DMARC: start at none, read the reports, then tighten
DMARC tells receivers what to do when a message fails authentication. Publishing reject before you know who sends as you is how legitimate mail disappears.
Most internal newsletters die of the same disease — they are written for the sender. The ones that survive are short, predictable, honest about bad news, and never contain anything the reader could not use.
An internal newsletter is a recurring update sent to everyone in an organisation, and it is one of the few communication habits that either works well or wastes everyone's time — there is very little middle ground. The ones that work share a shape: short, arriving on the same day every time, written in a voice that sounds like a person, and containing at least one thing the reader could not have found out otherwise.
The failure mode is consistent enough to describe precisely. It starts as a genuine update, then departments ask for space, then it becomes a newsletter about the newsletter's contributors, then it is twelve sections long and the open rate is nineteen per cent. Nobody decided this; it happened because nothing was ever removed.
Congratulations rounds that name half the company, restated strategy with no news attached, department updates written to prove a department is busy, and anything that is really an announcement needing acknowledgement. That last one matters: a policy change, a deadline or something requiring action belongs in a direct message to the people affected, not buried in section seven of a weekly digest. If it is important enough that you would be upset it was missed, it should not be in a newsletter at all.
Bad news is what earns a newsletter its readership. An update that reports a lost customer, a slipped deadline or a decision that did not work is read carefully, because it establishes that the good weeks are also being reported honestly. A publication that has never carried anything uncomfortable is understood, correctly, as a channel rather than a source.
Sending outside the company raises problems internal mail never has. Newsletter software handles subscriber lists, unsubscribe handling and the sender reputation rules that decide whether any of it arrives.
Ettex Mail sends it and keeps the template consistent from issue to issue, which matters more than it sounds — a familiar shape is what lets people find the section they care about in ten seconds. Where a discussion is wanted rather than a broadcast, a channel in Ettex Chat is the better home, and the tooling question generally is covered in internal communication tools. The formatting side, including the parts that break in Outlook, is a separate write-up in email templates.
What we do not do: there is no read-tracking of individual employees, no engagement scoring per person, and no intranet-style publishing portal. Deliberately, in the first two cases — measuring who opened an internal email is a fast way to make people stop reading it honestly, and the useful signal was never in the open rate anyway.
Weekly or fortnightly, at a fixed time. Monthly is usually too infrequent to build the habit, and daily turns into noise unless the organisation genuinely changes daily.
One screen, or about five items. Set the cap before writing; it is the constraint that forces the editing which makes it worth reading.
Aggregate open rates are a weak but harmless signal. Tracking individuals is a bad trade — it changes behaviour without telling you anything you could act on, and people notice.
Usually it contains nothing the reader could not get elsewhere. Add decisions and numbers — including the bad ones — cut the length by half, and give it one editor with the authority to say no.
DMARC tells receivers what to do when a message fails authentication. Publishing reject before you know who sends as you is how legitimate mail disappears.
A hard bounce is a permanent answer and must be suppressed immediately. Retrying it damages the reputation that gets the rest of your mail delivered.
Segmentation is only worth doing if the segments get treated differently. Most segmentation projects produce a chart, and every customer continues to receive exactly the same email.