Wanting to make sure I've done everything I can to prevent our legitimate email from being incorrectly marked as spam by other companies:
Exchange 2010 with SP1, no major changes in 6 months.
I've run email for this particular 30-user business for nearly a decade, but the last few months their emails are starting to be caught by their customers/partners spam tools. These are human to human emails outgoing from my Exchange servers to other companies. When one or two starting saying company X or Y isn't receiving my emails I think it's a fluke, but now it seems to be up to 5 people internally having issues with 1/2 a dozen of their customers/partners on every email, sometimes just new emails, sometimes replies.
- I have a proper PTR record matching hostname
- My email domain matches MX domain name and the EHLO response
- On no blacklists according to mxtoolbox.com
- senderscore.org is all good and shows score of 100
- SPF/Sender ID setup correctly
- Exchange 2010 doesn't support DomainKeys so I'm not doing that
- No spam or marketing emails are sent from this server/IP
- All sending email is human generated
- IP is from a colo not residential ISP
- no complaints of email not getting to email providers (hotmail, gmail, yahoo, etc.)
Any further ideas? At this point all I have to say to people is "totally out of our control", unless I want to work with each company to find out their anti spam vendor and contact each one for resolution...
Thanks for any ideas!
I've seen issues crop up when the company asks all employees to use a specific 'signature' block on their message that includes a gif or jpg (like the company's logo). That little embedded image can pump the spam score up needlessly.
It wouldn't hurt to ask one of the affected employees to resend a message that was caught in a spam filter to a Yahoo or Gmail account and then inspect the header after it arrives there. It could reveal those little issues that are below Y/Gmail's thresholds, but that may be over the threshold with smaller outfits that are being overly aggressive in filtering.
You need to find out the who/what/when/where. You could enable smtp logging if you are sending directly from the server. It could be something as simple as a signature. I once had a user that used the word "specialist" in his sig. ---> speCIALISt ---> CIALIS see what happened there.