Ask HN: Are you also getting extremely obvious spam bypassing Gmail's filters?
For the past weeks I've been receiving emails that are pretty obviously spam. Here's one I just got:
sender: Динасий Колпаков <[email protected]> subject: Q7425 7235 F0 8741 (empty body)
They all have similar formats, with a .htm attached file with ridiculous names like "Elon secret invitation" or "how to get free bitcoin".
They are all look like 90's era spam. Yet not only aren't they caught in the spam filter, they arrive to my main inbox, they aren't even classified as promotions or anything.
I can also see a long CC list, since it's not hidden.
Are any of you also having a similar problem?
(
282 comments
| 575 points
| on HN
)
Add to that that when that recruitment system was implemented, the recruiter probably didn't bother updating/customizing the default templates, making their emails just the same as everyone else's.
Then for every email interaction with a 3rd party you make a unique address like [email protected].
You can then just add rules for each incoming mail domain to send them to a junk folder, especially if it was was a one time transaction and subsequent emails are just going to be marketing junk.
It also lets you know who has had their mailing lists compromised as there is virtually zero chance someone would guess the email address like the example above.
Does that work for you? Have you caught many companies sharing your address?
I’ve been making unique addresses for a decade, and it has been a complete failure. I haven’t caught any companies sharing them, and it hasn’t stopped spam in the slightest. I’m not sure where spammers get my mail email address from, but they do, and so far I’m not seeing them get the individualized addresses, like ever.
Before setting up a catch all domain I had 3 distinct email addresses I used as high/medium/low priority which helped, but its much simpler with the catch all domain as it goes into the same inbox, just with different "To:" addresses that you can easily filter on - I just right click in outlook and select move all mail to recipient to junk.
To be fair most of the mail I get is marketing and I could probably unsubscribe, its just easier to make a rule to filter them to junk.
Literally everyone claiming they never miss emails in spam, never have delivery issues with their custom domains or mailservers, etc
If you are a corporation using gmail then you will likely have emails going into a shared mailbox or public folder and have team members triage emails as sales lead, spam, upset customer, etc.
Interesting how people can have such different experiences.
If you're a regular unsubscriber, your email box is only ever filled with relevant important items.
I do the same thing with my phone notifications. If my phone ever gets a notification, it's something that directly concerns me.
I don't know how people live getting bombarded with stuff all day, but like you said, everyone's different.
“Unknown / trusted / spam senders” lists are a basic implementation of this concept.
Gmail thought that adding Snooze was enough to get feature parity and kill Inbox, but Bundles were by far the feature keeping me on Inbox.
i get so few of them now i removed all those extra tabs and filters a year ago, but it was pretty useful to get my total inbox as clean as it is now :)
Unfortunately, all I'm probably doing is just confirming my email address to generate future "mailing list" subscriptions.
I drag them to the proper tab each time, but it doesn't seem to fix anything.
At the same time, all sorts of ads fall into my inbox repeatedly, even I deliberately mark them as spam or junk.
( Being cautious, I just checked the promotions again. And there is another important email lying. WTF!)
What I did to reduce the noise in my mailbox is to unsubscribe from all marketing emails, and I move the ones that still come through to spam folder. It was a bit tedious at first, but now gmail is doing a pretty good job at automatically filtering out senders that do not respect my request to unsubscribe.
Am I missing something obvious?
I'm really surprised most people aren't realizing this.
I haven't heard anything about this when I asked colleagues / IRL friends... I wonder what is going on over at Google
Except for the last one, all the calls to action seem to be crypto related, but I don't know if that's relevant to the origin of this attack. Perhaps it's just the most successful way of getting clicks nowadays (?).
It's baffling that google are letting these ones slip. Even marking some as spam does nothing to prevent new ones from coming.
Title: Rc 2 Xq 1677 Riyw 532
Sender: Чеслав Сальников <[email protected]>
Body: 3688 867 8383 4 3 784
Attachment: Free Bitcoin - DELL .html
Title: V10 YU L2 RUO T778 ZRD
Sender: Тельман Кудряшов [email protected]
Body: 3484 2 0812 61 3
Attachment: Blockchain Prizes 3883 .htm
Why would it be baffling? Google’s spam filters have always been extremely low quality, even mistaking email generated by Google itself as spam. They simply don’t have the pride of craftsmanship to improve it.
In my world, spam was a true problem that was killing email. Truly killing it. You would have to create a new address every 6 months to stay ahead of the spam and have a functioning inbox. Very different world.
Gmail fixed that. And for me, there have been issues from time to time and in fact i am getting this very spam myself, but gmail permanently stopped the hordes for me in 2004.
Now, don’t get me started on their insane categorization choices.
Hardly. They didn't invent the tech. And they didn't perfect it. They haven't even tried. The evidence is in my Spam folder. I mean they mark messages they themselves originate as dangerous spam. "Dangerous" being their word, not mine. Screenshot from within the last hour:
https://imgur.com/IjQUGaY
I was in that earlier world too. They papered the problem over, a little. But no, they didn't fix it.
The return address is:
Google Alerts - googlealerts-noreply (at) google.com
Now, my impression is that all of the well-regarded email providers do a good job of handling spam—it's no longer a differentiator. (Although it might start to be again if Google has genuinely let their guard down.) Fastmail does a good job in my experience.
Probably a whole lot of dont-give-a-darn-about-e-mail, because it's not new and sexy, and likely doesn't drive revenue.
Also, the people who suffer from gmail spam are often non-users of gmail. I.e. neither customers of Google, nor targets of its advertising, nor sources of personal information.
Google doesn't care if <[email protected]> is getting Gmail spam, because that's an outside entity whose existence does not benefit Google.
Plus, Google knows that Gmail is so huge, that nobody can just block all of Gmail. Unlike some small-time mail domain, they do not feel any risk that, if they don't take action to combat spam, they will be blocked as a whole.
If a small-time domain's machine gets listed in some DNS black-hole lists or other dynamic anti-spam databases, they have to care, or they don't get to send mail. It's a dire situation to which they have to respond.
If a Google machine gets listed in these databases, Google doesn't have to care. Anyone actually blocking Gmail machines is essentially just cutting themselves off from a huge e-mail communication hub. It's almost as if that operator were blacklisting itself.
Small fry: OK, that does it, I shall not receive Gmail!
Google: Hahaha; say bye bye to more than half your contacts, then!
In other words, Google knows that e-mail operators who are using blacklists have to pretty much whitelist Gmail servers, and so it doesn't care about blacklists.
Interestingly, they can change the font of the subject lines which no valid email I have ever received in gmail has a subject with a different font.
That would be an interesting filter: if subject is !Font, then spam.
It's now happening regularly with emails from people in my contacts with whom I regularly exchange messages.
Mind-boggling. I know spam filtering is a hard problem, but these are just obvious misses.
20% of what is in my "Spam" folder today is what I'd call "spam" in the classical 90s/2000s-internet sense. Obvious trash/scam stuff, usually sex-related.
Most of the rest of my "Spam" looks like what Gmail usually just labels Promotions. It's mail from legitimate organizations that I did indeed give my email address to and have a reasonable expectation of getting semi-regular email from, even if it's just trying to sell me more stuff. The Promotions auto-labeling works (worked) just fine for managing that stuff.
I figured enough users are clicking the "Spam" button on enough "legit promotional" email from real organizations that they did agree to receive email from, that Gmail just started classifying it all as spam, and now doesn't/can't distinguish between "classic" spam and "annoying emails I can't be bothered to unsubscribe from". Sort of a tragedy of the commons of crowd-sourced spam filtering. But maybe there's a better explanation.
I try to use unsubscribe links when 1) they exist and 2) the email is from a legit organization that I reasonably believe is actually going to unsubscribe me from something (if not from every place they've copied my address to by now), but I can see the argument for just flagging as spam in the case where an organization oversteps previously agreed bounds. Especially with the political campaign emails, the unsubscribe game can feel like futile whack-a-mole.
For ordinary commercial campaigns, I'd worry that flagging as spam would cause legit transactional mail (i.e. mail that belongs in the Updates auto-label) from the same organization to get flagged as well, but based on the state of my Spam folder, that's already happening anyway...
If an unsolicited email contains no unsubscribe, that is bad behavior and they get spam tagged immediately.
Fun fact: from time to time I still have mails in my spamfolder... seems the filter does not always get applied.
For the promotional stuff it's the report spam and unsubscribe button every time. I've worked in a small business that bothered to do double opt-in signups, so don't waste my time because you can't be bothered due to some vague metrics. It's spam from my POV.
For the rest it depends on my mood. If you have a noreply and it's going to take effort to reach customer service (more than an email), spam. I'm of the opinion I should be able to reply to any email and it's rude to shout at someone and then block your ears, but maybe that's old fashioned.
I once caught my boss doing this (he was not a native English speaker, but absolutely used to communicating in English so it shocked me.)
I do that all the time. I consider every list I didn't explicitly and intentionally subscribe to as spam and treat them accordingly. I wish I wouldn't have to do that but I find subscribing someone to a list when they thought they were just buying or creating an account for a product so unnerving and disrespectful that I don't feel bad about it.
Possibly the worst is not allowing replys. I mean if a customer sends me a message, and you block the reply as spam how is that serving that customer? Sure mark it, remove viruses, obfuscate links, but let me reply to someone!
Clearly shows they give more weight to things like scanning, IP reputation, etc vs common sense.
Funny enough, the mails it penalizes worst are GMail addresses for small businesses, like my vet or the pizza shop.
For me privately, switching to a better provider solved that.
But having >25% GMail customers and always landing in Spam is horrible. Pretty much any other provider likes our mail server, but GMail always says spam.
Then you're going double opt-ins but customers still mark mails as spam because unsubscribing is too hard. Thanks for nothing.
There is actually an industry in gaming GMails spam filter to somehow get into the inbox: some offer automated replies and unmarking spam, some manually run hundreds of mailboxes and don't do anything else all day than unmarking mails as spam.
I feel like it used to be near perfect. Something must’ve broken the models.
If this was all part of some clever conspiracy then it would be, well, more clever...
My personal account is pretty clean and rarely does anything get through that shouldn't. Occasionally I find something in the spam filter that shouldn't be there, usually password resets.
My work account is seeing a big increase in "professional" spam. "Hey Guy, did you see my last email I sent . . .", "Hey Guy, we are the top network security consultants . . ." Many of them are getting tagged Important. Some of them are so left field it doesn't make any sense that they are listed as important. Here's a good one:
"You asked for it, and we made it. We are delighted to present a complete redesign of our Merge Rules in Duplicate Check." - coming from mta.exacttarget.com
How is that not spam and how did it get tagged important? I have no idea who this company is. I've never done business or corresponded with them.
As I write this I'm coming to the realization that I'm a mechanical turk working for Google to find and report.
What's interesting is that many of the emails are coming from clear email marketing sources like HubSpot and Exact Target. Why would those get bumped up? I also notice some coming from something like xxxx.outbound.protection.outlook.com - not sure what that is exactly, but it mostly comes from companies directly marketing using their spf and dkim domain, but seems to passing through outlook.com.
Email is dying but will never be dead.
Those servers send a very high volume of legitimate marketing email that is not spam. Those services do a lot of policing within their platform to prevent or discourage use of their servers for actual spam. So when someone does manage to get spam out through that service, it often gets past more filters downstream. Theoretically, Hubspot or Exact Target will get notifications and shut down those senders so Google trusts them, gives them some leeway and doesn't ban the servers outright.
But what is much more interesting is Gmail is pushing the spam up by tagging it as important and it ends up at the top of my email list and not in the lower section of everything else. It should be in the spam folder with all the other junk, but out ranks internal JIRA tickets, daily sales report etc.
Could there be an implicit or explicit incentive? Am I the spam detection?
from: kate.random.but.reasonable.name38 @ gmail . com
Sorry for the delay, here is my email.
Best, Kate
Sometimes they are more obviously spam, but other times it seems nearly impossible for a spam filter to stop, like the example I provided.
I was always assuming that if I reply, then they would send the spam links, but your theory is better.
I think Googles spam detection is a bit too much lax when the sender itself is using gmail.
These might as well be hacked accounts which have already proven themselves to be valid and "human" at a previous point in time? I doubt gmails spam detection would let a brand new account spam CC'd emails without any sort of detection.
a) replies to emails I sent b) have anything to do with topics I'm actively involved with c) from senders who I have marked as not spam dozens of times
I never used to check the spam filter, now I do almost daily.
Fun fact: Back in 2002 ycombinator founder Paul Graham wrote an article on spam filtering. (See http://www.paulgraham.com/spam.html ). I emailed him that his method can be defeated by sending an image of the text, as opposed to the text itself. PG replied and pointed me to this FAQ: http://www.paulgraham.com/spamfaq.html
In the FAQ there is an entry named "What if spammers sent their messages as images?" The answer indicates that is not going to be an issue, because there's still plenty of signals to go by.
Guess PG was wrong!
I mean, he wasn't wrong. He was describing the state of spam at the time.
The way I see it, you were trying to corner him into giving the answer you wanted, and he was trying to corner you into asking the question that would force him to give the answer you wanted.
All machine learning models make intrinsic assumptions about their input distribution (no free lunch theorem and all that). If all your objection was that the model will fail to work if the assumptions it makes about the input distribution will fail to hold, that is a criticism for the entire field of machine learning, not this one spam-filtering product.
onload="document.location.replace(window.atob('aHR0cHM6Ly9ibG9jay1jaGFpbi1ib3gudGsvbXpwaWwvP3RldHRoa3Yg'));"
Which if you decode you get a strange domain.
I assume gmail only looks for urls which in this case is not visible without decoding it
"Profitability 28388 .htm"
Maybe these emails are coming from real users that got hacked? That's probably the easiest way to get past the filter.
I have also been getting several of these super obvious spam mail messages recently.
Also, I use a desktop email client (Spark on MacOS) with IMAP/SMTP. Massive improvement over any webmail client, especially GMail.
The administrator of my email servers has a vested interest in making sure I get all valid emails and in junking the garbage.
Gmail administrators are interested in pleasing their boss and getting paid -- the quality of their spam filters and especially your interests aren't even in their top ten important issues.
The only problem with my set up is that if my email administrator gets bored, I'm screwed. Fortunately, that's unlikely as I am my email administrator.
I expect that many folks will dismiss this email as just some rando who doesn't understand just how important it is to have gmail or some other provider's email service because reasons.
But the truth is that quality service is based on having the right incentives. And Google (and by extension, their employees and contractors) have zero incentive to consider the needs and interests of their product.
And why should they? If you're a car salesman, do you worry about how a particular car will feel if sold to a jerk? If you're a barista, do you care if the lattes you make are consumed or thrown away?
I could go on, but I expect I've made my point.
I keep marking them as spam but more keep coming. About 1-2 per day and of varying content but similar visual layout. 90s era spam is a good description.
This last month, maybe two, I've had extremely obvious spam hit my inbox repeatedly. Picking two cases from today, the subject is the same "FWD: FINAL CALL", from two different senders, "A P P L E" and "NET FLIX". The pattern is pretty much always the same, it's immediately obvious that it's spam. No idea why it's slipping past when they're still catching hundreds a day (I've had this hotmail account from the early days of the platform, used it a bunch all over the place)
If all the bigger providers are having trouble too, perhaps there’s a new kind of spam? Or higher monetary incentives now?
What I did was to switch from Gmail to a paid email provider. Then I started giving every single business a unique email alias, though my friends all get the same email alias. Currently I have 370 active aliases. I've had to disable only 20 aliases in the whole decade which works out to only about 5% of my contacts.
As I said, I use no spam filtering whatsoever, so I find it amazing that Gmail users with spam filtering have such a different experience.
https://imgur.com/4efNttg
Another kind of spam mail like this is they use google docs or sharepoint to add a huge CC list.
However, defending them is hard and I think I can share my knowledge on topic of spam filtering(given I run an email forwarding service (https://mailwip.com) and have to deal with spam a lot)
When an email come from gmail/hotmail(any popular free email service) itself, it's harder to detect spam, especially if the email is in a non English language.
It has a few way to flag spams:
- Look at the IP address: - Look at the structure of emails: follow best practice, such as html/text plain part, has right mime encoding etc - Look at header of emails: no weird header, no "bad" ip in received chains - Look at attachment file type, virus scan those attachment - Finally, tokenize content of the email to find similar email that are flagged as spam
When the email come from their own IP, send out by gmail themselves, email format looks good, DKIM/SPF all pass and this is the first kind of email then the only way to flag spam is by analyze content. And if the email is in non English language, it's harder to analyze. Especially if not enough people flagged it as spam to train the naieve bayes tokenizer then we're out of luck here. The long CC list looks like a legitimate indicator for spam, but librarian/scool has a tradition of sending out a huge CC of entire class/department, sometime even BCC which make the email looks very suspicious (undisclose recipient) yet they are legitimate email so the CC alone cannot easily be used for spam indicator.
Yet, at the same time, legitimate emails form your own server get flagged because low reputation or a history of previous owner send spam...
I thought that Google had some sort of metrics and would spot this issue if it was widespread, but it seems they don't even bother.
Do they even care about Google search and GMail? They're getting worse and worse.
A random IP is the first text in it, and a vast majority say they originate from .co.uk. It's generally trying to get me to participate in a survey, or I've won a card from Home Depot (a big DIY supply/hardward chain in the US).
(edit) I just got one of the Russian ones noted by the OP; first one I've gotten so far.
From: [email protected] To: several apparently real Gmail addresses Subject: (no subject) Blank body apparently, PDF attached which the web and mobile previews indicate is of a pretty young woman, possibly porn (I've not opened any).
Another variation, the same as above with random words as the subject, such as "shallop escape clause unfilled orders crockitude". This is an actual example from my inbox.
Another, same again, but with nonsense in the subject and body, e.g. "48046256 of hlbezmy". Actual example again. I noticed that the preview image of the PDF in this example has a yellow-green tinge, possibly an attempt to disguise flesh-tones?
And so on. And on, and on, and on. It smacks of a very deliberate, industrial scale approach either at training the filters, or at exploiting weak spots that have already been discovered.
As another commenter has pointed out, this spam really does feel like 90s spam all over again. This made me wonder, is it possible that 90s-flavour spam is a result of 90s-flavour filters falling out of favour and being dropped or becoming under-trained? Are we doomed to endless cycles of spam revivals?
Since 2018 I am getting almost a spam rain, see statistics & logs of spam in this published google sheet https://docs.google.com/spreadsheets/d/e/2PACX-1vSdZyRvDd0ES...
Almost every email has no text, just an image, with a text footer with Unsubscribe. But clicking each unsubscribe tells spammer this email address is hot & live. I made this mistake of clicking Unsubscribe few times in 2018. Now about few weeks ago I loaded images on one of the email by mistake.
Domains these emails sent from are typo, subjects are stupid offers.
A filter adds a label to these emails & deletes them. A google apps script runs once every 24 hours, fetches data from these emails with label X in trash, adds that data to Log Sheet I linked above. Then script removes the tag.
Some other examples: Чарльз Некрасов <[email protected]> Фома Авдеев <[email protected]> Порфирий Угримов <solodqez@g...
Fortunately they typically set a "X-phishing-testing" type header so you can make a rule to get rid of them automatically.
We get tons and tons of gmail spam inbound to our non-gmail email server - every single day for years.
If I complain to Google we get more, its uncanny.
Seriously, fuck Gmail. Gmail can die in fire. Biggest spam service on the internet.
These A/B tests are quite common for products like these to test new features, such as a new spam filtering algorithm. Or even test the efficiency of an existing spam filtering algorithm by degrading the experience of a subset of users.
[1] https://en.wikipedia.org/wiki/A/B_testing
Some years ago a not so bright product manager decided that dots are not important.
My address used to be [email protected], and I've been receiving email for [email protected] for years.
I analized 2 arrived this morning: both have 2 message-id header, one from gmail the other from sendgrid; both have a in-reply-to header (that is used for ARC and DKIM signature); both have 2 DKIM-Signature... both have the same CaMelCasE html... both are flagged as spam by a corporate antispam and my own spamassassin... one message has fake Received: headers with date in year 2019...
gmail has all the information to mark them as spam
Gargh.
What's more annoying than the emails is that I'm getting Google drive documents shared with me with in a similar format from Russian senders. There doesn't seem to be a way to block or spam sort Google drive docs.
I use a Bayesian filter trained on old spam (I haven’t retrained it in years) combined with Spamhaus blocklists. My guess is that spammers found a clever way to fool the Bayesian filter, or they found new hosts to send email from in order to bypass blacklists.
I heard that one gang makes/burns 1000+ gmail addressed per day.
However, even weirder is a series of emails I've been getting for nearly a month now. They are always from randomized emails and claim to be different people but they are always about someone named Becca and some guy. Sometimes they are accusing me of being this guy and harassing Becca. Sometimes they are saying someone stole Becca's phone and was impersonating her friends. The weird thing is that none of these have any links.
Here's one from yesterday: "Beverly hacked all your stuff made fake account of you on Facebook apparently you have a new house and going to hurt Becca if she doesn't stay away from you . Apparently according to the rude message Becca Lynn received y'all aren't friends anymore ... That made Becca cry " But there aren't any links or anything like I would expect if it was typical phishing emails. Maybe they are just sending things out hoping someone will respond?
https://www.bleepingcomputer.com/news/security/new-elon-musk...
I find that every year or two there's a spammer that figures out a new technique and gets a couple of emails past the filter, but usually no more than a couple, and none recently. I don't see a trend toward less or more effectiveness of spam filtering overall.
And like you, obvious spam sits in my Inbox without being flagged as spam. It seems on a long enough timeline, spammers just get better at making email look legit.
Simple filters in free software don't do much unless you spend your energy tweaking it, sophisticated filter like Gmail's has a hidden agenda against you.
It's very obvious; I'm sure the gmail team must know about it. I'd be curious to know whether they're planning on returning the spam detection back to the previous low false-negative rate.
Of course they do. Its from all of the phone calls they've been receiving from their help lines and helpfully answering? Or maybe the bank of humans they have responding to emails sent to support.
I tried Googling for a way to block mail in a language or from a country, but apparently the technology just isn't there to do so(/s). Getting rid of Russian and German mail would completely fix my spam problem.
You could also try another provider or DIY. Another option is something like Mimecast which you put in front of Gmail - https://www.mimecast.com/products/platform-selection/mimecas...
If you use a mass email provider, you get a rather generic experience. How on earth are Goog n co supposed to work out what you consider to be spam or UBE? I imagine that you get some basic filters (naughty words and a few IP deny lists) and a tokenizer n bayesian classifier that does the really hard work. Keep tagging the crap mail as spam and after a while it should learn, that for you: German - probably unwanted, Cyrillic - probably unwanted etc. Unfortunately you don't get to see under the hood/bonnet to understand what is actually going on.
I recommend that you stick with manually notifying spam for a while and see if that works - you need a lot of samples. If Goog allows you mark ham then use that too but I suspect they implicitly mark a mail as ham if you don't mark it as spam. When I say a lot of samples, I think you need at least 200 ham and 200 spam samples minimum, ideally 500 each and the more the better. Moving to another mass mail supplier will almost certainly not help.
We're out of luck until then.
It does not care whether it is in German or Cyrillic, it will learn the characteristics gradually. It does require effort.
If you teach a bayesian classifier with around 500 ham and 500 spam, you will see very little spam and if you continue to teach it, it gets better and better. I am assuming that is what Google give you but if I was them, that's what I'd do. I'd also add a few lists and stuff but the gold standard is a trained bayesian filter. You get to do the training, there is no shortcut. I suspect that if you don't mark a mail as spam then it will be implicitly marked as ham.
Getting an efficient spammy feedback mechanism working in a mail system is surprisingly hard. Email changes at each hop as it gets from source MUA via MTAs to the destination MUA. Headers are added at a minimum at each hop. Anyway, that's my problem - not yours!
So, no they do not need to run up a language detector but given that Google have an online translation service, I doubt that would be tricky. That sort of thing may be added to an "enterprise" offering.
Try teaching your mailbox what you want and don't want and see if the clever buggers at Google have actually mastered the basics. They probably have but you need to do the work to provide the data that corresponds to showing what you want.
Very weird to see the spam filter be obviously and dramatically worse than it was 10 years ago. Are they about to move to a paid model for Gmail? Hehe
But they seem to have stopped... for now. It was quite aggravating.
Phones must be insufferable for most people.
It's pretty rare for me to receive more than a couple notifications a week at most.
Something's up here.
I also have to rescue valid List mail from the spam folder though.
Subject: X 6673 B 11 KZPV 5 V 6 BO 720 XGWH 8
Body: EC 3753 YK 6321 UXNS 487
Most are German, but now a lot of non-english letter based ones in last few days.
I'm guessing i'm part of the new test spam audience.
brutal.
filename:html filename:htm
The most obvious 2000s era spam is coming straight to my inbox.
No, No one wants to live in Arkansas doing contract to hire for $30/hour working on some windows mess.
This is just junk mail
Here's a thread where I walk through our hassle trying to get Gsuite support to try to acknowledge they even consider this an issue, let alone do something about it: https://twitter.com/JustinMcCammon/status/141761476919279206...
tldr; remove all bit.ly links from your emails
Google has massively messed up spam filters and we got confirmation that they are aware of the issue from Gsuite support (although it seemed like Google did not consider it a problem and was just the absolute worst to try to work with via support. Absolutely terrible at every interaction except one rep who had to fight the system to help us investigate).
We use Gsuite at work and ran into issues where in the middle of an email thread, with contacts we'd exchanged dozens of messages over many weeks and even months, suddenly the emails were being sent to our spam folder or worse, rejected entirely (which ends up being a silent failure unless you are really on top of your email logs or you have clients that pick up the phone and say "why haven't you responded to my email?" we had the latter).
We reach out and spent weeks going back and forth daily with Google "support". I'd spend hours on the phone with them going through steps to recreate it and trying to find workarounds. Aside from one good rep who acknowledged many other people were writing in about it at the same time we were it really seemed like Google could not care less.
At one point I got so desperate I searched on twitter to find other people complaining about things. I found a person who was willing to help me - she was on the other side of things - someone NOT using gmail trying to send emails to gmail users and getting the rejection bouncebacks all of a sudden. She helped me figure out some of the root causes. Turns out Google decided that all bit.ly links were bad and if one appeared in your email it was either rejected or sent to spam (we couldn't figure out why one or the other). With her help we figured out clear steps to reproduce the issue and I did so on emails I controlled to send all the email headers and such to Google thinking they would realize the obvious issue.
Turns out we had bit.ly links in our own company email signature and so what was happening is when a client would reply to our email and it would include our own signature in it then google would flag that email as bad.
In addition there were some cases where links to google docs or youtube (the irony!) were also getting flagged.
The only thing we did that worked was to set up custom exception rules in gsuite to always allow emails through that contained bit.ly links or gdoc links as well as turn off ALL spam filtering. Naturally we all got lots more spam but we also could get regular emails again, which was much appreciated.
I had a phone convo with someone at bit.ly since I figured they might like to know and maybe could apply some pressure to google but after running it up the ladder there they ghosted me.
It must be all the pesky Russian disinformation that's everywhere, right? Surely google wouldn't tweak an algorithm to further anyone's narrative?