This whole discussion of VERP versus batched for better bounce
processing seems to have overlooked an arguably preferable solution:
MTAs should return the Internet standard delivery status notification
This has, in fact, been occurring. I've seen a dramatic increase in the
last 6 months in the use of the standard DSN. I don't have any hard
numbers although I can say that my hit rate for fully processing bounce
messages has inched to 85% from 80%, for those elists that are a
problem. What that means is the larger an elist gets or the lower the
frequency of messages, the more problematic managing the bounces
becomes. I divide my elists into two broad categories: those for which
the bounce processing is 100% for all practical purposes and those for
which I just can not get ahead.
By the way, I have trouble believing that SmartBounce gets 97% because I
see a new error message at least once a week, but then such statistics
are based on your subscriber community so anything is possible.
Overall, my community is worldwide and quite varied. Heck, I get error
messages in foreign languages almost every day, which result in frequent
updates to my parsing routines.
In fairness, I do have to admit I've been thinking a lot about VERP
distribution, for all the reasons already discussed. Even now I have
too many to think about failed messages that come back with no
indication or an ambiguous indication as to which address failed. It is
a personnel drain to deal with them, although lately we discovered that
letting them fail and keeping good records counts for a lot. We've been
getting lucky lately getting similar failures across multiple elists,
and the intersection can be quite revealing.
James M. Galvin, Ph.D. Principal
eList eXpress LLC +1 410.549.4619
607 Trixsam Road +1 561.619.2450 FAX
Sykesville, MD 21784 http://www.elistx.com
Delivering your email, your way.
There are only two ways to live your life; one as though nothing is a
miracle. The other is as though everything is. - Albert Einstein