On 07/20/2023 03:17 AM, Lilium via mailop wrote: > > > On 20/07/2023 03:50, H via mailop wrote: >> On July 18, 2023 1:33:20 PM EDT, Lilium via mailop <mailop@mailop.org> wrote: >>> >>> This solution is based on Sisimai: >>> https://realsender.com/for-email-broadcasters/bouncehandler-app/ >>> >>> It helped a few customers with deliverability issues due to >>> wrong/inactive recipients in their lists. >>> >>> The article "how to handle BOUNCED EMAILS" is a summary of our >>> experience on this matter: >>> https://realsender.com/email-trends/2021-how-to-handle-bounced-emails/ >>> >>> Andrea >>> >>> >>> On 17/07/2023 03:47, H via mailop wrote: >>>> Is anyone using sisimai to analyze email bounces? I have just began >>> using it and am interested in sharing experiences etc. as I have not >>> found any user mailing list or forum on the net (there is commercial >>> support though.) >>>> >>>> _______________________________________________ >>>> mailop mailing list >>>> mailop@mailop.org >>>> https://list.mailop.org/listinfo/mailop >>>> >>> _______________________________________________ >>> mailop mailing list >>> mailop@mailop.org >>> https://list.mailop.org/listinfo/mailop >> >> Thank you for the links. After letting Sinai extract the bounce information, >> I import it into a database for further analysis of the bounce reasons. >> >> The two biggest issues analyzing bounces are: >> - The information returned by different mail handlers is not standardized. >> - Many bounces are classified as ”undefined” which I have to dig into >> further. One reason could be - have not checked this yet - that some mail >> handlers may return information in a language different than English and >> some return information in English. Our campaigns are in a non-English >> speaking country. >> >> Do you have any experience with the latter? >> _______________________________________________ > > I haven't made any change in the application settings. > In RealSender configuration the "undefined" errors are skipped. > > If you look at the Github releases page: > https://github.com/sisimai/p5-sisimai/releases > > There are changes like this one, maybe you can find some hints there: > FrancePTT : add laposte lpn code #441 > https://github.com/sisimai/p5-sisimai/pull/441 > _______________________________________________ > mailop mailing list > mailop@mailop.org > https://list.mailop.org/listinfo/mailop
The accuracy of the Sisimai bounce reasons might depend on where you are located geographically and which specific mail handlers are involved. It looks like the software is developed in Japan and therefore might handle bounces from Japanese mail handlers most accurately. Looking at a few of the several thousand bounces I have collected, I found that the bounce information is not correct, replycode is missing, diagnosticcode is missing, rhost is not correct and reason becomes undefined. It is hard to do a reliable analysis given this. Hardbounces due to mailboxes missing seems to work but that is only a small part of an analysis. Presumably adding parsing engines for more mail handlers would improve this and I have sent an email via their website. _______________________________________________ mailop mailing list mailop@mailop.org https://list.mailop.org/listinfo/mailop
