AWSuddenly Dropping Open Rate
Hey guys,
About a month ago, we switched email service providers from Constant Contact to MailChimp. It's been a huge improvement from a designing/developing standpoint. But over the course of the month we've noticed a sharp dip in our open rate. Our average has been a consistent 45-49% for the last 12 months. August (so far) is sitting at 37-38%. It's a pretty big difference, especially considering our mailing list is only 2500.
We're trying to narrow down what exactly may be causing our sudden decrease. We haven't changed the frequency of email traffic, the type of emails sent, nor our email list, which is comprised of all our active business dealers (hardly any unsubscribes and delivery rate is 99.75%). We've also gone back through all of our data collected, just to make sure we're comparing apples to apples, and verified the numbers.
Gmail.com and Yahoo.com addresses comprise of the largest chucks of our list at 18.6% and 14.5%, respectively. We noticed that, according to MailChimp's reporting, about 62% of each group are not opening our emails. At first we thought it might be a sorting issue, as tests have shown that our emails are now being delivered to the Promotions folder in Gmail. We were preparing to send a specific Gmail campaign notifying our dealers that due to our ESP switch their emails might be found in a different location than they're used to. However, I'm not aware of any folder system in Yahoo that would push our emails out of the primary inbox and cause the same unopen rate there.
We're also tentatively ruling out design, since the main reason for the switch was to drop CC's unresponsive platform. All of our emails are responsive/mobile friendly and have been testing fine (or as fine as they can in Gmail, HA). Our mobile opens are actually up a little bit from past months.
So what could be going on with our open rate?
Has anyone else noticed drastic differences in reporting after switching mail providers? Should we expect a "warm up" period and, if so, for how long? I understand that a month might not be the longest interval to get an accurate picture, but at the same time we don't want to sit on any potential issues for too long without taking preventable action.
Could there be a difference in the way the different platforms report an open? I've seen that before, for clients moving platforms and seeing wild fluctuations on rates.
For example, one platform may also count a clicked link without a downloaded tracking image (ie. images off but still clicked), whilst another may not.
I'm assuming that both platforms are reporting an email "open" solely as clicking the email in the inbox to open the message in the reading pane/window, but I haven't done any research about that yet. Each ESP tracked click-thru rates separate from open rates, and we haven't noticed any substantial difference during the switch (maybe 1-2%).
just had a look around constant contact's docs - http://support2.constantcontact.com/articles/FAQ/1611
this doc suggests they count an open as one of a) a tracking image being downloaded, or b) a click, so that metric that would include users who click without downloading the tracking image (ie. images off but clicked).
I can't find a definitive answer on Mailchimp's site, but if they only included a) in their metrics then that would explain a difference. Yahoo and Gmail users could quite likely have images turned off, which would put them under column b), so if Mailchimp don't count them then that cause a difference.
Of course it could equally be down to IP reputation, deliverability etc. It's unlikely but possible that it could be the size of your HTML. eg. if you have a lot of code from Mailchimp vs. Constant Contact, the tracking gif could drop past the 100k HTML limit and therefore not be downloaded as often, reducing the opens counted. Responsive does add a bit of weight to HTML, though you'd need a fair amount of content to hit that limit.
Ah, I misunderstood your first post. I found MailChimp's tracking information - http://kb.mailchimp.com/article/about-open-tracking. They use an invisible graphic as well, but I'll have to read further about their troubleshooting.
We also tested MailChimp's general send IP addresses, but they all tested clean.
I just checked Elliot's theory, which I thought would probably account for it.
Unfortunately (for this scenario), MailChimp also counts a click as an inferred open ... So, the open rates are definitely an apples-to-apples comparison technically-speaking.
stumped
Welp.
I was secretly hoping that something like Elliot's suggestion was happening. But since that's not the case, I guess we'll be monitoring our campaigns over the next few months and see where the open rate goes. It could just be a bad month (although that's not something we've ever seen).
Hey Abby - would you mind sending me a bit of information? A DM on Twitter would be fine. I can ask our delivery folks to take a look at what's going on.
Well, I would but it won't let me send a DM to a non-follower. ;) Is there another way I can send you a message?
Sounds like a deliverability issue? If you're not getting into the inbox with your new IP which I assumed is a shared IP, you're open rates will drop. Do you have a warm up plan for your new IP/domain? Perhaps put a banner in to ask people to add you to their safe senders list?
Perhaps run a delivery test? Check out the sender score for your IP?
There was a problem with delivery to Yahoo in the last month by the way.
As far as I'm aware, we did not experience any change in delivery rates. We didn't get any abnormal bounces or unsubscribes either. Our list is a set number, so the concern is that these dealers were opening their mail one month (and consistently every month prior) and then suddenly weren't in August.
We already checked the IPs; those were clean.
I don't have the final numbers for August. Considering the original post is a few weeks old, I don't have the final data for the month.
Are you emails actually getting to their inbox though? Are they landing in junk or going completely missing.
It would be the first thing I looked at if nothing else has changed e.g. templates, lists. Its sounds like delivery issue to test. Might be worth looking into Return Path for some advice if it hasn't improved.
The MailChimps guys (who are amazing btw) looked into it, and didn't find anything significant. They did add us to a seedlist to monitor if our messages are going into a spam folder, but assured me that these types of fluctuations in open rates are normal. I think our opens got bumped up a bit by the end of the month too, so it's not as drastic as we initially thought.
Knowing that, we'll keep an eye on our account for the next few months. If the opens plummet or go wonky again, we'll figure out something else.
Hi Abby,
I came across your thread and I'm wondering if you ever found a solution to this issue. My organization is facing almost identical drops since switching from Constant Contact to MailChimp. I really like the MailChimp platform but can't figure out why our open rates are dropping so drastically. Any information would be much appreciated! Thanks.