quality control. It is important to understand that the IM barcode is not human readable; it requires computer interpretation of a scan. If that were the only issue, we would be talking about simply adding hand-held scanners to the existing processes. More importantly, you need to know that the various services available with Full Service can be selected on a piece by piece basis. So, not only will the person performing "QC" checks need to scan the barcode, but they will need to know what is expected for that specific mailpiece. Will it be getting Address Change Services, and/or CONFIRM (planet code tracking) services? To do this, the device used to scan should be able to scan the address and determine if the services match what is supposed to match that specific address/addressee. This is another large investment in equipment, programming and people not generally required today.

Why did Harte-Hanks invest in this effort? We firmly believe in a "data rich" environment. The more data points we have on the position of a client’s mailpiece in the process of delivery, the better we can do our jobs in meeting their needs. Obviously, we desire every scan possible in the entire production/service processing of the client’s mail. Our solution sets are heavily integrated together making the changes required more difficult and extensive. However, these changes offer the opportunity to provide a far more robust solution leading to better delivery performance. The same holds true for the Postal Service. We knew it would take time and it has.

The ultimate participation in Seamless Acceptance when it goes live remains in doubt. The reason for this becomes apparent when looking at how it works. All mail detection will take place live in postal processing. Think on this. This means that nearly every piece of mail will be scrutinized and detected as to the containerization to the barcode imaging on the individual mailpieces. MERLIN looks at a tiny sample of the mail. We all know that inkjet imagers can "burp" every now and then, causing issues. Service providers like us receive tiny fees for our services compared to potential postage penalties that could bankrupt companies with a single failure. Clients will not pay this; it will be the service provider. As a result, we would need to invest in barcode evaluators for every inkjet and laser imaging machine we have. This would be a $4-million-dollar-plus investment with ZERO ROI. How many organizations would invest that much for no return in today’s environment? About the only way the USPS will get providers to participate is to offer a postage discount to the service provider – a novel approach.

The bottom line for Harte-Hanks is that we will be ready as long as our major software provider for mailstream management is ready. All other aspects are already in place. The reason for this is due to careful advance planning and development, development that took more than three years.