Is the "Discourse-Gate" plugin harmful for SEO?


#1

Hi,

I’m talking about this plugin:

My questions:

  1. How many of you have used this plugin (with the close button (X) disabled)
    and haven’t noticed anything going bad with SEO?

  2. Even if there is an overlay-pop-up, search engines like Google can still read the content - its just real visitors who won’t be able to read the content - Will this really affect SEO (since the crawlers are anyways able to read the content)?

  3. I disabled the close button (x) in the overlay pop-up - is this bad for SEO?

  4. Some say that - with overlay-pop-ups the bounce rate will increase. I installed this plugin 2 days ago, I checked the stats, the bounce rate didn’t really change at all. Its still at 3.4 % (which I think is a pretty good score). Since this isn’t really affecting the bounce rate of my site - can I assume Google won’t mind me installing the overlay-pop-up?


(Justin Veenema) #2

I was just chatting with @nixie about this via private message.

Some preliminary reading can be done on the topic by our good friend Rand Fishkin here:

Here’s my take on it.

It’s Google’s job to find the absolute best content for users who are searching. They factor many things into this equation, such as content, user experience, bounce rate, etc.

By deploying DiscourseGate you could be making the experience worse for the majority of people visiting the page. Sure, you might see 40 new signups… but how many people see that overlay and immediately leave?

Most people don’t like pop-ups – especially those that they can’t close. If I visited a site that forced me to sign-up/login to view the content, I would press ‘back’ in my browser as fast as I could.

This causes your bounce-rate to increase drastically. Google now sees that +80% of the visitors that go to your site leave immediately. That must mean that either the user experience is bad, or the content wasn’t a good fit for the search.

However…

I was just chatting with @DiscourseMetrics.com this morning about whether or not his custom sign-up overlay has had a negative effect on his SEO, and he’s explained that his bounce rate has not actually taken a hit.

I guess this would be something that you really need to scope out and keep a close eye on. Discourse has fantastic SEO built-in, so that could give us the upper hand, but I would really hate to see someone’s long-term SEO take a nose-dive in exchange for a handful of signups each day.

Test, Test, Test…

With all of that being said, we’re going to run some tests over the next couple of weeks and I’ll be sure to keep you guys posted. We’ve already deployed a few Unbounce Convertables

3db50e000e7c20f29ad47d23303d114e6463fc87_1_690x420

but we’ll also soon be testing a custom, baked-in overlay that @DiscourseMetrics.com helped us build, which you can preview in action here.

14 AM

I’ll report back soon.

Happy converting!


(Mittineague) #3

I’m not sure bounce rate is a good metric to use here. If the overlay was a “before you can see any content” maybe. But because it’s “after you’ve seen some content” bounce rate has already been affected at that point.

I think it would be more insightful to measure the close rate of the overlay and whether or not those visitors return to register at some point or not.


(Justin Veenema) #4

Pretty sure the example @nixie is referring to doesn’t actually have a close button. It loads immediately on-load, and you’re forced to either login or navigate away.


(Cody Campbell) #5

Hey @nixie,

@Justin_Veenema is pretty much on point with his answer, but I’ll add a few things.

Firstly, removing a close button from an overlay won’t hurt your SEO in and of itself. It’s the user experience metrics that Google will be looking at as a result of the overlay you’re showing.

The worst scenario for you is if users pogostick. That means they land on your site, immediately see the gate and then hit the back button. If that happens repeatedly, then that’s a very negative signal to Google.

Since you’ve said your bounce rate hasn’t changed, you might be in the clear. I’d watch it carefully over the next month or so as you’ll need a fair bit of traffic to run through this experience before you’ll have accurate data. Be mindful of your bounce rate on mobile too. Google may treat this version of your page differently. Like Justin said, be sure to test and monitor your rankings and organic traffic levels.

Also, you mentioned your bounce rate is 3.4%. That sounds fishy to me. I’ve seen that happen on a few sites, but that tends to be when there are analytics tracking errors. I would look deeper into why that number is so low. Typical bounce rates for articles like a community or a blog tend to be anywhere from 60-90%. For example, the average organic bounce rate for the Unbounce Community is at about 65% right now (and that’s quite healthy).


#6

@Cody_Campbell
I’m little worried hearing about the bounce rate. I was in an impression that my sites bounce rate is really good.
This is the stats for the last 2 days:
Is there any chance the bounce rate can be really 3.55 % as seen in the below pic? Or are you sure that is wrong?


(Kane York) #7

A 3.5% bounce rate and 60% new visitor rate doesn’t make any sense. Typical Discourse forums have a high ratio of returning visitors to new visitors. (But don’t forget that new users are essential to survival, etc etc.)

It’s possible that the analytics isn’t properly associating returning users cross-device after they log in.


#8

@riking
I’m wondering what could be wrong.
Will it be because I added the Analytics code in CSS directly (under head) and didn’t use the built-in textbox available in settings?


#9

Thanks to @riking and @Justin_Veenema for informing me about the Bounce Rate.

I think I fixed it. I monitored things the last few hours and it looks normal now.

To fix it, I removed the code from head section and then used the built-in feature of Discourse to add Analytics.


(Justin Veenema) #10

Good to hear, @nixie!

Let us know what happens to your bounce rate going forward.


(Cody Campbell) #11

Glad you got this solved @nixie. Your bounce rate should look a lot higher now. And that’s a good thing in this case :wink: