Please keep the post in one place
This should now not be possible as of a little over a month ago. The system was changed so that you canāt upgrade a nomination within the first 24 hours of submitting it, which is so that the Machine Learning has a chance to look at it. After this 24 hour period has passed, the Machine Learning will no longer reject a nomination and thus you are free to use the upgrade to get the nomination through the voting phase more quickly.
See this post for more details:
I appreciate the suggestions and ideas Iāve received from the community so far, but the core technical issues (the automated ML system auto-rejecting these valid nominations) remains unresolved.
Its not bug. Niantic have been working to improve AI but there is nothing you can do for now but appeal or resubmit if you want your nomination to pass AI check
Hi, I have merged your posts, letās indeed stay on the same topic for the same question.
Considering the number of thoughtful replies you got above itās a little unfair to the community to say no one helped you.
These are indeed very typical features of many historic European city centers, so I think the suggestions you got above relating to researching exactly why this specific frieze or bas-relief is more interesting than the one next to it, or why it is more important to nominate the decorative feature than the entire building (which would also be worth considering if weāre that interested in historical architectural heritage) will also make your nominations stand out and help you get past ML.
Hi, thank you for merging the threads, and I want to sincerely apologize for my previous comment.
You are completely right, and it was unfair of me to say ānobody helped me.ā That was poorly phrased out of sheer frustration. I am actually incredibly grateful for all the thoughtful advice, tips, and time the community members have given me in this thread. My frustration was strictly directed at the official Niantic Support and the automated ML bot, which keep rejecting things blindly, not at the players here who are genuinely trying to help.
Regarding your point about nominating the architectural feature vs. the entire building: that is a very helpful perspective! The reason I focus on these specific details (like the Savoy mascaron or the Neo-Baroque gable) is because the historical articles and architectural journals I found specifically highlight these exact ornaments as unique, standalone works of art. Sometimes the building as a whole is too broad, but discovering a specific, 100-year-old carved face hidden on the facade fits the āExplorationā criteria perfectly.
I will definitely take your advice and try to emphasize why the specific detail stands out compared to the rest of the building in my future descriptions, hoping it might finally satisfy the ML bot. Thank you again for your patience and for guiding me!
Hello
Iāve now submitted this nomination multiple times and carefully followed all guidance from the forum discussions ā including improving the description, photo quality, and supporting information.
However, the nomination keeps getting rejected during the initial AI/ML review phase and never reaches community reviewers.
The object in question is a unique, architecturally distinctive building with strong cultural/visual value, but it seems the AI system fails to properly recognize or categorize it.
At this point, Iāve exhausted my appeals and Iām not sure what else I can improve on my end. I would really appreciate guidance on:
- how to better present such ānon-standardā architectural nominations to pass AI screening
- or whether this is a known limitation of the system
I enjoy contributing to the game by discovering and submitting interesting locations, but this experience has been quite discouraging.
Thank you for your help.
Each appeal returns 15 days after being, regardless of whether the appeal was accepted or rejected and you have two appeal slots. This allows for up to 4 appeals each month.
Yes, Iām aware of how the appeal system works and the cooldown timing.
The issue isnāt really the number of appeals, but that the same nomination keeps getting rejected at the AI/ML stage before it even reaches community reviewers. Since appeals are limited, it would take several months just to retry a single nomination multiple times, which can understandably be frustrating for many users.
It would be really helpful if the AI system handled more unique or non-standard nominations better, or if there were clearer ways to improve submissions that currently get filtered out too early.
I appreciate the clarification though!
This doesnāt make sense 
Each POI should need appealing at most once. Therefore, if you have 12 different architectural features you want to be wayspots, this would take 3 months to get them all into appeal (if they are all rejected by the ML system).
If you are submitting the same thing multiple times, only one of those needs appealing.
Unfortunately noone but staff can tell you how AI work. My best advice is to use fresh foto on every resubmission . Try different zoom level mayb it will work. Old rejected photo will be rejected again by AI
I understand your point about each POI only needing one appeal in theory.
The difficulty Iām running into is that these types of nominations (architectural details) are consistently rejected by the ML system before reaching reviewers, even when I try to refine and improve them based on previous feedback.
In my area, these kinds of features are actually relatively rare and visually distinctive, so itās a bit frustrating that they donāt make it past the initial stage. Because of this, I end up needing to use appeals just to get a fair evaluation, rather than as a true exception mechanism.
Thatās why it feels less like a āone appeal per POIā situation in practice, and more like being stuck at the first step.
I do appreciate your explanation though 
If you find your appeals getting rejected and you feel the submission needs to be improved and then retried (and re-appealed), you could post here to ask for advice on getting the details right.
Not everyone would agree the POI is eligible (I myself am on the fence), but you would definitely get assistance and reduce āredundantā appeals.
Thanks, that makes sense.
Iāve actually been trying to improve each submission with new photos and adjustments, but the limitation on submissions makes it a bit difficult to iterate quickly, especially since I was submitting many locations during a recent walk/tour through the city.
Some of these spots are not easy for me to revisit regularly, so Iām trying to make each attempt count as much as possible.
Iāll keep your advice in mind for future submissions and try to approach them more carefully with fresh photos and framā
Thanks, that makes sense.
Iām still fairly new here, and this was my first post on the forum, so I appreciate the guidance.
Could you please point me to the right place or thread where I could get more detailed feedback on submissions like these? Ideally somewhere active where people can review and give advice on what exactly might be causing rejections.
I could also share a few of my recent nominations there to better understand what I should improve.
Thanks again for your help 
This is the correct category for posting individual contributions for discussion.
No one is able to tell you what the ML model looks for. It would be abused if Niantic gave us this information. All we can do is recommend that you take photos with a focus that make it very clear what you are submitting, and you do that. And that you link to evidence that this is something significant as opposed to just decoration. The appeals rejection you show does state that you can resubmit if you have additional context to provide.
I am also on the fence about details like this being eligible.
If something is within 10km of your base, take several photos, mark the location somehow, then you can submit at your leisure from your base (in Pogo or on wayfarer website).
Actually u can just take multiple photo, so if you resubmit, you can just use other photo
Yes but I did on the app these :c
I will do my best next time:)