You click a link to go to an interesting-sounding page. Your mouse is still wherever that link was on the previous page, but now, blammo!, it randomly lands on some menu, which won't wait to find out whether you wanted to see it or not. It just throws up all over the page and leaves you looking for a way to get rid of it.
You scroll the new page a bit, click another link, and blammo, another menu jumps up and covers the new page. You see a link near the top of the page you want to click and as your mouse heads up to it, blammo, you pass over another menu that jumps out and covers what you were trying to click.
Usually, a popup signup ad will stay closed after you close it the first time, but these menus won't EVER stop jumping up in your face. Tooltips or items that highlight themselves in place on hover are no problem, because they don't get in the way, but menus are meant to cover other things. That's why all native OSes have menus wait until users explicitly ask for them.
And touchscreen devices essentially have no hover, so if your menus had to have these obnoxious dynamics for users to discover that they were menus, it would mean that your design was nearly unusable on devices with no mouse. Make your menus obvious to users without a mouse and you no longer have to make them obnoxious to users with a mouse.
Tell your designer friends: when you have caught someone's attention and made them want to see something on your site, don't then suddenly block their view with something they DIDN'T want to see, something they then have to escape from, whether "Sign up now!" or "Click here to tell your friends about us!" or "What do you think of us now?" or "Here's a cascading menu of things you weren't asking for instead of what you wanted!"
The (usually) unasked questions are whether those people represent actual customers (or just the terminally clueless) and whether getting a few more email addresses for your newsletter is worth irritating the crap out of the other 90% of site visitors.
Sadly, they work! One claimed a 78% larger sign-up rate and the other was something like a 65% larger sign-up rate.
I did not discuss churn rates with them - but overall traffic to the sites weren't largely impacted I was told.
Note that the popups on each of these sites was a one-time deal on a user's first visit. After closing it you never saw it again (unless you clear your cookies or browse incognito, etc.)
I have noticed that some sites make the popover happen when you mouse off of the window to the back button, which kind of makes sense because you're leaving anyway.
Unfortunately you're in the minority. Every A/B test I've seen shows popovers performing much better than end-of-article placements on all dimensions.
Important, but unanswered questions (because they are hard, so let's ignore them!):
(1) How many of these users just mark all of the newsletters as spam?
(2) How many users don't realize that they can dismiss the dialog, and enter their email because they think that they are forced to? How many of these are really "engaged" users vs. just users that are willing to jump through a couple of hoops before they disappear in the ether?
(3) If those extra sign-ups are due to people thinking that they have to enter their email, how is this any different (ethically or practically) than actually forcing them to enter their email? Are you only making this distinction in the smallest way possible so that you don't run afowl of anti-spam laws?
(4) Did the A/B test include a "No thanks, I want to read the article instead" instead of just a "No thanks" button? How much does that affect the number of sign-ups?
(5) How many of those users will actually open their newsletters vs. getting lost in the clutter of other newsletters that they automatically sign-up for?
My newsletter open rates are around ~45% and click rates are ~15% which both seem to be pretty good and I attribute this to the fact that you have to really want to subscribe in order to do so. Given that I pay based on the number of subscribers I have, and that the purpose of my emails is to point people to the website, I actually periodically unsubscribe people that never click/open them.
To answer your questions from my experience (dealing with a non-tech audience):
> (1) How many of these users just mark all of the newsletters as spam?
The deliverability and negative feedback for users who come in through the popover is not statistically different from users who sign up in other positions (bottom of the page, for example).
> (2) How many users don't realize that they can dismiss the dialog, and enter their email because they think that they are forced to?
We did some basic user testing and the vast majority of people had no difficulty with finding the close dialog. If this were the reason, I think it would be born out in (5).
> (3) Are you only making this distinction in the smallest way possible so that you don't run afowl of anti-spam laws?
IANAL, but I don't think there's actually any law preventing an email wall. We even tested one at some point but it did have a meaningful effect on bounce rate, and it did tend to get more undeliverable emails than the dismissable version.
> (4) Did the A/B test include a "No thanks, I want to read the article instead" instead of just a "No thanks" button? How much does that affect the number of sign-ups?
Why would a business do this? What business goal would including the additional copy fulfill?
> (5) How many of those users will actually open their newsletters vs. getting lost in the clutter of other newsletters that they automatically sign-up for?
In my experience, ~20% open the emails and then 40-50% click a link within the email. These rates were not statistically different for people who subscribed through popovers vs. less intrusive forms.
That being said, I still personally dislike popovers and never subscribe through them.
That is, by far, the most ridiculous click through rate I have ever seen for a newsletter.
We have a pretty great fanbase. :)
"Hey, you were probably about to give us some actual money, but would you mind stopping that and doing something else instead?"
Jordan Coeyman's OptKit is a light but intelligent popover tool that actually works, and I signed up for his newsletter based on how well the tool worked on his blog.
Ironically, it's the proliferation of unnecessary third-party tools (as described in [1]) which is a major source of the web slowness described in the original article.
[1] linked from the article: http://pxlnv.com/linklog/tools-dont-solve-the-webs-problems/
Click away from modal content; Try ESC button; Close browser tab
Yes. This is a very successful way to increase those numbers
> Does this actually work?
A more important question! The issue is that the numbers that the website owner is trying to get up are really low to begin with - usually <1% or so.
So you have to think like a Bayesian here. What are the chances of someone who isn't computer literate enough to know there's a way to dismiss the dialog and believes that giving their email address is the only way to continue?
Even if 99.8% of users are smart enough to x out of that dialog, then bam, you've increased your newsletter membership by 20%!!! (simply by confusing 0.2% of your visitors) Congratulations! Look at how effective you are! You're now giving a bunch of novice computer users email they don't want and shit they won't read!
Either these things aren't looked at holistically or they are and people don't care about user experience.
Think about it, you have to do this using CSS because all the browsers block doing the same functionality with another window.
Once you start asking questions like "how many exits happened at dialog versus how much convergence?" you will, as I have, see ~60x more exits from service then engagement with dialog. You lose 6,000% more than you gain.
And yes, the engagement is so low on most analysis that you can't dismiss marginal user error in many of the cases as constituting a large percentage of engagement.
We can do better
A pop-up or -under window and a modal dialog are also different in that the modal is confined to the window or tab it exists in and when you leave the page you're on it goes away.
If the website is honestly just a bunch of scraped content which solely exists to solicit donations then go ahead. But if there is another purpose, any other one, than the popup dialog detracts from it in an unnecessarily injurious way.
In the meantime, the rest of us will learn to write greasemonkey rules for those popovers and the next generation of adblockers will evolve.
Or are the statistical signals not strong enough to distinguish them from user-requested pop-overs like "I want to login"?
For a less "nuclear" option, use Adblock Plus. And this suggestion might help the defaults a bit: https://adblockplus.org/forum/viewtopic.php?t=27776
And if that doesn't work, get Greasemonkey/Tampermonkey and write your own by playing around with the debug console. (Chrome: Shift-Alt-C or Shift-Cmd-C - firefox: similar...)
Yes. I've personally seen seen multiple A/B tests on different sites showing newsletter popups to be highly effective on both a rate and absolute basis. Their increase in conversion rate is more than enough to offset the (tiny) number of users who leave because of them. (We also ran this A/B test over multiple weeks to see if it had an effect on return frequency. It didn't.)
That being said, good publishers should cookie that you closed the popup and not bother you about it again. I imagine the numbers on this would also be different for a technical audience.
As many have pointed out, most are unlikely to subscribe to the newsletter/install the app/whatever before they've even seen the site.
It seems like the least annoying way to do this would be to detect multiple visits (preferably over multiple days) and then ask for the signup. If the person says no, don't show it to them again for a very long time, if ever.
"We also ran this A/B test over multiple weeks to see if it had an effect on return frequency. It didn't.)"
Does your A/B test detect how annoyed the readers are?
We actually tested that as well (though I think we did it after 6 pages). The conversion rate was of course higher than showing it immediately, but the users didn't open or click any more frequently. (Hence, showing "immediately" on first page load is better business-wise.)
> Does your A/B test detect how annoyed the readers are?
Haha, that is indeed why I pushed against this in the first place (if I hadn't, we wouldn't even have tested it so thoroughly). But the fact that such "annoyance" didn't show up in any metrics ultimately won me over.
One thing to keep in mind is that I've only tested with non-technical users, who I think are in general much less annoyed by this sort of thing than the typical HN person.
This is actually very very useful, but every single implementation is bad. Ideally your browser would have all this figured and allow you to one click sign up for the newsletter. The popup should be noticeable, but away from the content (above the reading area).