Hey, I am finally back at a PC.
So here are some better screenshots of my data.
As you can see, the user activity seems pretty normal to me.
The likes and collects even did go up due to a new release.
But prints dropped massively. For most of my models, you would usually print more than one. That resulted in mostly having more prints than downloads. Not for all, but the rate is usually around 1:1.
Some examples of those models:
I can not argue that they are hiding my models or making them less visible, as also the views did go up. I am also regularly checking that the models are well-placed within their keywords and titles so users can find them when they are looking for something like that. (I am a web guy, so yeah I did on-page SEO)
I put some effort into the descriptions and visual representation of the models, with better high-quality images to also better show the expected results for the newer ones. Very few CAD screenshots, and only if they would benefit the user instead of real pictures, but always backed with real pictures.
From what I can see, tested, I think @Uhl already pointed out.
There is a limitation on downloads for a while.
You will only receive 1 download from a user per model.
It seems to be new that the same behavior is now also applied to prints.
If that’s true, it is indeed a bummer. I can verify that a first download and print per model is counted. Additional prints seem to be ignored.
This makes it unattractive not only for systems but also for larger more complex models to be shared in a single models, that have multiple build plates to be completed.
Examples:
Because it means that a single filament clip is as much worth as a Big Ben for example:
I like this one 
. But it’s multiple Kilograms, and in this case 63 hours of prints in contrast to 5 minutes.
When I read what they like to see from models or also take into account the feedback from the questions regarding less popular models:
It seems to be the wrong turn if they want to achieve high-quality models, that people like and print. For now, I still hope there is a bug in the reporting from the slicer to the platform.
I also checked the new BETA of Bambu Studio because there seem to be some updates to the network component, but it still did not count repeated prints/prints of different plates.
Edit 1:
For me, the prints are more interesting than downloads. Since I can check if a version finds adoption and what the preferred sizes of the users are. Since my models mostly should not scaled in the slicer, I always provide the source models also, so the users can build their own if no profile matches. Having no access to the real prints anymore in the data, it makes the analytics more or less useless to me.
Edit 2:
I´ve also noticed that some data seems to be lost in the analytics. If I switch to the monthly view, there seems to be a limitation, or they only keep data for a certain amount of time.
It also makes the analytics feature, even more unreliable. And would force me to extract the data on my own if I want to see a long-term history.
Maybe it´s time for a little crawler to persist the data long-term on my own?
