It's time for another Spam Resource Tuesday Tip. The goal? To share short and actionable deliverability tips and tricks. More often simple hacks than deep deliverability dives.
This week, it is just something that I've observed, and that other people have observed, and asked me about. When comparing notes, we've found that we're all seeing clients run into this issue. And the issue is: It's taking longer to resolve Gmail spam folder issues compared to years past. And the point if me sharing this here on Spam Resource is to confirm for you that it's not just you -- other folks are seeing it, too.
Here's what the Gmail remediation process looks like.
The guidance can vary, but when dealing with spam folder placement at Gmail, the remediation process usually looks something like this:
Make sure all the technical bits are right. SPF and DMARC in place, and especially: DKIM signing mail as yourself, your own domain, not your platform's domain.
Set up Google Postmaster Tools (GPT), if not already configured. Necessary to watch for domain reputation score changes as you work through this process.
Adjust segmentation to temporarily (or permanently) suppress subscribers who haven't opened or clicked on any email message you've sent in the past X months. X varies, it'll be based on list size and industry. Need a starting point? Try six months.
Send your usual (often weekly or twice weekly), but now only to this "tightened up" list of more highly engaged subscribers. This boosts your engagement, as measured by Gmail.
After around three or four weeks of sending, mail will usually start going back to the inbox. If it doesn't, the data is usually suspect. This cleanup process only works with opt-in data and generally active subscribers (meaning re-enabling 10 year old dead addresses is not compatible with this process).
What's changed nowadays is that step five now seems to take longer than in years past. Microsoft Outlook/Hotmail's "reputation restore" often took much longer than this; sometimes twice as long, sometimes longer. But now that seems to be the case with Gmail, too. What was previously a 3-4 week process of sending to restore your domain reputation through boosted engagement is taking longer for some folks -- not all -- but enough that it is clear that Gmail's filters are handling this differently than in years past.
How long? I don't have millions of example to draw from; the limited anecdata I've got so far suggests that what was 3-4 weeks before, might take 5-7 weeks now. But it will vary. With better data comes better chances of quick success, though I am guessing on that.
What to do about it? That's a tricky one. In a perfect world, you'll want to avoid having to deal with this at all, because that pound of cure is now even worse than in years past; that ounce of prevention is looking ever better. That involves implementing a subscriber lifecycle policy to automatically suspend mail to subscribers after a period of time. I typically recommend implementing a subscriber flow that drops an unengaged subscriber into a series of 1-3 winback campaigns, trying to get one last sign of life from them, before dropping them into the "dead" bucket and no longer mailing them. The more you can do to get unengaged subscribers out of the way, the better off you'll be.
And what of proxy opens and Apple MPP? We know that opens are less accurate than in years past. Marketers still track opens (and that might be a metric that stops being tracked only when you pry that tracking pixel out of their cold, dead hands), but while I think that open tracking is not generally safe for decision tree routing of individuals in an email flow, I do find it to be "good enough" to continue to be used for engagement-based suppression of inactive subscribers. As open tracking counts are over-inflated, that translates to you suppressing fewer subscribers than if you used the same process prior to the rise of Apple MPP, but the net effect is still generally the same: successfully boosting engagement and improving your sending reputation.
Maybe that longer "reputation repair" step is due to lesser suppression due to less accurate open tracking? It's hard to say, but my guess is that it's not related, as this seems new to 2024, and we've had Apple MPP for a few years now.
TL;DR? Gmail remediation of a low-engagement driven spam folder issue can take longer than in years past, making it more important than ever to proactively avoid falling into that hole, if at all possible.
It's time for another Spam Resource Tuesday Tip. The goal? To share short and actionable deliverability tips and tricks. More often simple hacks than deep deliverability dives.
The guidance can vary, but when dealing with spam folder placement at Gmail, the remediation process usually looks something like this:
- Make sure all the technical bits are right. SPF and DMARC in place, and especially: DKIM signing mail as yourself, your own domain, not your platform's domain.
- Set up Google Postmaster Tools (GPT), if not already configured. Necessary to watch for domain reputation score changes as you work through this process.
- Adjust segmentation to temporarily (or permanently) suppress subscribers who haven't opened or clicked on any email message you've sent in the past X months. X varies, it'll be based on list size and industry. Need a starting point? Try six months.
- Send your usual (often weekly or twice weekly), but now only to this "tightened up" list of more highly engaged subscribers. This boosts your engagement, as measured by Gmail.
- After around three or four weeks of sending, mail will usually start going back to the inbox. If it doesn't, the data is usually suspect. This cleanup process only works with opt-in data and generally active subscribers (meaning re-enabling 10 year old dead addresses is not compatible with this process).
What's changed nowadays is that step five now seems to take longer than in years past. Microsoft Outlook/Hotmail's "reputation restore" often took much longer than this; sometimes twice as long, sometimes longer. But now that seems to be the case with Gmail, too. What was previously a 3-4 week process of sending to restore your domain reputation through boosted engagement is taking longer for some folks -- not all -- but enough that it is clear that Gmail's filters are handling this differently than in years past.What to do about it? That's a tricky one. In a perfect world, you'll want to avoid having to deal with this at all, because that pound of cure is now even worse than in years past; that ounce of prevention is looking ever better. That involves implementing a subscriber lifecycle policy to automatically suspend mail to subscribers after a period of time. I typically recommend implementing a subscriber flow that drops an unengaged subscriber into a series of 1-3 winback campaigns, trying to get one last sign of life from them, before dropping them into the "dead" bucket and no longer mailing them. The more you can do to get unengaged subscribers out of the way, the better off you'll be.
And what of proxy opens and Apple MPP? We know that opens are less accurate than in years past. Marketers still track opens (and that might be a metric that stops being tracked only when you pry that tracking pixel out of their cold, dead hands), but while I think that open tracking is not generally safe for decision tree routing of individuals in an email flow, I do find it to be "good enough" to continue to be used for engagement-based suppression of inactive subscribers. As open tracking counts are over-inflated, that translates to you suppressing fewer subscribers than if you used the same process prior to the rise of Apple MPP, but the net effect is still generally the same: successfully boosting engagement and improving your sending reputation.
Maybe that longer "reputation repair" step is due to lesser suppression due to less accurate open tracking? It's hard to say, but my guess is that it's not related, as this seems new to 2024, and we've had Apple MPP for a few years now.
Comments
Post a Comment
Comments policy: Al is always right. Kidding, mostly. Be polite, please and thank you.