New email vendor? Expect your deliverability to plummet… at first


“I’m switching to another ESP. Will my deliverability improve or get worse?”. Of course you are expecting your open/clickthrough rate to consistently improve with the new ESP… but you’re still in for quite a shock, as deliverability tends to drop off in your first few mailings.

Wait, don’t run, come back! This is par for the course when switching email vendors, although not many are honest enough to warn you about it.

So why will your results with the new ESP get worse before getting better? Even if you stick to email marketing best practices and keep in good graces with Internet service providers?

Your hard bounces will spike

Hard bounces are dead or mispelled email addresses which wreck your sending reputation. The reason?

Major ISPs such as Gmail, Hotmail and Yahoo look at your mailing list hygiene for signs that you’re spamming. Lots of undeliverables usually mean your list is not opt-in or that permission has gone stale, so ISPs will rightly shout “blasphemy” and flag your emails as spam.

Any ESP worth their salt should automatically scrub hard bounces off your list the first time you mail it. If you are switching over from an email marketing service with no bounce handling (or if you used to do it yourself), this will be especially noticeable because you’ll see your list size shrink dramatically.

Even switching between bounce-supressing ESPs often yields a higher bounce rate, as their detection criteria can vary. The damage caused by those extra bounces in your first mailings with the new ESP means your deliverability will initially plummet, only to improve after the list is fully cleansed.

You’ll get more spam complaints

Switching to an ESP with a higher deliverability rate might actually work against you at first. Weird? The explanation is simple: your emails will now reach the inbox of people whose accounts were previously filtering you. Seems like great news, unless they aren’t interested in receiving your emails anymore, in which case they may report you for spam! This in turn lowers your ISP sending reputation, hurting your deliverability.

Don’t worry though, good email marketing vendors maintain feedback loops with ISPs to take care of complaints and automatically unsubscribe spam reporters. After the first few mailings, your complaint rate should drop, allowing you to get your sending reputation back.

ISPs give the evil eye to any change in your mailings

Rule nr. 1 of spam filters: when in doubt, sniff it out! As soon as Gmail, or Yahoo realise your “postman” is different, they’ll set the alarm bells ringing and give both your new ESP and your emails a thorough pat down.

Ever noticed how a dog’s attention will immediately zero in on someone new coming over? Yep, same thing here. While ISPs double-check that you’re a good sender and the email marketing provider is not attempting to phish you, they’ll be wary of your mailings until they give you the all-clear. You can make it easier for them by using your own sending subdomain instead of the ESP’s. Most email vendors allow you to set it up manually and some even provide you with a custom subdomain by default.

Once ISPs finish sniffing your emails all over, they’ll move on to gauging how people respond to them (ie., whether subscribers open, click through and reply). That’s what engagement is all about and as long as you manage to build a relationship with your audience (always go for highly relevant, targeted mailings), your open/click trough rates should gradually rise.

“Slow down, you’re sending too fast”

Gmail, Yahoo and many other ISPs have traffic rules in place governing how many emails you can send them in one sitting. This is known as throttling and if you fail to abide by the rules, ISPs will defer or block your emails altogether. So, if your old email vendor didn’t take this into account and just blasted out your mailings, it may look as if your new ESP takes longer to deliver them. That’s exactly what’s happening, but going slower is required to ensure the highest chances of making it to the inbox.

So what deliverability steps to take before switching to another ESP?

First off, deliverability is just one step in the onboarding and switching process, there is a lot more to expect and deliver when switching ESPs, but deliverability is an important part.

1) Don’t just export your entire mailing list from your old ESP. Make sure to export only those subscribers who are active and engaged. This will prevent you from importing invalid or already-unsubscribed emails into your new ESP, which would ruin your sending reputation.

2) Think it’d be cool to halve the mailing list and send out a couple of mailings using both ESPs at the same time to compare how they perform? Think again! Besides skewing the baseline (the mailing lists won’t be the same and even the way both vendors calculate your open/clickthrough rates may vary), ISPs can frown upon it. Save yourself the trouble and just run with the new ESP.

3) If you want to compare between vendors, go for the right way to do it: create a few test accounts in Gmail, Hotmail and Yahoo and add them to your mailing list. You should only use these test accounts (also known as a seed list) to monitor delivery. Never open or click through any email landing there – look but don’t touch!

Before migrating to the new ESP, use your old ESP to get a few mailings out the door to your whole list and check for delivery in the test accounts. After switching to your new ESP, do the same (make sure the mailing’s contents and subject line is not too dissimilar) and see how delivery shapes up against the old ESP.

4) Do you have any kind of integration, autoresponder or drip campaign in place at your old ESP? Take the necessary steps to reproduce it in your new email vendor (it might also be a good time to review it with your sales staff). If you’re getting a dedicated IP, the folks at your new ESP should help you warm your dedicated IP up.

5) This is a given, but it bears repeating. Make sure to always follow email marketing standards and best practices while using either vendor. Remember: there’s no magic formula to guarantee inbox delivery, but if you follow all the rules, you’ll have much better chances of making it through.

Image via flickr

Ruben Alvim

About Ruben Alvim

Rúben Alvim is Content and Development Manager at E-goi. He’s also part of the Deliverability Team, breaking down thousands campaign samples a day to gauge email delivery trends and pinpoint what actually takes to make it through the inbox.

You can skip to the end and leave a response. Pinging is currently not allowed.
  • Andrew Bonar

    That all sounds very dramatic if you are seeing massive spikes in hardbounces and complaints then you are not doing the migration right at all and maybe you need to be re-thinking the move to that ESP.

    You link to several articles. Two in particular particularly provide solid advice. One is from Loren who states “Moving forward with a company that takes deliverability seriously can keep your rates high.” Loren then links to Andy Thorpes post with solid advice about migrating platforms.

    That said… What however is your measure of deliverability? Those address if genuine hardbounces were hardbouncing before, they are not a new vector in your mailing program. You say ” The damage caused by those extra bounces in your first mailings with the new ESP”
    But there are no extra hardbounces, the hardbounces were more or the same the day before as the day after a switch of mailhost.
    Why then would sending to the same hardbounces be worse at ESP A than ESP B?

    You go on to say later in your article

    1) to move across the engaged users.. “export only those subscribers who are active and engaged” surely you don’t mean only active and engaged and dump the rest? How are you defining active? engaged?

    Gradually migrate users across sure, upload suppression lists, all the good stuff Andy T/CaptainInbox talks about in the post you link to. But as a marketer you will want to migrate everything across eventually.

    Either way if you only move the engaged, or only move over the engaged initially of course means much of the preceding information would be completely negated.
    (a) If you migrate engaged users you dont get bounces and if the new ESP is getting high bounces from that list of engaged users there is a problem at the ESP.
    (b) Likewise in respect of the spam complaints, the engaged users will have been getting your mail, so no need for an immediatte spike their either

    (2) If a marketer accepts point 1 of your advice.. Depending on how big a list you have, how you plan the migration, how often you send, it is not always best practice to stick with one esp. Further the new esp shoulod be getting much better results as you gave them teh best list. You still want to identify engaged people and keep the info flowing if its a protracted migration.

    In respect of making comparisons between vendors, as a marketer you know what response rates were on your former platform, what was converting how many uniques were opening or clicking, or how many sale/signups or whatever it is that is your mark of success in your campaigns as a marketer. So flawed inbox testing is simply not required.

    I’m not sure what to say, I hope I have not come across as unduly harsh, but there is some solid advice available and this post just serves to confuse and muddy the waters in my opinion. It is contradictory in many parts and does the marketer a disservice. I note this is a couple years old, so I am sure in that time you have changed and refined your process having learned from the experiences of past.

    All too often I see people lately talking about misinformation they see and read on blogs about deliverability, but people for one reason or another are not prepared to call it when they see it. I will try and stuck to my principles and advise when I see advise that is a concern.

    • Rúben Alvim

      This post was written with the worst-case scenario in mind, Andrew – that of someone using a really sketchy email marketing software with no bounce handling whatsoever who then suddenly switches over to a proper ESP.

      That’ll get you a spike in bounces and hurt your deliverability. The steps we suggest at the end are meant to prevent that.

      A spotless marketer or brand with high delivery rates will certainly want a carefully planned, multi-stage migration. However, a well-meaning but misguided beginner using Super-Mailer McBlaster to mail a couple-years-old list just wants to get rid of the chaff as quickly as possible when moving over.