Back in 2002, we were crowded into a conference room at USPS headquarters – anyone who had an interest in tracking mail with the new PLANET code was there. There were a lot of good things to report – tracking letters was proving to be very reliable. But that’s not what anyone wanted to discuss. “What about flats? You need to get better at tracking flats!”
Frankly, flats were much more difficult to track in the days before Intelligent Mail and, more importantly, Informed Visibility. There were no pallet scans, no assumed scans – if a mail piece did not go through a sorter, it didn’t get tracked. If you could get 25% of a flats mailing to track, you were doing better than most.
That all changed with Informed Visibility – sort of. Now, FAST appointments are tracked, pallets are tracked, bundles are tracked – and all that data is assumed to the single piece. So now, everyone is delighted with flats tracking…well, not everyone. Tracking flats is much better than it once was, but it’s a long way from where we’d like it to be.
How bad is flats tracking? I’m glad you asked! In addition to tracking mail, we track tracking. As you may imagine, we track a lot of flats, and several years ago, we formulated a flats tracking report that shows how many pieces received at least one scan, and how many pieces were tracked to delivery, based on our business rules. We break it down into sortation type – identified by those first two digits in the IMb – the barcode ID. Here’s a summary from January:
You can see in this report that we tracked a bit over 31 million flats in January. Of those, 30,845,613 (98.80%) got at least one scan. Almost all of these are assumed scans based on pallets being scanned as they are taken in at the Postal Service. So, to prove you mailed on time, this is a very valuable metric. Unfortunately, only 60% received delivery scans – but we can gain additional insight by looking at sortation levels. Carrier route sorts are consistently the worst of the bunch, and because we tracked a lot of carrier route mail in January, it dragged down the average a bit. Other categories showed better scan rates to delivery, although not as much as we’d like.
This report has nothing to do with how long mail is taking to be delivered – it simply shows how many pieces are receiving tracking data. Just because a piece was scanned to deliver it does not mean it was delivered on time, and just because pieces didn’t receive tracking data does not mean they were not delivered on time. We just couldn’t see it.
Why are scan rates so different than letters? There are some inherent differences in how they are handled. Virtually all letters are run though automated sorting equipment, while many flats are not. Tracking flats really relies on the scanning of pallets and to a lesser extent, bundles.
Here’s how it should work: Mail is prepared and stacked on a pallet. A pallet placard is put on the pallet. The placard and its IMbc (which is an IMb for the container) is recorded in the Mail.Dat and uploaded to PostalOne! This eDoc tells USPS what pieces are on each pallet.
When the pallet arrives at the Post Office, it receives its first scan – 98.8% or the time. Then the pallet arrives at a facility for processing, and it should be scanned again. In the case of carrier route mail, that processing facility is often the DDU where the mail is delivered. At some point, those carrier route bundles will be removed from the pallet and distributed to carriers for delivery. BUT, before doing that, the pallet should be scanned, and that scan, along with its operation code, should be uploaded into the Informed Visibility system, where it will be reunited with the mail pieces that were on it, and that pallet scan is then assumed down to the pieces on it. When scan data comes to SnailWorks, a bell rings, an angel gets its wings, and we post the scan. That will generally serve as a last scan, and based on when it occurs, we will declare when the piece is delivered.
So why aren’t scan rates for flats better? There are a few possible reasons:
- Some flats bypass automation altogether and are hand sorted.
- Some flats are scanned on bundle/package sorters and may not be oriented correctly and cannot be scanned.
- Some flats may be moved to new pallets within the system and their connection to the pallet data may be lost.
- Sometimes the process breaks down, and folks don’t scan what they are supposed to scan.
- Sometimes they can’t find their scanner.
As it turns out, that last one is a big issue. A few facilities are simply missing scanners – in some cases, hundreds of them. It turns out USPS does not have much control over that inventory.
We’ve been measuring flat tracking statistics for years now and have seen very little movement. Recently, the Postal Service has been putting more effort into identifying and correcting the things that cause flats to go missing. We remain, as always, hopeful. The technology and systems are there to track a high percentage of flats. We’ll keep watching.