Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Can I force Big Email Company to not mark my emails as spam?
10 points by kuboble on April 27, 2022 | hide | past | favorite | 19 comments
We are running a business where people buy a product in a webshop and receive an email "Thank you for purchasing the product. Here is your licence key".

Now Big Email Provider are marking those emails as spam. Even though

- every single email we send is an important email that our users really want to read - our email is following all known to us good practices - our domain has never been used to send any unsolicited emails ever - they are simple plain-text emails without links - we usually send not more than 3-5 emails a day to Big Email Provider - we ask our customers to mark them as "Not Spam" and they claim they do so.

Big email providers refuse to not mark our emails as a spam. "We know it's not a spam, but we are still going to mark them as spam to our customers and we won't tell you why."

It's hard to calculate the exact damages but it does damage our reputation because frequently pepole don't check spam and send us angry emails or publicly complain that they haven't received an email for a week. (Our replies to these emails aren't sent to spam).

Is there any precedence where someone has succesfully managed to force Big Email Provider to not mark legit emails as spam or have there been any lawsuits anywhere around this topic?

Edit: we do have proper dns settings, send emails from big email provider ourselves and did our due diligence including discussing topic with external experts etc. For the purpose of the question please assume that we do everything by the book and the reasons for the issue are hidden even from the big company in their ai- based spam filters. We did get response from those providers that "your emails are fine, you do everything right, but our ai marks them as spam and we can't and won't do anything about it"




Obviously, make sure you have dkim, as well as strict spf and dmarc.

I would suggest sending a simple HTML part along with your plaintext part.

Make sure you're handling bounces and if you can do feedback loop programs[1], do those as well.

Consider making your message a bit longer and specific to your business, so if there's any Bayesian stuff, you might get good vibes. Include your business name and url, etc.

Make sure your mail looks like a real email and not a machine generated low effort mail; have a reasonable subject, put the user's name (as you know it) in to To: with quotes, do something nice for From and Reply-To... etc.

[1] https://support.google.com/mail/answer/6254652?hl=en Google program as example, other large services may have similar


Some providers, I suspect gmail here, have new requirements for e-mail. SPF entries, using DKIM to sign mails perhaps.

I don't recommend actually following these guidelines to the whims of Google although it might be good practice for some applications. But this is again not against spam, there are other interests involved here and not necessarily to the interest of the users and the mechanism can be used to make email exclusive overall.

If it is just a spam filter, you might want to include some personal information although the filters have become far more aggressive because they had to. Scammers already use personalized info to send out fake mails. If you have a customer where this happened perhaps ask their IT what methods they are using. Perhaps adjusting the mail for a popular methods of spam filtering can solve the problem.


I only use spf and seemingly they go through. I never send unsolicited mails though for years. In any case i always ask people to check their spam.

It is very irritating of course that professional spammers (i.e. mailchimp) always jump the line


You can use the likes of https://www.mailgenius.com/ to test an email and it'll let you know why it's hitting the spam.

It can be anything from trigger words in the subject/content to the email being image heavy etc.

Usually checking the headers of the email when it's been sent to spam can identify why.


It's kind of a pointless question when you haven't provided any details. Are you self hosting? If not, which provider? Are you using an email delivery service? Have you correctly configured DKIM and SPF?

To answer your question, no you cannot force them to do anything. Either fix it yourself or hire someone who can fix it for you.


Yes, we have all DNS settings configured correctly. We use big email provider to send emails (Gmail & fastmail). We send order emails via smtp client.

We already did try to figure out why our emails are marked as spam including external help and available support channels and do follow all known good practices.

I'm not looking for advice on how to make my email not look like spam to their systems. Please assume for the sake of the quotation that we did due diligence here.

I'm really asking if there is a precedence for someone trying to sue big company for damages caused by their refusal to not mark legit emails as spam.


> I'm really asking if there is a precedence for someone trying to sue big company for damages caused by their refusal to not mark legit emails as spam.

No. Email is an altruistic delivery system per the RFC. You have no right to have your email delivered at all, let alone a right to have it marked as not spam. You should assume your email is still triggering the undesired spam label because of your configuration, content, or delivery approach and try to resolve that.


Some companies use also plaintext emails to improve deliverability https://useplaintext.email/


Plaintext vs HTML email is an interesting one - I found that getting deliverability to MS 365 from a very low volume, self-hosted email server was actually improved by sending in HTML format, as opposed to plaintext.

This was the exact opposite of what I expected, but every other reasonable variable was controlled as much as possible.


My best advice would be to try an actual transactional email service. A while back I was experimenting with this... Absolutely couldn't get past spam despite all the right settings. Then I tried sendgrid. Worked like a charm without any domain configuration. In fact, it would let you use a completely made up domain and it'd still work (because you're really using sendgrid's domain for reputation, yet most UIs will still show the other domain or both)


Are you Elon Musk? Then you can buy the email company.

If not, you must check your email server's reputation, check you are not using trash netblocks like linode, and check you follow best practices like having a sane RDNS for your outbound mail server.

Some places check there is a valid MX that has a sane HELO matching the sender.

In short stop seeing this as 'big email' problem and as a technical quality issue about your mail arrangements which you do have control over.


Where are you sending these mails from? Mailchimp, Mailgun or similar?

Or a self-hosted server? If so, have you checked if you're on a blacklist or something that would lead to providers marking it as spam (or refuse delivery altogether)?

If your customers mark something as NOT spam, I don't think that will do anything at the server end unless they have also set up Sieve?


We used to send them from Gmail, then we switched to Fastmail. It was the same in both cases. We're not aware of being on any blacklists. Our emails are delivered, just land in spam for few Big Email providers.


I send a low volume of "critical" emails from my FastMail account, and have been for a number of years. I never received any reports of them ending up in spam.

It's unfortunate that many email providers no longer provide a detailed reason why something was marked as spam, but considering you say it's with several providers I'm inclined to suspect it's something non-obvious in your setup or email content somehow.


You mention that your email responses to complaints are not sent to spam.

How are those different to the emails that are marked as spam? Therein may lie a clue as to what is causing the problem.

Fighting big companies is a David vs Goliath battle. Except in the world of lawyers and deep corporate pockets, the Davids end up being chewed up and spat out.


This is common [1]. Most email provider detect replies [2] and it's easier to pass the spam filter. Also if the user sends you an email, you are added to a secret allowlist [3]. So we added a message asking people to send us an email if they don't receive our initial message.

[1] And Gmail/Hotmail/Yahoo have slightly different rules/heuristics, so each trick may work in some of them and not in the others.

[2] Real replies, not fake replies that start with Re:

[3] A different list for each user.


Force is a difficult word. You can certainly help, by following their guidance on DKIM, SPF and DMARK.


No some email will always get marked as spam.

It's the world we live in. Your customers are used to it.

Most people are willing to hunt in spam folders for transactional email they care about.


Now it's common that the server just eat the email, not spam folder, no error, it' just disappear without trace. Specially Hotmail/Outlook/Live.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: