New, harsher algorithm with faster drop off in page rank?

[DogSauce]

So, I am aware that the CB algorithm is based largley on the amount of colored (preferably purple) viewers + tipping action.

I have however observed in the room where I hang out that the requierments for staying high in page rank seems to have become higher.
This girl can get good tipping going and have around 5-6 purples which would probably land her at around page 2.
But then as soon as the tipping ppl leave, it just seems to die off so fast.
The viewer numbers decrease rapidly and the page rank tanks to page 5-10.

Up until recently it has been my impression that landing on page 1-2 would keep you there for a while but now the drop off seems to come very fast.
Any1 else who has observed this or any other tendencies that would indicate a change in the algorithm?

 

[heavensangels]

Tipping is only one part of the algorithm, as the number of unique registered viewers decreases so does the ranking. I have always noticed that the unique registered users are more important long term than the tippers.

 

[Dan Epstein]

Do we know if there’s a direct link between the frequency on the homepage, to the actual amount of converting traffic? In SEO, the homepage is usually the less converting, and with the highest bounce rate. With that being said, how do we know if this rank directly effects the performance sales wise? We are after all, not talking about SERP, which is search engine results page. Just my two cents as a professional in digital marketing

 

[Smoothie]

Do we know if there’s a direct link between the frequency on the homepage, to the actual amount of converting traffic? In SEO, the homepage is usually the less converting, and with the highest bounce rate. With that being said, how do we know if this rank directly effects the performance sales wise? We are after all, not talking about SERP, which is search engine results page. Just my two cents as a professional in digital marketing

I guess the only thing you could really count on is more impressions for the model’s preview thumbnail. Whether they get more traffic or tips from being discoverable on a higher traffic page (first page or thereabouts), is probably always a bit if a game of chance.

It still boils down to how many visitors prefer that model’s “type,” AND happened to be online during the model’s broadcast, AND browsing (instead of just visiting a followed model), AND found that model’s thumbnail preview before settling on a different model first, AND had tokens to spend. None of those factors can be fully controlled by CB or by the model. Best you can do is cast the widest net possible and hope to catch a big fish or two.

Home page (Page 1) would be the “widest net,” although your point is totally valid too. As with any website, a pretty big portion of the traffic are people passing through, not necessarily fully engaged. If a model is on Page 1 and there are 100,000 visitors on that page, realistically only a fraction of those visitors is looking to convert their experience into spending time and tokens in a room.

 

[tokensarentforkids]

the models on the front page stay there because the viewers stay there, even an hour after the broadcast ends many are still there...

 

[DogSauce]

I guess the only thing you could really count on is more impressions for the model’s preview thumbnail. Whether they get more traffic or tips from being discoverable on a higher traffic page (first page or thereabouts), is probably always a bit if a game of chance.

It still boils down to how many visitors prefer that model’s “type,” AND happened to be online during the model’s broadcast, AND browsing (instead of just visiting a followed model), AND found that model’s thumbnail preview before settling on a different model first, AND had tokens to spend. None of those factors can be fully controlled by CB or by the model. Best you can do is cast the widest net possible and hope to catch a big fish or two.

Home page (Page 1) would be the “widest net,” although your point is totally valid too. As with any website, a pretty big portion of the traffic are people passing through, not necessarily fully engaged. If a model is on Page 1 and there are 100,000 visitors on that page, realistically only a fraction of those visitors is looking to convert their experience into spending time and tokens in a room.

Valid points. I've always prefered to compare it to a "net" as well. Being at front page increases your chanses of being discovered by potential buyers substantial. That being said there could very well be models on lower pages that often make more because they've attracted a few people that pay a lot for a particular niche or look.

My recent experience has just been that CB's rankings switch a lot faster and perhaps the requierments for activity has increased.
Even models that used to be mainstays on page one I've seen struggle around page 2-3.
Maybe it's just that the competition has increased or CB has chose to go for an algorithm that opens up more for new models that does good.

 

[DogSauce]

the models on the front page stay there because the viewers stay there, even an hour after the broadcast ends many are still there...

Ye many of the users in these rooms are probably "fake" accounts or accounts that comes from pop ups or services.
Can't say for sure but some of the top models definetly has a way to ensure they never start of from 0 viewers.

 

[DogSauce]

Tipping is only one part of the algorithm, as the number of unique registered viewers decreases so does the ranking. I have always noticed that the unique registered users are more important long term than the tippers.

For sure. And I seem to recall that CBhours also pointed out before that the amount of tokens that these account has makes a difference too.
But it seems like latley that as soon as these ppl leave the page ranks drops instantly. No time to stick around and fish for new big tippers.

 

[Dan Epstein]

Well, what is the source of that new traffic, affiliate, or organic? What are the keywords that people look for that convert into tokens? Algorithm is a nice buzzword, but in this case a misnomer. CB does not act like a search engine of an e-commerce platform-I.e Amazon.

 

[DogSauce]

Well, what is the source of that new traffic, affiliate, or organic? What are the keywords that people look for that convert into tokens? Algorithm is a nice buzzword, but in this case a misnomer. CB does not act like a search engine of an e-commerce platform-I.e Amazon.

Well in general when speaking of CB and the algorithm it is in regards to how the site postion rooms in the page rankings. Even if it is not a known algorithm there are some factors that can safely be said to play a part. Like the number of registered users and the number of tokens these people has tipped recently. Stable connection and good quality camera seems to also be favored. CB tends to switch up the algorithm every now and then from what I know so that it can't be manipulated and probably also to serve a different purpouse if they feel like the site needs a change.

Models can of course market themselves and use services that could potentially affect the amount of viewers they get. But what I am mostly interested in right now is if anyone has observed any changes to the algorithm or "the flow" that would indicate that page rank drops much faster if a model stops recieving tips or the high tipping accounts leave.

 

[cbhours]

I seem to recall that CBhours also pointed out before that the amount of tokens that these account has makes a difference too.

the opposite, the amount of tokens the account has makes either no difference at all or so little difference it cant easily be tested for.

 

[Dan Epstein]

We don’t have evidence that page rank matters to the actual performance in terms of sales, so we keep comparing apples to oranges.

 

[Smoothie]

We don’t have evidence that page rank matters to the actual performance in terms of sales, so we keep comparing apples to oranges.

True, in the sense that there is no set number of visitors that guarantees any set amount of token expenditure. And, in the sense that it’s fully possible for a model with one visitor to earn more tokens in a shift than a model with a hundred visitors. There is no keyword or algorithm that can prove or disprove the whims and free will of a visitor with tokens (or with the money to buy them).

But, not knowing which visitors will spend and which ones won’t, the best a model can do is try to be discoverable to the most visitors possible. Which doesn’t guarantee any of them will spend, but it makes sense that a model would want their page to be found by as many potential spenders as possible.

The algorithm does matter; calling it a buzzword is a bit dismissive IMO. It’s just that the people that want to know the specifics don’t have the necessary info to fully understand it anyway, despite having a valid reason for caring about it. I don’t blame anyone for being curious.

 

[DogSauce]

the opposite, the amount of tokens the account has makes either no difference at all or so little difference it cant easily be tested for.

Oh, sorry. Didn't mean to missquote you. My memory just failed me. I really do appreicate the work you do to figure out the workings of the site.

 

[DogSauce]

We don’t have evidence that page rank matters to the actual performance in terms of sales, so we keep comparing apples to oranges.

I never said anything about sales. This question was strictly about the algorithms effect on page rank.
Not that sales isn't an interesting topic. I don't mind that the discussion branches out.

 

[Olive]

I personally think that on a site like CB being more discoverable is always the better option.

Why? Because tippers (and their ability to tip) are transient. If you have no visability then when your current base is slowly fading, there is next to no way to regain others (and if you can then very very very slowly).

How do I know this? I used to do alright 2 years ago, until the same thing happened to me. Low-ish ranking and tippers fading out.
Now I'm ranking a lot better (FP) and the diversity in my tipper-group is so much higher. Do I make the same as some models with big whales on lower pages? Not even close.. But I have a lot more peace of mind that I don't need to rely on a very select few. And a bigger shot at higher tippers seeing me now too.

On the topic of the algorithm changing: I'm not sure I have seen a recent large change like that myself (yet). There seem to be a lot of factors that go into placement (a lot of which are also "indirect" factors). Specific viewers leaving has had the ability to make quite dramatic shifts in placements for quite awhile that I'm aware of. The higher up the placement the lower the votality (usually) due to being able to more easily have an influx that balances out the outflux (because of visability on the site).

Things like show content & tipper and viewer behaviour in the specific room can have an impact on the change you are observing without the algorithm playing a different role then before.

There might be something going on though. CB actively tells models that they constantly tweak the algorithm, so if that's true then it might play a role. IMO as a model it's always good to observe/analyze/adapt and try to finetune things no matter what.

 

[Dan Epstein]

There is a chance though, that the ranking on CB’s pages is not algorithm per ce, but more like a comparison website or a media buying platform. That means putting the model with a conversion and profit that best fits the website there and may increase profit to the website, not the model necessarily. There’s a model affiliates use, it’s called Earn Per Click. PPC’s like my self rarely use it, but we understand the logic behind it, and it’s not impressions, adult industry doesn’t give a damn about impressions as a metric in general( wrongfully if you ask me). They historically take the whole amount of clicks, minus anything they may have cost, and divide the total profit with it. Models with a higher EPC can get a higher rank. It’s a shame that we’re in streaming, yet video performance metrics are not used- show completion and engagement for instance

 

[Bellboy2829]

I’ve heard that is the amount of users that actually have tokens in there account is the biggest contributing factor but who really knows exactly except CB themselves.

 

[Tinydickman]

The algorithm is very complex but one of the contributing factors is users with tokens