Dropping ad revenue

All this is very interesting, but the spirit of the subject (at least the beginning of it) was some folks reporting they were earning 10 times LESS ad revenue by switching from one forum software to discourse. And it wasn’t coming back, even after time, and they didn’t manage to solve that issue, to the point where they were thinking about stopping using discourse. Infinite scrolling and the “lack” of pages seems to be believed to be the cause, although I don’t think this has been confirmed in any way.

I got until here that the “problem” seems to be very Google adsense related. Now, I understood that how many ads are requested/displayed may have an importance, and the content of where they are displayed may also have one.

The question quite remains of why these people were able to earn 10 times more with just a different forum software. Maybe they were displaying “too much” ads before. But they were able to. Is it because you can show more ads with different pages than you can with infinite scrolling ? I can’t help but to notice that “clickbait” websites filled with ads make you surf a lot of pages. They are designed this way. They are not using infinite scrolling.

If a user takes time to read a topic and goes through 20 or 30 pages with a “pages” forum software, is it possible that there is no way you can display as many ads with the same amount of “scrolling” with discourse ? (with google adsense, I mean). I see the URL is changing when you scroll in discourse (my first thought was too much ads were maybe displayed on the same URL - But maybe it is still considered the same “page” somehow !?)

Anyway, the base question remains, I believe: Why are some people reporting this “ad revenue drop” with discourse compared to a “page forum” ?

7 Likes

I am also having this issue. Are you using any cloudflare service?

1 Like

Let’s take another step back and look at online ad revenue.

Online ad revenue is primarily of three types, CPA, CPM, and CPC. Each of these models have different characteristics.

Let’s assume you are talking about Google Adsense CPC display advertising.

CPC revenue will be realized by this simple forumula:

Total Ad Revenue = Total Ad Clicks = Total Ad Impressions * Click Through Rate (CTR)

So, let

  • TAR = Total Ad Revenue
  • TAC = Total Ad Clicks
  • TAI = Total Ad Impressions
  • CTR = Click Though Rate

Then, obviously:

TAR = TAI * CTR

If your TAR has changed, either your TAI or CTR (or both) have changed.

What are the possible causes of each?

Total Ad Impressions (TAI)

There are many reasons for TAI to fall:

  • Total traffic to the site (page views) has decreased.
  • Total ads per page have decreased.
  • Other factors (this is not meant to be a exhaustive bullet list)

Notes on TAI:

Search engine referrals are a primary driver of revenue for many, but not all sites. You must compare your search engine referrals between various sites to determine which web site has better SEO. There are more opinions on this topic than we can imagine, and many are just pure speculation, and every site is different; so it’s best to just look at your own numbers (traffic stats). The same is true for total ad impressions, in general.

Click Though Rate (CTR)

There are many reasons for CTR to fall:

  • The ads being served are not interesting to the users (ad context issues).
  • Total view-ability of ads has decreased (based on placements, scrolling, etc).
  • Users are engaged with the content and not interested in clicking on ads.
  • Other factors (this is not meant to be an exhaustive list or discussion)

Summary

It’s not really possible to answer the question of:

"Why are some people reporting this “ad revenue drop” with discourse compared to a “page forum” ?

… unless “some people” (whomever they are) provide detailed site-to-site statistical comparisons of CTR and TAI for various forum software and ad traffic models and stats. However , in closing, here are some general possibilities:

  • Sites with infinite scrolling might have less overall pages in the search engine indexes. However, I do not have enough data to say either way. It could be the opposite, I simple do not know (yet). There are discussions about this on the net and can be found from search analysis.

  • Sites driven by javascript may be indexed differently by various search engines, since each search engine spider / bot handles javascript differently. Some have speculated that long infinite scrolling pages are not as fully indexed compared to the shorter “pages” model, because (speculation by others) is that search engines are not designed to “scroll” like humans and so therefore, the pages are indexed differently. Personally, I do not know the answer to this, but many discuss this on the Internet and there are many opinions about this. You can search and read them.

  • Sites which are more fun and engaging, like some Discourse sites, may have less CTR because users do not want to click on display ads and just “scroll past them” and engage with the great forum and wonderful content. Anything is possible. CTR stats and analysis will help answer these types of questions.

  • Other possibilities, but this is not meant to be a full study into this topic.

Brief Conclusion

Each web site must look at their own traffic statistics to determine what is going on with their site(s). You can study your traffic and overall page impressions using Google Analytics; and you can study how many ad impressions you are getting and the CTR with each ad network. If you run you own ad server, you can look at those statistics.

In general, many people have speculated over the years that javascript driven sites which are based on the single-page-request model have lower SEO than their counterparts. There are many discussions about this on the Internet and are not germane to Discourse, per se; but apply to all javascript framework driven sites and their SEO characteristics.

Finally

I hope my test page I tossed up a few days ago helped a few people, in some small way, to understand that there is no problem with iframe-based ad tags being loaded in a Discourse topic. This test page will be “torn down” tomorrow as I need to use that VPS slice for another task. Note: That test page did not address async javascript tags.

Also, hope the short summary above, which is not meant to be exhaustive nor complete, on a few aspects of how and why CTR-based ad revenue can “fall” was useful; and kindly forgive me if I drop off this discussion for now. Hopefully, this “bird’s eye view” was helpful to one or two of you in some small way.

18 Likes

Yes. It would be helpful in troubleshooting if those who have experienced ad revenue decline could provide some additional numbers like these.

2 Likes

I do think you answered it very well :+1: And people with this problem should indeed look at all these aspects and maybe provide them here if they want to discuss that subject further.

I compared the ad reports between two periods - in the old forum (2017) and discourse.

It looks like a combination of several factors are at work.

  • A tenth of the earnings is THE problem of course.
  • Fewer page views that never got back to the old numbers
  • Even fewer impressions (the result of some blanks?).
  • A quarter of page views but only a sixth of clicks - clicks’ drop is steeper.
  • Lower CTR - why?
  • Lower CPC - did it generally go down over the years?
Old Discourse
Earnings 10x x
Page Views 4x x
Impressions 5x x
Clicks 6x x
CTR 1.5x x
CPC 2x x
Page RPM 3x x
Impression RPM 2x x
5 Likes

How did your user base respond to the migration? Did they stick around? Did you lose a percentage? If that’s at play we’re not talking about just the ad plugin being at fault (but don’t want to negate it either).

5 Likes

You should also try to have a look at the amount of visitors from search engines (mostly Google). Someone mentioned it above, if I remember correctly. About what Justin is bringing up, maybe your members came as much as before, but it’s on the visitors (non members) that the difference may mostly be coming from ? (I don’t know if you still have access to that kind of data. Hopefully, yes)

1 Like

Online ads have collapsed since 2017 for a lot of sites, regardless of software, so that’s another variable… you can’t expect to get 2020 rates in 2017 even if you changed nothing.

9 Likes

Agree and we need to discount for that.
Hence I noted “Lower CPC - did it generally go down over the years?”

But ad revenue should not be impacted by visitor or member?

Hard to estimate but a poll after migration generally gave positive results from the small group that responded.

2 Likes

I don’t understand what you mean.
You have your members, who are people registered on your forum and posting messages (and also reading, obviously). And then you have people only reading, not registered. A lot of them may only come from search engines because they are redirected to your forum thanks to a search they did there. It’s people that come once in a while, read a few things, and leave (for most of them).

If search engines used to send you a lot of these , and stopped (maybe for different reasons, but the indexation of discourse and infinite scrolling may be one of them - I know there have been discussions on the subject) that may explain your ad revenue drop. Which would simply be: Less people coming, because less are sent by search engines.

About what Justin said:

You wouldn’t even notice it unless you analyze it.

3 Likes

Just occurred to me that perhaps there might be a connection between blank ads and the page change not being pushed:

It’s a long shot, but it seems plausible to me (understanding that I’m mostly ignorant about these things). Just trying to brainstorm things within Discourse that could be different from other forums.

4 Likes

This seems to make sense to me, and could be the explanation of what I suspected above:

An option to be able to set a certain number of posts scrolled (really displayed, not if someone just jumps x posts further) after which a new URL is “pushed” may then solve the ad revenue problem this discussion was all about ? (would it be the same as having clicked to a new page ?)

On the other hand, it would first scroll back x posts when people press the back button in their browser. But if it’s an option, each forum admin can choose knowing the pros and cons (if it indeed works to solve the discussed problem).

1 Like

For anyone reading this… monumetric.com is working to monetize our Discourse forum when we thought all hope was lost.

Message me if you have questions about it.

3 Likes