We have a 99% email reputation, but Gmail disagrees
blogfontawesome.wpcomstaging.com348 points by em-bee a day ago
348 points by em-bee a day ago
How do you get email addresses? Do people freely and explicitly choose to sign up to your mailing list, or is it baggage that you're forcing on them without their consent?
I notice that when I go to https://fontawesome.com/ and click "Start for Free", I'm asked for my email address. This isn't necessary for me to use the icons. I just need a page that tells me to add the necessary tags for cdnjs [0].
I think your problem is dissonance between what you think your users want and what they actually want. If I had to sign up for a mailing list in order to use every frontend development library I've ever used, and their emails actually made it past my spam filter, I'd never see anything else.
I think Google's doing the right thing here. You need to separate your newsletter and product updates from people who just want to set up the icons and move on with their lives.
I don’t know if this is true with Font Awesome, but more and more companies are spamming my inbox despite disabling any promotional emails in their settings.
So, I mark any unwanted email as spam in Gmail immediately, and even leave bad reviews.
Having my email address is not the same as having my consent. Stop trying to roofie us with malicious EULAs.
I remember there was a thread some years back with an article complaining that you get emails immediately on sign up, but that it can take up to 10 days to stop receiving emails when you unsubscribe.
One spammer said they could use the same servers for both but when you unsubscribe you have immediately signaled that you are now losing him money. So he uses the slowest cheapest part of the stack for removal. He will never fix it and doesn't care if you get some more spam after you unsubscribe since he has done the bare minimum.
If I get a single email after I've unsubscribed I go back in my inbox and mark every single email I ever received as spam.
My phone network provider ran some "12 days of Christmas" promotion last year which entailed a spam email trying to hawk me crap I don't need every single day. When I tried to opt-out they told me it would take a month. I emailed the ICO and the network provider's complaints team and miraculously they were able to remove me from the mailing list immediately.
> but that it can take up to 10 days to stop receiving emails when you unsubscribe
This is a really bad business practice, people will just mark your mail as spam and the likelyhood of other people seeing your mails will drop
I still do not understand how marketers haven't understood that quality > quantity.
But often $ generated by quantity > $ generated by quality. And that’s the metric everyone really cares about.
Inside the marketing org bubble, quantity is the "any moron could see that" metric. So anyone who wants to get ahead, inside that bubble, had better be willing to optimize it.
> One spammer said they could use the same servers for both but when you unsubscribe you have immediately signaled that you are now losing him money. So he uses the slowest cheapest part of the stack for removal.
Hmm, wouldn't you want to remove the money losing people as soon as possible, so you don't waste even more money on them?
Doesn’t compute.
He probably meant that “customer” is not making him money, therefore not worth the time. The only reason unsubscribing works at all is probably a legal requirement.
It does, probabilistically. For any given customer C, the expected value EV(C) is... not much. By starting an opt-out process, a customer Co is revealed to have EV(Co) = 0, which is less than not much.
It’s because transactional email and marketing email are two different systems.
That's not really relevant here. The complaint is that you start getting promotional emails right away, meaning that adding you to a mailing list is instant, but removing you somehow takes ten days. Normally you can't unsubscribe from transactional email, as they serve to provide you with information you're legally entitled to. There might be companies that are foolish enough to use the same system for both transactional and marketing email, but normally you'd never do that, because you exactly risk having things like order confirmation, recalls, invoices and so on, be tagged as spam, if it uses the same system as the marketing emails. Frequently you can use the same provider, allowing for tracking bounce rates, open indication and so on, but even if it's within the same interface or set of APIs, the two things are kept very separate on the backend. They'd at least use different email addresses, but frequently also different domains/sub-domains.
I've done both transactional and marketing emails, and I've never seen a system that could not remove a user at least within 24 hours. I can imagine one, but you're doing something very wrong at that point. Ten days is deliberate.
As the end user: not my problem, I don’t care, I don’t need the implementation details.
I only care about what I see.
Sounds like an engineering problem that can be solved, and, more importantly, not my fucking problem.
This is when I send them a GDPR data request followed by a deletion request. If they send me spam after, I sue them with the receipts.
> If I get a single email after I've unsubscribed I go back in my inbox and mark every single email I ever received as spam.
Fuck me, that is brutal and could absolutely ruin your SES complaint rate - even with the suppression filter on, as the emails are already in your inbox.
When I worked on a notification system that sent over a billion messages a month. We received spam complaints on emails sent 6+ years ago. No correlation, just a one-off spam complaint. I always wondered why this was happening.
Probably because people like me finally had some time to go through an inbox with 20,000 unread messages. Almost anything that's been unread was either (most likely spam) or (very rarely) just simply unimportant.
Good. Don't send spam. If you're sending spam, then you clearly don't care about your complaint rate.
I have done the opposite We had a million people enter their email over the last decade We haven’t messaged a single one.
Now we plan to start sending out a newsletter. For many, they may have forgotten downloading the app, but they might still appreciate it. If not - they can u subscribe.
Don't do that, it will be disastrous for you.
Instead, send them a politely worded one-time announcement with an invitation to subscribe. Clearly mention that if they don't, this is the last mail they'll get from you, and keep that promise by deleting their address. You'll still get some pushback, but I think most people would find that acceptable.
At least with your suggestions there's some chance that their newsletter won't instantly get flagged as spam.
I'd do what you suggest, but send the newsletter from an separate domain once subscriptions have been confirmed.
That one-time announcement is called an email. And therefore that first announcement itself can be flagged as spam.
And naturally, unless they click a link in the first email, gmail should consider anything subsequent to be spam anyway. They have no idea whether consent happened somewhere else or not.
The unsubscribe links must work without even opening the email, according to gmail rules.
What I'd be concerned with is that if you have never sent anything to these users, they might have forgotten where and when they gave you their email address and simply mark your message as spam.
We've trained users to not use "unsubscribe" because some spammers once used that to verify addresses, or they may simply click "Spam" because they forgot who you are and think you got their address illegitimately. Gmail also doesn't make unsubscribe as visible as "Spam", making flagging the easier option. So now Gmail will see some percentage of users manually flagging you as a spammer, tainting your sender. This is why I'd switch the newsletter to a new domain or at least a new sender address. That does mean preparing that new sender, give it a bit of time to mature and send a few emails to Gmail accounts you control and ensure that they are not flagged as spam.
Probably also test with a list of Gmail account you control and check if you're tagged as spam and fix that, before doing the big push.
As a gmail user who may or may not have had to enter an email address to do something on the web, and who gets annoyed by spam, let me describe my decision points (anecdote is not the plural of data, of course, but here I am) when it comes to "unsubscribe" vs marking something "spam."
If your email reminds me (upfront!) how and when and why I specifically gave you (and not some other third party) my email address, and promises that you are advertising this newsletter one time, and it is opt-in, and you keep your promise, I am highly unlikely to mark it spam.
Now, this presupposes that it was really me who gave you my email address. I have a fairly generic email address because I got on gmail early. There are many variants of it, but sometimes people forget to add the trailing numbers or letters, so I get misdirected email all the time.
If the misdirected email is personal, I usually respond letting them know of the issue.
If the misdirected email shows a clear understanding that I might not have been the one who really signed up then I give them a pass.
If the misdirected email blithely assumes that I am the one who signed up, then I blithely assume that its senders are too fucking stupid to use the internet and it goes straight into the spam bucket. (And this is usually an easy call because they use the name of the person with the similar email address, which is not my name. My email address is firstinitiallastname@gmail.com and there are many different first names that start with the same initial.)
Any failure on any of those other points starts to increase the likelihood of it being marked spam, and...
> The unsubscribe links must work without even opening the email, according to gmail rules.
So here's where I'm a hard-ass and maybe even worse than google's rules.
If I see the RFC8058 unsubscribe link, it is too late. I only notice that link after I've decided to mark your email as "spam" and google asks if I'm sure, or if I merely want to unsubscribe.
Why did I decide to mark your email as spam? One possible reason is that I read through it, decided that the sender legitimately had my email address and was acting honorably, and then clicked the unsubscribe link embedded in the email.
When I do that, one of two things happens. Either I get some form of "thank you, you've been unsubscribed" or nothing happens because the sender assumes that I am OK with them executing javascript on my computer.
This is a privilege I jealously guard and only reluctantly offer to as few websites as possible.
Even if I previously gave you my email address, that did not come with an open invitation to use my computing resources for your own purposes.
So by your own description, ANYONE sending you a newsletter, by complying with Google’s rules, they piss you off and make you mark their email as SPAM because, according to you, they made “javascript execute on your computer”. Actually, gmail is the one executing tons of javascript. The mandatory unsubscribe LINK uses HTTP, not even HTML. Google just requires that the unsubscribe instant.
It is an unwinnable situation.
With all respect, why would I care what an impossibly hardass tech person would do if I sent them an email in an unwinnable situation? The vast majority of our users are not this technical, let alone a hardass HN denizen who advertises the fact that the mere compliance with Google’s rules will piss them off due to a misunderstanding of how unsubcribe works.
Here is what we might both agree on: email sucks. You shouldn’t be reachable by anyone who just has your address, and it is not your job to be vigilant. Then all these problems go away.