Hey @Tanklet I’m wondering what exactly the criteria is for the models that are selected when searching on makerworld using the “Relevance” filter.
I’ve produced a number of Pokemon keychain packs which are moderately popular, but when I search for “Pokemon” using “Relevance”, my most popular pack doesn’t show up until around the 230th search result.
If I change the filter, searching “Pokemon” using Boosts, Likes, Collects, or Downloads, more than one of my packs appear in the top 10 search results.
Since “Relevance” is the default search criteria, i’m wondering what exactly does the algorithm take into consideration when searching with that tag, so I can help boost my models closer to the top of the page.
I would assume that since my models score so high in all the other search results, they would appear at a similar level with “Relevance” too, but that’s not the case
Any info you can share would be appreciated! Thanks!
Thanks for the clarification here, @Tanklet. I’m wondering if this ranking is working as intended then? My models rank really high in downloads/likes/boosts/collects, but aren’t showing up in a similar position with relevance.
I am using all 50 tags in my model pages, in the hopes that the models will be discovered more - and I started adding extra descriptions in the name too, all hoping this would add to increased discoverability. But if what you’re saying is accurate, it looks like that extra info i’m adding is actually working against the discoverability of the models?
Is there a middle ground here? Should I remove tags and that would actually help? That kinda seems counter intuitive to me.
This is interesting. I’ve varied my use of tags between models (usually out of forgetfulness or laziness tbh), but to me it seems like it shouldn’t really make a difference how many tags there are. But I will keep this in mind
Yeah, I never thought I would be punished in the search rankings for using more tags, which I thought would make things easier to find. Seems really counter-intuitive to be honest, I hope they change it!
No, you trade discoverability for ranking. Using fewer tags increases their value as search terms, enhancing their impact.
That depends. Do you want to appear in results for all search terms, or rank higher for specific terms that accurately describe the model? You can experiment with the tags, as search results update in real time.
OK that helps me understand this a lot better! - I still don’t totally agree with that, since I think more tags would be better, but this makes sense. I’ll monkey around wiht the tags on one or two models and see what happens
