Deliverability is not a promise anyone can make
Once a message leaves our infrastructure, another company's classifier decides what happens to it. Here is what we can control, and what nobody can.
No email provider can guarantee that a message reaches somebody's inbox, and any provider implying otherwise is describing a decision that is not theirs to make.
Once a message leaves our infrastructure, the receiving provider decides where it goes: inbox, spam folder, or refused at the door. That decision is made by their classifier, using their signals, against their thresholds, and it can change on a Tuesday for reasons nobody publishes.
What is actually in our control
Quite a lot, and it is worth being specific rather than shrugging.
- Authentication done properly. SPF, DKIM and DMARC written and verified for your domain rather than left as a documentation page you were supposed to follow.
- The sending reputation of the infrastructure, which is why outbound mail is scored as well as inbound, and why bulk sending belongs somewhere else.
- Delivery logs and event history, so when something does not arrive you can see what we did with it rather than guessing.
And the same honesty about filtering inbound
No spam filter is perfect, including ours. It will put something in the wrong place in both directions — a newsletter in the inbox, an invoice in spam. The useful posture is to treat that as a fact to work with rather than a claim to make, and to give people something to do about it. Finer controls, including trusted senders and per-domain rules, are planned and not built yet; when they exist they will be announced here, on the day they work.