This is the third time in a month that an email I send to one of my clients gets flagged as SPAM by OVH, and OVH therefore decides not to deliver it without informing me. I complained to support, but incompetence seems to be their hallmark, so apart from a "we've modified the SPAM filter, this won't happen again"—which clearly doesn’t prevent the issue from reappearing at some random moment—they are neither:
Able to tell me what triggered the SPAM classification (you understand, it’s their security policy, tough luck for you if it causes you a problem mistakenly)
Able to give me any guidance on how to prevent this from happening again
Able to escalate to someone who could do better than their responses from the GPT‑3 free tier.
And I don't know what to do anymore. I’ve already almost lost a project because of them, and every time I don’t get a response within an hour of reaching out, I start freaking out, thinking this service will block me for no reason.
For context:
My SPF, DKIM, and DMARC records are correct (DMARC is in pass, which is fine for most companies without it causing issues)
The content of my emails is standard and shouldn’t trigger any heuristic rule (they are completely random exchanges with my clients). I have no attachments, and my signature contains a few normal links (one to my calendar at cal.eu, and some image links hosted on my own WordPress at ovh[.]net). There was nothing at first glance suspicious—just text and my signature.
My domain sends a normal volume of mail (10‑50 per month), and never more than 3‑4 distinct recipients per send or subject. Nothing that could be mistaken for SPAM in terms of volume.
My Spamhaus score is perfectly fine, my domain isn’t on any blacklist. In fact, it’s OVH—not the recipients—who are blocking or flagging the messages as SPAM.
Of the three example emails I know were blocked, one was to a recipient I had never contacted before, but the others were not; one was sent while I was connected via a VPN, but the others were not.
Even stranger, if I forward the same email to another personal address (Gmail or Outlook) that had been identified as SPAM 10 minutes earlier, it goes through without issue (unlike my client, who never sees it even in their SPAM folder, because the block comes from OVH).
My mail client (Thunderbird) is configured to send via ssl0.ovh[.]net, with my OVH login & password. I don’t want to switch to a dedicated sending infrastructure (like Mailgun, etc.) because I don’t send marketing mails; I just need to communicate with my clients (I’m a freelance consultant) and I’ve already paid for the service.
I’m currently traveling in Asia, ever since I subscribed to the service. Sometimes I use a VPN, sometimes the Wi‑Fi at my accommodation, sometimes my Free Mobile 4G (which exits France). It doesn’t seem to matter—I have examples of emails sent and flagged regardless of the outbound IP.
Has anyone else been in this situation? What solution did you find?
Otherwise, does anyone have an idea, something I haven’t thought of?
My last resort is to give up and move my mail handling to a competitor.
Mail problems are always a huge pain and it’s easy to make wrong diagnoses.
OVH has changed its anti‑spam; they now use Vadesecure.
Some sent or received emails are indeed not routed (DISCARD) and it happens silently.
Personally this also causes me problems because for sysadmin reasons I have to receive all emails, even spam (before we could disable the anti‑spam).
It’s hard to say why an email was flagged as SPAM, especially since now there must be some AI component perhaps.
and some image links hosted on my own WordPress (on ovh[.]net)
Which domain? You mean .ovh? If that’s the case, then it’s a good candidate for increasing your spam score.
I'm sorry that the antispam system is causing inconvenience.
To give some context to this issue, as you can see the use of signing and encryption systems such as DKIM, etc... has been becoming much more standardized recently, because email fraud continues to rise on the internet.
One security measure is to be "scrupulous" with the emails that leave the OVHcloud infrastructure precisely to avoid our IPs being considered as "spammers" and being placed on blacklists.
As @TTY points out, we currently use VadeSecure and sometimes emails get mistakenly labeled as spam (false positives). Since they are an external provider, we give them this feedback one by one (because if they applied a fix that was too broad it could mean that real spam would leave our infra).
It's an annoying problem but not permanent. VadeSecure also acts according to internet threats and sometimes adjusts better and other times worse.
If you are having blocked emails that follow a certain pattern (same sender, same IP, same message body, etc.) we can try to open a ticket directly with them so that your situation gets a more permanent "fix".
I hope I have helped you and given you a bit of technical explanation of this problem.
Do not use redirects. Redirects are the worst; there is a probability (I would say rather a certainty) that emails will get lost, either because OVH fails to forward legitimate messages (false positives) or for any reason, good or bad, on the part of the target server of the redirect, which refuses to accept the mail.
Emails detected as spam are never retransmitted by OVH.
In Zimbra offerings, you need to check the Spam (or Junk) folder.
After several useless and ineffective exchanges with OVH support, they were unable to give me any precise information on why some of these emails were flagged as SPAM.
They gave me a list of general criteria for recognizing outgoing emails as SPAM. None of these criteria applied to the messages I sent, which were nevertheless identified as SPAM.
I was forced to sign up with an email‑sending provider, which gave me SMTP access. Today I receive via ssl0[.]ovh[.]net but I send through the SMTP of mailersend.
It’s really ridiculous that OVH sells an email service yet blocks its own customers without even warning them and without being able to guide them in resolving false positives. At least I now have a solution that works...
I’m referring to the OVH domain name of the server (in clusterXXX[.]hosting[.]ovh[.]net). I find it strange to flag links to its own architecture from OVH as spam, but that’s all I see.
Or perhaps it’s the fact that I’m connecting from certain Asian countries, since I’ve been traveling since I opened the account.
Still, I think it’s ridiculous to drop sending an email just because a user connects from a suspect region of the world (and not simply alert them. If we consider it potentially malicious, that implies the malicious actor has the client’s password, which I believe the client deserves to know), or links in ovh[.]net...
I've been fighting against this for years. It's a professional misconduct on OVH's part to steal mail while OVH knows exactly which client it's /dev/null'ing their mail for.
Clearly, the discard feature for a professional email account deeply annoys me as well.
Free services like Outlook, Gmail, etc., discard without warning—in fact, they do whatever they want, and that’s exactly what they do.
But in my view, emails should be delivered, even SPAM.
That being said, given the security situation we’re dealing with right now
Personally, I've stopped hosting my own emails, and I'm gradually migrating my clients to shared/standardized email solutions.
Drops should never happen, we agree. Either the email should be delivered, or the sender should be notified.
But in some situations, this can involve dozens, even hundreds of messages per day for certain mailboxes. And at OVH's scale, everyone can imagine how horrible it must be to manage.
What I recommend is not to include external links in these messages; it's VERY annoying, but that's my recommendation when there are occasional delivery problems.
For my part, I'm amazed that email still exists today in the form it was conceived decades ago. It has become a nightmare for everyone. Users who get spammed, admins who have to handle staggering amounts of spam. Email is even worse than the web, which gets flooded with bots...
