Ben Greenfield wrote:
>> - You're looking a little too closely at daily changes, particularly around
>> implementation time. Allow the thing some time to settle, perhaps a month,
>> before considering next steps. Bear in mind that there are multiple,
>> independent good and evil actors here, each reacting to the others all the
>> time. This will take time to settle, a single day's (or week's) change is
>> unlikely to be actionable. Note in particular that the larger receivers are
>> almost certainly comparing their user feedback ("This is [not] Spam") with
>> your DMARC policy ([un]authenticated messages that get reported as
>> [not-]spam) as an input to their decision making. On the fairly small
>> numbers that you're talking about, this calculation could take weeks to
>> converge.
> I think what you are saying is that my small traffic volume 15,000 messages
> are such a small blip in the spammers world they will be doing some monthly
> analysis to notice and adjust their routine accordingly.
No, each of the larger actors is analysing others' actions and adjusting their
own actions constantly. The relevant point is that the system as a whole is
slow to stabilise because each actor's analysis affects their actions, which
affects other actors' analysis, which affects their actions, etc.
>> - How much is on-domain (vs. cousin-domain) impersonation costing you in
>> fraud/support/churn losses? If it's costing you thousands of dollars a
>> month, then by all means bring in the professionals. If you can't price it,
>> or you haven't done so yet, or it's a trivial amount, then you're probably
>> done.
> That is good to know since nobody is directly paying to do this. I imagine
> that I should just keep an eye on ARC and I could always become an early
> adopter of that as way to improve things.
For most forwarders, ARC will be "good practice", but unlikely to provide a
measurable benefit for quite some time. It might eventually make it possible to
undo DMARC workarounds for forwarders (John Levine just posted a link), but
this is far from certain.
- Roland
_______________________________________________
dmarc-discuss mailing list
[email protected]
http://www.dmarc.org/mailman/listinfo/dmarc-discuss
NOTE: Participating in this list means you agree to the DMARC Note Well terms
(http://www.dmarc.org/note_well.html)
