"Google May 2020 Core Update" affecting your xF Forum?

askandy said:
So you mean the Siropu Ads Manager is not a good (well coded) add-on?
Click to expand...
Well, I'm not saying that. )
I'm saying that with Siropu Ads Manager a picture loads longer, than 3 times bigger one without it. And again, if it would be just one picture, may be the difference would be not noticeable. But you loaded AdManager with bunch of ads that has be processed at the same time. It's like with a browser, when you open too much tabs. It doesnt necessary means that the browser is not well coded, it means that you overloaded it with tasks, that it hardly can handle.
 
askandy said:
Yeah, I'm working on getting rid of Viglink.

In the process of working with a developer to create a Viglink alternative, that hopefully will be less bloated then Viglink, and give me 75%+ of the core functionality.

Should have an MVP running in a week or two.
Click to expand...

This could be done with Ad Manager 2.

Go to VigLink stats, find the domains that generate you the most revenue.

Add a URL replacement in Ad Manager 2 to rewrite those domains to:
http://redirect.viglink.com?key=KEY&u=URL

It will do the redirect for those domains, and you can get rid of the js.

arn
 
arn said:
Add a URL replacement in Ad Manager 2 to rewrite those domain
Click to expand...
That's actually pretty cool.

I'm not familiar with Ad Manager 2, but I'll need to look into it to see how it does the rewrites.

The main thing is I plan to get rid of Viglink completely -- so I wouldn't want to redirect to their affiliate link. I prefer to link to ours.

The affiliate revenue we earn from Viglink has plummeted, so I don't want to run their code on our site at all, and want to figure out a way to earn more affiliate revenue without them.

Plus, I need a site-wide solution, even though the lion-share of the traffic currently goes to our forum.
 
askandy said:
That's actually pretty cool.

I'm not familiar with Ad Manager 2, but I'll need to look into it to see how it does the rewrites.

The main thing is I plan to get rid of Viglink completely -- so I wouldn't want to redirect to their affiliate link. I prefer to link to ours.

The affiliate revenue we earn from Viglink has plummeted, so I don't want to run their code on our site at all, and want to figure out a way to earn more affiliate revenue without them.

Plus, I need a site-wide solution, even though the lion-share of the traffic currently goes to our forum.
Click to expand...

Ah yeah... then you can do the rewrite for your own affiliate links.

Which prob makes more sense if you find the high earners. Here's an example of one of our rewrites:

Screen Shot 2020-07-10 at 2.15.26 PM.webp
 
arn said:
Ah yeah... then you can do the rewrite for your own affiliate links.

Which prob makes more sense if you find the high earners. Here's an example of one of our rewrites
Click to expand...
Wow, that's very similar, if not identical, to how I'm planning to make my software work. Didn't even know it existed.

Can it also just cleanse a URL, and append your affiliate id?
 
@Anatoliy @Masetrix @arn @ozzy47 @rdn @djbaxter

Just a quick and HUGE thank you for all your help & time these past couple days. It's been really very kind of you to help me/us.

I'm not sure exactly how I can repay the favor, but if you have WP on your site, I may be able to offer a nifty little plugin I just had developed.

It finds all articles within a certain time period, with fewer than X views (i.e. articles with fewer than 5 pageviews total over a 6 month period) and sets them with a 410 gone header.

There are many in the SEO world (Yoast, Ahrefs, SEMRush, as well as other SEO consultants) that believe that stale content that doesn't get many views can hurt a website's SEO.

The problem is that many sites have thousands of articles, and it's very hard to manually go through the articles, and see how many pageviews there are. If you only look at GA, you won't find articles with 0 pageviews over a period of time.

So, my plugin queries the articles in WP, looks-up data in GA, and then queues up articles to be marked with a 410 gone header.

If you could use it on your site, or want more information, hit me up via PM, and I'll be happy to give it to you free of charge.
 
askandy said:
Wow, that's very similar, if not identical, to how I'm planning to make my software work. Didn't even know it existed.

Can it also just cleanse a URL, and append your affiliate id?
Click to expand...

Yep, you can have it recognize specific urls and replace parameters on it.

Screen Shot 2020-07-10 at 2.42.00 PM.webp
arn
 
Anatoliy said:
what is your AdSense Ads CTR?
Click to expand...
We don't use AdSense. All of our "demand" network ads are served by/through AdThrive.

If your site serves more than 100k+ pageviews/month, you can apply to see if they would approve your site.

We don't get paid by CTR for demand ads, we get paid by impressions, so CTR doesn't really apply.

Generally speaking, the stats provided are based on RPM (Revenue Per 1000 Pageviews), and RPM can vary widely depending on the site. TBH, I can't even find a CTR report inside our AdThrive dashboard.

Some sites can yield $5-10 RPM, while others can get $50+ RPM. It just depends on your site's particular vertical.

For all of our direct advertisers where we use the Xenforo & WP ad managers to serve ads, we do track CTR. But again, CTR ranges quite a bit, depending on the ad and the placement.

Generally speaking mobile CTR is MUCH better than desktop. For example, we have 3 mobile ad placements on the forum, and the CTR ranges from 0.22% to 0.54%. Desktop ranges from 0.05% to 0.10% currently.
 
askandy said:
Yeah, we're trying to consolidate our ad placements even further -- especially on desktop.
Click to expand...
what do you get in Search Console under Core Web Vitals? Google says that largest contentful paint should be less than 2,5 sec, you got like11, so what, you have there like all 150k in red?
 
we use Adthrive, so it's possible with ads.

edit: to be fair though, this partially depends on your audience mix. If you attract a lot of users with slow connections, it's harder, since this is based on live data.

Screen Shot 2020-07-10 at 7.35.20 PM.png
 
@arn What did you do in mid/late June to take your performance from Need Improvement to Good?

We use AdThrive too, plus we have some other direct ads served through WP & Xenforo ad managers
 
askandy said:
@arn What did you do in mid/late June to take your performance from Need Improvement to Good?

We use AdThrive too, plus we have some other direct ads served through WP & Xenforo ad managers
Click to expand...
These stats are lagged so it's hard to say. But we were doing ok on performance before I think. A lot of that was CLS.

  • CLS was a big issue. So we set up specific fixed-sized divs that ads would get inserted into. So there would be no shift. Sometimes that results in ads with large gaps around them. I had to work with adthrive on this.
  • I reduced our font files around that time: link
  • I lazy loaded our avatars, which was tricky -- since the current addons don't do it. I ended up doing it through cloudflare workers.
  • We use cloudflare to cache our guest pages, and do a custom purging system so things don't get outdated.

You should see what you are getting dinged for on those pages -- it makes a big difference on what you should optimize.
 
Anatoliy said:
Teach me, masta!
Click to expand...
So the caveat is that I was already using Cloudflare workers for our thread pages. Cloudflare workers have an incremental cost. But I was already paying for them, so it's easy to add more functionality to it.

Cloudflare Workers are basically javascript run on the edge servers. So they can manipulate the page on the fly before its delivered to the end user. One thing they can do is rewrite the HTML before they get delivered to the end user.

So cutting out all the rest of the code, here's what does it:

Code: 
addEventListener('fetch', event => event.respondWith(handle(event)))

async function handle(event) {

    let response = await fetch(event.request)

    let html = await response.text()

// Lazyload Integration
    html = html.replace(/<img src="([^"]*)" srcset=(.*?)class="avatar/g, '<img src="" data-src="$1" data-srcset= $2 class="lazyload avatar')
    html = html.replace(/<img src="([^"]*)"(.*?)class="avatar/g, '<img src="" data-src="$1" $2 class="lazyload avatar')
    if (!html.includes("lazysizes.min.js")) html = html.replace( /<\/body>/ , '<script src="/js/sv/lazyimageloader/lazysizes.min.js"></script></body>')

  return response
};

It basically replaces the avatar images with the right format to do lazy loading. If the lazysizes.js isn't included in the page, it adds it.

It would make more sense to do it via a XF addon... and maybe I'll look into that.

arn
 
arn said:
I lazy loaded our avatars, which was tricky -- since the current addons don't do it. I ended up doing it through cloudflare workers.
Click to expand...
My little lazy loading addon adds lazy loading also for avatars:
xenforo.com

Unmaintained - Lazy Load Images

This add-on adds 'lazy loading images' to your forum with minimal impact. This will have a positive effect on your PageSpeed score, especially for the mobile test. Simply check the difference with and without this add-on. Loading image lazy...
xenforo.com xenforo.com
..but as you said, too bad that the tag this add-on adds has no Safari support. And that is your main audience ;)
 
You must log in or register to reply here.
Top Bottom