Hello everyone,
I propose opening a thread here dedicated to truly understanding the Vinted algorithm.
The goal is not to share impressions or feelings like « I had a lot of sales last week » or « I think the algorithm likes this ». This type of information is unfortunately too random to be useful.
The idea is rather to centralize all concrete, technical, or verifiable information regarding the platform’s operation.
For example:
- How search ranking works
- Possible criteria for highlighting listings
- The role of favorites, clicks, messages, and sales
- Frequency and effect of relisting items
- Impact of listing modifications
- The role of photos (original / AI / compression / metadata)
- Interaction between account activity and visibility
- Quality signals used by the platform
- How the recommendation engine might work
- Any observable or documented technical element
Particularly useful contributions could be:
- Comparative tests
- Reproducible experiments
- Observations on large volumes of listings
- Information from technical documentation, patents, or engineering of similar platforms
- Elements confirmed by responses from Vinted support
In summary, the idea is to move from feeling to analysis.
If we gather enough solid information, we might be able to start identifying the real levers for improving visibility and sales.
Thank you in advance to those who agree to share their observations or research.
Goal: understand the system rather than be subject to it.
