I have had issues with it over the years. Many will blacklist entire cidr nets for a single bad actor. I get this on my linodes frequently if I proxy traffic through them. Ie: tons of captchas on google/YouTube.

When I ran my own mail it was similar. Often having to spend time getting IPs off rbls and the like because some other node on my subnet was malicious.

In the end, I just moved my email over to workspace. Not ideal. But it works.

One thing I did notice was that as soon as I registered my domain in workspace (but hadn’t even setup mx records or began moving mail) a lot of issues with google immediately stopped, and thus, same with Office.com. I actually ran this way for a while but then google axed freed accounts and I just moved my stuff to them and pay.

Maybe because I use a gTLD? I dunno. But it was a headache.

I also host my own mail and there’s been little issues.

Microsoft is a pain in the ass if you’re in an IP space they don’t like like DigitalOcean. Which is ironic because they have the worst spam filter by far in the industry.

If you want to get through to everyone you will have to:

• Use a “good” TLD ( not .to, not .xyz, …)
• Don’t use cloud platforms that are regularily used for spam (mostly DigitalOcean)
• Use SPF
• Use DMARC
• Use DKIM
• Use a PTR record
• Don’t make an open relay by accident
• Use proper ports and certificates
• Register an abuse account at the big players (Google, Microsoft, …)
• Don’t use an dynamic IP
• Keep it up to date
• Minimize downtime

I can’t recommend mailcow enough, it makes setting up a mail server a breeze.

https://github.com/mailcow/mailcow-dockerized

Use the MXToolbox to verify your server(s).

https://mxtoolbox.com/diagnostic.aspx