There’s a long history of conflict between double opt-in and single opt-in list confirmation. Ultimately, email senders have to choose between list size and list quality. As an ESP, we tend to care a lot about the quality of your lists. After all, our IP reputation is at stake.
A handful of bad lists can ruin the party for our million good users. That’s why you’ll hear us talk so much about double opt-in list confirmation. If you don’t know what I’m talking about, check out Ben’s blog post from 2005. It’s kind of required reading if you want to understand the stats below.
When talking about double opt-in, we tend to hear customers say, “Why should I make it harder for people to subscribe to my list?” I get it. After all, double opt-in does require people who want to get emails from you to… open an email from you. It’s an extra step, but I think it’s one you can afford. Cutting corners in list collection tends to generate bounces, blacklisting, spam trap hits, and other nasty stuff that can tank your emails.
The question is, will using double opt-in really improve the stats that matter: opens, clicks, bounces, and unsubscribes?
We took a random sample in our database of 30,000 users who’ve sent at least 10 campaigns. We wanted to make sure they’d have enough stats to make the analysis interesting. These users had list sizes from 500 to 1.5 million. Some of them send every day, and some of them send a couple of times a month.
Higher Stats
After pulling this broad sample, it was fairly easy to aggregate their campaign stats and let the numbers do the talking. I think you’ll like what they have to say. As you can see, double opt-in lists have better open rates.
Unique opens means we don't count multiple opens per subscriber.
It’s important to keep in mind that these are unique opens. That means we only count the first open per subscriber. It turns out that while double opt-in has a 72.2% increase in unique opens, they have a 75.6% increase in total opens. So not only do double opt-in lists get more opens, they also get more multiple opens per subscriber. Cool, right?
Okay, so people are opening your emails. They’re also clicking your emails. The chart below shows that double opt-in lists have a 114% increase in clicks as compared to single opt-in lists. That’s right, it’s more than double.
Lower Stats
There are a few stats you desperately want to lower. The first would be bounce rates. Bounce rates can get you kicked out of MailChimp, which would be terrible. We’re much more strict on hard bounces than soft bounces, and using a double opt-in list can lower your bounce rate by 48.3%. The soft bounces are 25% lower, which isn’t bad either.
Finally, you’ll be pretty interested in the unsubscribe rate. The higher your unsubscribe rate, the more aggressive you have to be about list growth. For the desperate email marketer, that can mean making risky choices with your list. Double opt-in lists have a 7% lower unsubscribe rate as compared to single opt-in lists. I wouldn’t call that a miracle number, but it does give you some breathing room. It also proves that a lot of those single opt-in subscribers don’t want your email. Keep in mind, unsubscribing is the nice way a disinterested subscriber responds. They’re just as likely to hit the spam button and forget about you. If at all possible, let’s avoid that!
By the way, we’re not the only ones who care about your list quality. Major ISP’s are watching what you send, how many opens you get, how many clicks you get, and definitely how many spam complaints you get. They can choose to throttle your campaign or simply send it to the spam folder. It definitely pays to be vigilant. As Freddie Bueller once said, “Email moves pretty fast. If you don’t stop and look around once in a while, your subscribers could miss it.”
Thanks Matt. I’m having this ongoing conversation with a client of mine right now. They are very much on the up and up but we are philosophically at different ends on this one. I’ll be sure to share this post. For what it’s worth, I use MailChimp for my list, so yes, I side on double opt in.
How does this relate to the size of the list that you can get/acquire as a company, though? If I can get 5,000 double opt-in subscribers where 1500 of them open (30%), that might not be as useful to me as getting 10,000 single opt-in subscribers where I get 1700 opens (17%). Are there any stats for people who don’t sign up due to the double opt-in hurdles they have to jump through? (Though I agree, it takes minor effort.)
Hey Ben, this is something I’m desperately working on because I knew someone would come along and point out the white elephant in my post. Congrats on being the one! :)
The thing is, a list growth analysis is going to be really interesting. I’m trying to pull enough data with sufficient detail for us to see the nuance between double and single opt-in list growth. In other words, it would pad my nerd ego if we could make some cool graphs.
Hah, sorry to be “that guy.” :P But it interests me. It makes sense that in the long term double opt-ins just matter that much more, but often time businesses overlook long term impact and want the most *now*! lol
Mathew, any follow up on this? Ben’s question is a good one for sure.
Following on from Jeff, any follow up on Ben’s questions, or more specifically, can you simply add the actual numbers alongside the percentages above? Thanks.
(PS, I note that I didn’t have to double opt in when I made this comment…)
Note that it *was* moderated by a human before publishing though. ;-)
:)
Hey, we definitely have a followup to this on the radar. We decided to juice up our reporting database, which of course means we also created an endless number of new projects to work on. Cool new toy, must play with it :)
Anyway, we are working on doing the list attrition analysis for a couple of different projects, and the double opt-in followup is written in large bold letters on that list. You know, right next to a note saying we need to make an AI that monitors our other AIs and just below an integral equation involving pizza.
Hi Matthew,
Your Blog posting is most relevant to our recent discussions on ways of growing our list. Currently we have 4700 on our list with nearly all being Double Opt-Ins. Our % of Opens is 60% week by week and the Clicks are 18% — both way better than average for our sector (retail).
Over the past year and since we started with MailChimp, we’ve also have about 1500 customers who failed to complete the Double Opt-In procedure when invited to do so via a personalised email, with a follow up request, usually one month later.
I’m curious to discover why we lose so many. I’m starting a small survey by email with a modest incentive for each reply. So far I’ve had 4 respondents from 21 customers surveyed with only 1 respondent providing anything in the least bit useful.
This is starting to look as though the ones who didn’t bother to follow the Double Opt-In request were not really worth having as subscribers anyway — thereby adding further evidence for the value of a Double Opt-In subscription process in terms of list quality.
The people who didn’t follow up with the double opt-in either didn’t get the confirmation email (that’s the part most people are concerned about), or weren’t really all that engaged or smart (also important) or were spam bots just submitting fake email addresses (which can result in some of the worst deliverability problems). I’ve seen customers’ signup forms get hit by spambots, and suddenly thousands of email addresses — some of them spamtraps — are submitted to their list. In the last scenario, using double opt-in can help you prevent blacklisting by admins.
Hi Ben & Matthew,
In my surveyed sample of the 21 folk who had not opted in, all were genuine MailOrder Customers who had actually purchased goods from us. Thus none could have been spam bots.
I don’t believe we’re dealing with any spam bots in our subscription process either as we record the full postal address in our Customer database of anyone who wants to join our list. That enables us to snail mail them our printed catalog twice each year. Any addresses that are undeliverable result in an immediate purge and folk that don’t actually purchase in the previous two years are asked confirm that they still want to receive it. Their mailing status is set to “Conditional” when asked to confirm and to “NonResponder” when they fail to do that.
It occurs to me that waiting 30 days to ask for the opt-in is probably long enough for the subscriber to totally forget why he/she subscribed in the first place. You might have better results if you follow up quick and with a nice letter in the first day or two of the signup (if not the first minute or two), so you don’t have to fight the “interested in you” battle a second time.
Interesting stuff!
I’d also like to suggest you take a look at the delivery statistics associated w/the double opt-in signup process, so you compare apples to apples. E.g. I understand that the *future* bounce rate of double opt-in lists is low, but isn’t that just because the list was pre-cleaned by the delivery of the initial “click here to sign up” emails?
From a marketer’s point of view, all they want to do is capture the highest percentage of deliverable email addresses from their website visitors. And I’m curious if you see any evidence people type more carefully when they know they will later have to click on a confirmation email.
Very interesting post. I have seen a post or two comparing subscriber growth using single opt-in and it seemed great, these stats show the other side and now it doesn’t look that good anymore…
Very interested in hearing about the list size difference, too.
Call me a #datageek, this stuff excites me. Double Opt-in vs. Single Opt-in Stats – must read for #emailmarketers. http://t.co/lV3lTG2r
Great insights from @MailChimp. Love the service! "Double Opt-in vs. Single Opt-in Stats" http://t.co/JPbXIdaF
Double Opt-in vs. Single Opt-in Stats: There’s a long history of conflict between double opt-in and single opt-i… http://t.co/Kbnlt1Se
MailChimp Email Marketing Blog – Double Opt-in vs. Single Opt-in Stats http://t.co/xWfe415p
How to nearly double your email newsletter open rate: double opt-in your subscribers http://t.co/3jbY9TPc via @mailchimp
RT @jimcaruso: Double Opt-in vs. Single Opt-in Stats http://t.co/fgjwNYqQ
Dus daarom: dubbele opt-in – "Double Opt-in vs. Single Opt-in Stats" http://t.co/ME2Ihf4B #email
Double Opt-in vs. Single Opt-in Stats | MailChimp Email Marketing Blog http://ow.ly/6Dj6W
Double Opt-in vs. Single Opt-in #Email Stats http://t.co/wmDsbPC6
Double Opt-in email campaigns have significantly better stats than Single Opt-in | MailChimp Email Marketing Blog http://t.co/6nQQyQWj
Email Double Opt-in vs. Single Opt-in Stats http://t.co/jOlkXMlG Quality over list size
Double Opt-in vs. Single Opt-in Stats for email marketing lists http://t.co/5Yp7rAwW
Double Opt-in vs. Single Opt-in Email Marketing Stats http://t.co/R4iaBrzN
Worth a look: Double Opt-in vs. Single Opt-in Stats from (author unknown) http://t.co/5c6sHF59
#EmailMarketing Stats: Why you should use Double Opt-in vs. Single Opt-in http://t.co/DI0BbSqE
Matthew. This is off topic but could you do a post soon on setting up Outlook so that is doesn’t send _all_ messages from mail chip to the junk folder? I have read your support article on the subject but it doesn’t help. Nothing, but nothing, stops messages from the several Mail Chimp-powered newsletters that I subscribe to from getting sent going to junk because of the ‘random sender domain’ problem.
Does it mean that the only way for me to get mcsv emails into my inbox reliably is to become a MC customer so that I can ask your support people for the full mscv domain list file and import it into my safe sender list?
I can understand the need for this when it comes to flat out email marketing. The purpose of my email account is strictly for autorepsonder purposes for my membership product. My members are paying for monthly membership and agreeing to terms and conditions upon joining. Why do I need to bother them with additional unnecessary emails? Please explain how this applies for those of us in the autoresponder membership world. Thanks!
A confirmation email can prevent someone from inputting an incorrect email, and your product getting sent to the wrong person. http://www.honet.com/nadine/
Are there stats on Double Opt in Abandonment? Like how many people start the process that don’t finish it versus those who complete it?
I think this is the primary missing stat from above that would tie the information all together.
thanks!
Thanks for the stats. I’ve been waffling between which is my better option and now I know. As a brand new business, I want to do as many things right the first time as possible.
Thanks Matt. As Ben pointed out, open and click rates are really not the right metrics here. Ultimately you would have avg. lifetime value metrics for members of both groups. That, plus avg. double opt-in rates gives you a better sense of which is a better strategy for a marketer.
Using LeadSpend’s real-time email validation service, our clients are identifying and removing invalid email addresses in the form, before they ever send an email. The service identifies bounces, disposable accounts, some spam traps, and other invalid email types.
In order to be a fair comparison, the experiment would have to split a sign up process into two equal parts, one would require double opt-in and one would not. Then we can test the ultimate performance of both halves of the test – conversions, transactions, and revenue.
Since double opt-ins have a much lower subscribe rate, you have to balance a better response from the few subscribers on that list vs. the lower quality response but higher volume single opt-in list.
I get what you’re saying about balancing the subscribe rate with engagement stats, but don’t forget that sending a high volume of email to a disengaged list can hurt your overall delivery.
I’ve seen subscribers who open the email every day suddenly notice the email they love is getting delivered to the spam folder. This happens when your overall list is so disengaged that the receiving ISP decides it’s safer to treat your entire campaign as spam, even if there are a few very loyal subscribers.
It can take weeks to rehabilitate a list that’s gone sour. I guess I’m not adding a lot of data here, but if I can throw another weight on the scales, I’m happy :)
thanks for this – I will try and change it all before my next newsletter – I am also trying to weed out the single star people I have who never bother to open the emails. I’m sure the increased quality of the list will help us all. As a small business, I’m only really interested in customers who want our emails. A large list of disinterested or irritated customers is of no use whatsoever.
Hi how do we create a double opt in list I have a sign up to win a $250 gift voucher and have received a lot of gmail and hotmail emails which either don’t open my email or unsubscribe this option sounds like a great way of stopping people joining who have no real interest bar the $250 gift voucher
You can use a double opt-in signup form, and dangle the $250 incentive behind the confirmation: http://blog.mailchimp.com/personalize-your-welcome-emails-with-custom-freebies/
People will still double opt-in just to get the voucher, and then they will indeed have a higher tendency to report you as spam if/when they realize they didn’t win. That’s just how freeloaders are. But at least with double opt-in you’ll have proof that they signed up if an abuse desk questions you.
Hey Matthew, unfortunately, clicks, opens, and bounce rates are important, but you forgot the stat that really matter: sales.
How do the two compare when it comes to actually buying your product?
My guess is the more opportunity I have to deliver my email, the better my chances are of getting a sale.
What does your test say about that?
Justin
Matthew – Fascinated by these numbers. I realize this post was way back in September of last year, but curious if you’ve done any follow up. What I really want to know is how list growth is impacted by a single vs. double opt-in approach. To me, that’s the key. Return Path just came out with some data showing that only 76.5% of all commercial email gets delivered to the inbox. If this is the case, one could argue a double opt-in approach may be losing 23.5% of it’s potential subscribers based on confirmation emails not reaching the inbox.
I’m not certain my logic is solid on that (ha!), but it’s something to ponder.
Any follow up?
I’m more interested in conversion and click-thru rates than open rates and experience with client list comparisons have shown stronger results with double-opt ins. Those folks who want to hear from us are more likely to buy from us or refer us. We even go the extra mile and add an additional unsubscribe graphic at top right inviting them to say bye-bye if they don’t find our content valuable. Had 1 (former) client who insisted on single opt ins despite repeated warnings and they got blacklisted over time because the list was large and each typo added up, people had short memories and marked as spam, and sometimes, because it was political in nature, people would be signed up by others… all this resulted in a large number of spam dings. I’m firmly on the side of using double opting in and maintaining a strong, clean list built on transparency, which builds trust and goodwill. Useful post; thanks!
I agree with Ben. I would like to see the numbers… Because a small favourable bump in the quality of subscribers at the cost of a large unfavourable drop in total subscribers may be concerning.
Thank you for this statistics.
Double opt in has a better open rate for sure.
But how did you count lost opt-in, people who didnt receive the confirmation message, or got it in their trash box?
Hi John, I had a great chat with Matthew about this study. It was really focused on quality comparison of both types of lists. While those other figures might be interesting to look at, albeit hard to track, they wouldn’t have impacted his findings. If email trends and stats are your thing, you might be interested in our Email Genome Project. You can read a blog post about it here: http://eepurl.com/pxhA9 or you can even sign up to receive updates here: http://eepurl.com/ixnB2
One statistic i really want to see is the dropoff rate between somebody who has entered their name and email but never pressed the confirm button. If the number as a percentage is higher than the advantages that your argument doesn’t work. My goal is to get a large list of people, and over periods of time- shrink it to people who really want the information
Sorry if I am missing something but shouldn’t the number of opens be counted not percentage? Obviously single opt ins will have a far higher number of opt ins, so the percentage of opens will drop, BUT the total number of opens may actually be higher than the double opt in. Its about numbers not percent. Your 28% of double opt in could be about 28 out of 100 people, whereas the 16% for single opt in could be about 160 people of 1000. I would prefer 160 people opening it than 28. The numbers may never be this different, but its an example of why I think numbers are important over percent.
While I completely understand the double opt-in & its efficacy – it doesn’t seem to work in all cases. Unless of course the only forms you are creating are in fact mailing lists. Which the bulk of mine are. However, and I didn’t see this addressed – what if you have someone requesting a Quote? Obviously making them double opt-in is counter productive & works against getting a response by creating an extra barrier.
In those cases, and those alone, since you are presumably going to contact them with follow up info – directly, it seems to be a great flaw in the otherwise great product you’ve designed. Obviously I would love to have you administer not only my mailing lists, but also use you for the request a quote since presumably I’d contact them later (whether they bought or not) especially to offer them a second chance.
Obviously I can create one on my own & have it go directly to a sales email, however then they would have to be subscribed manually to the MailChimp site…and seems that this sort of thing should be baked in.
It’s hard to believe it but it’s clear double-opt-ins are the best ones to focus on, still folks want to short cut via single opt-in even if the results are lesser. Those who double opt-in get the most from their users – thanks for the sharing the great info.
So question. Is it against the data privacy laws in anyway to only use single opt – in… I heard there were some new EU directives?
This is great info. I’ll be sure to keep this in mind for the next campaign.
This study makes a nice looking stat in favour of double-optin, but it doesn’t help me (one individual marketer) make a decision.
For simplicity let’s say there are “good” and “bad” marketers. Bad marketers are the spammy type who are less discriminate in their list-building and more spammy afterwards. They are likely to be using single optin more often than the good marketers. Their results (opens and clicks) will naturally be worse regardless of the optin method (because of spammy followup emails etc), but in doing so the single optin gets a bad rap.
Now if I’m an individual “good” marketer, which do I choose? Really I want to know the results from “good marketers only” to help me make a decision. I think we’d need a study of two campaigns by the same marketer to shed some light on this. Since the study here MAY be comparing two different moulds of marketer.
Hi Dave, Yes, that’d certainly be a different study altogether, but for the purposes of this study I should point out that it wasn’t so much about saying double-optin is better than single-optin. In this case is was simply more of an observation that (in theory) using the double-optin process should simply increase the likelihood that people on your list really do want your email.