This topic is for discussion about the December 2025 Identifiers Survey blog post We welcome constructive feedback about the survey and results.
Is there a way to sort the results by category? For example, could we see how the 100K+ identifiers group voted compared to the 2-4K identifiers group?
9 Likes
Pretty good post and happy to see the results!
To add another interesting metric would be, what general taxonomic group? People involved with snails may be really happy or find some things more difficult than people involved with Eels.
Overall it is a step in the right direction and i look forward to future posts.
17 Likes
One thing that wasn’t addressed in the survey is the difficulty of dealing with notifications as an identifier. I wish there was much more ability to filter, sort, mark “read” and “unread”, etc.
29 Likes
Notifications can be a real mess and an utter disaster for some by becoming nearly unusable. You’re very correct.
This potentially little known feature could possibly help management of them, though only slightly.
1 Like
Is this cutoff of 2,000 IDs a common thing? I thought the cutoff for surveys and experiments was more often set at a level like 25,000+ IDs? Because I’m not at 2,000 IDs yet, but I could easily hit it, if it allows me to engage in surveys etc.
One thing that I’m grateful for, due to my low number of IDs, is that my notifications are very manageable. With some days having NO notifications at all. So I can easily read and respond to all of them.
I currently view excessive cognitive load as the biggest problem in our modern world. And getting worse all the time. It doesn’t seem fair that the most prolific identifiers, who help the most people, basically get punished by being sentenced to an unmanageable storm surge of notifications, with many being unimportant.
4 Likes
One thing I didn’t mention in the survey (because I forgot about it until straight after completing and going to do some identifying) is how slow photos stored on Amazon are to open in Identify sometimes. At times I abandon any observations with Amazon photos, because it can take 30 seconds or more for a photo to display, and I’m not patient enough to wait. If I can’t enlarge the photos, or easily switch between several photos in the same observation, I just move on.
I know I have seen posts here suggesting that if Amazon photos are slow it is because the user’s ISP is slowing them down, but I’m not convinced that is the issue.
8 Likes
Yes to this!! I don’t know the cause, but slow-loading photos (both on the web platform and especially on the android app) are my number one reason to make me give up identifying and go do something else.
2 Likes
No difference in speed on Amazon photos for me, but as this is a specific technical issue that seems to affect a minority of users, I don’t think it would be a good target for something like a broad-based user survey. I would suggest a Bug Report is likely more useful to report this and effect change. There are a lot of existing ones that may have some suggestions for how to address this ( https://forum.inaturalist.org/search?q=amazon%20%23bug-reports%20order%3Alatest including responses from staff on some). If you don’t find anything in there relevant, you could create a new one and give details on how it is different from previous ones.
5 Likes
I’d love to have the opportunity to trade ID expertise with someone: I’ll teach you some IDs so you can help me get a block of observations to research grade, and you can do the same for me. Has this kind of thing been set up before?
8 Likes
While I don’t know about the 2000 ID number, I suspect it was chosen for the purpose of the survey. I honestly think it’s a fair cutoff - at that point, it shows engagement in the community, and also that someone’s probably experienced a fair enough range of the usual identifier aggravations to be able to meaningfully contribute to the survey.
when I started identifying more, I shut off notifications for agreeing IDs, and it made life so much easier. The downside is that I don’t see when my own observations hit RG, but such is life.
10 Likes
I only mention it because in question 4 on the survey, “What parts of the identification process do you find difficult?” one of the factors listed was speed, “Identify page is too slow”. I forgot when I was filling out the survey that that was a real issue affecting my identifying behaviour. I didn’t rank it as highly as I should have.
6 Likes
One thing I’d find interesting in the results summary is a measure of “experience bias” in the responses. Question 7 classifies respondents into categories by numbers of identifications - is there a way this could be expressed relative to the number of users in the various categories? For example, the largest number of responses is in the 2000 to 4999 IDs category, but I suspect this group is much larger portion of the user base that was invited to respond; and conversely, the 100K+ IDs category seems over-represented in the respondents (unless there are a lot more of those dedicated people than I imagine).
3 Likes
I think the survey was only presented to people with a minimum ID count. IMO that’s completely fair, especially for a first pass. Maybe on the next one they can lower the minimum to 1000 and see how that changes things.
In any case, I’d like to have these surveys quarterly or twice a year, with the data shared publicly as this time.
6 Likes
I miss the point for frustration, that observers have at the beginning have no clue what they had found (not even a broader suggustion like “animal” or appropiate for the “thing” they found) and after a first suggestion to species level, they only need to agree and get “research grade” for their observation. That’s not my definition of what should be called “research grade”.
That’s one larger part of my own reasons, that I reduced identification to a minimum (at some worst wrong cases I unfortunately can’t resist to help to improve for my region).
According to your resumé of the questionnairy, it seem to be that I’m the only one, who is caring about that, but I know there is a forum topic about that point of view somewhere.
4 Likes
I open the image in a new tab and zoom. Not going to wait, there are gazillion obs queued up!
Not yet - but we do have a forum thread for annotations and the IdentiFriday thread.
There is an open request to rename RG. Two agree is, ultimately, not how any of us would define Research. But iNat has its own jargon and its own definitions. Casual is another disconcerting one - Casual Friday come in pyjamas ?
Notification management has been an issue since 2019 (and my post about it is a year old). Hope that is top of the list. Please. We lose identifiers altogether due to that.
7 Likes
Yay! thanks for summary. Excited for next product update post.
Especially the emphasis on CV control, observer and identifier onboarding and better help pages, streamlined identify on mobile in survey results. Hopefully newer surveys will cover more long standing things missed before.
I do wish there comes a better iNat voting algorithm change as noted in survey result for cases of disagreements. I currently don’t have robust proposal solutions but the valid issue of voting algorithm on disagreements is discussed in forum few times and to summarise better as this comment of @tristanmcknight
4 Likes
Thank you for these initial results! I regret not having been contacted for the survey; I assume it was based on a sample of identifiers.
Some points have already been mentioned above, but my main frustrations are:
- Notifications: it would be great to be able to categorize them. I receive dozens of them every day, so I only look at those where a comment has been written (because I see words “comment” or “mentioned”). However, I would also like to see those where someone has added an identification that contradicts mine. Or to see the identifications made on observations that I have bookmarked.
- “Thanks ID confirmations” are a disaster (and btw account for the majority of the notifications I receive). I regularly try to tell people to only confirm if they are sure of the identification I have proposed, but most people ignore it. It’s annoying to have most of the observations in Research Grade based solely on one serious/expert opinion, and it adds another weight on our shoulders (even if we already try to give high quality ID, we sometimes make mistakes). Perhaps we could create a new stage called “Confirmed Research Grade” (or whatever name we choose) that is only reached if the observation has been identified by two people other than the observer, unless the observer has reached a certain number of identifications in this taxon (or any other rules that allow us to estimate that the observer has a certain amount of experience and has given a serious identification)?
- It could be nice having a way to add a “unsure ID”, when for example we suspect it could be a species, we don’t want this ID to be used by the CV or be used as a truth, but we would like the observation to be filterable more easily in the futur rather than stayed in a very large taxonomy level.
- The difficulty of removing a bad annotation: 3 downvotes are not enough (I think you need 4?), whereas given the small number of people who make annotations, I think 1 downvote is more than enough for the annotation to be reset.
- The difficulty of adding annotations: for example, for the Life stage annotation, in the Identify tab, you have to go to each image one by one to add it. Whereas very often, you can see from the thumbnail image whether the insect is a larva/adult/other.
- The lack of filters in favorites: I would like to be able to filter by taxon and annotation.
- In the Explore tab (and identifier tab), I would like to filter by Observation Fields (by parasite, by food plant, etc.).
Other than that, I really enjoy identifying species on iNaturalist. Thank you for everything you do.
12 Likes
afaik inat doesnt reset annotations from votes as of now? i guess u only saw correlation where owner of observation/annotator explicitly reset (so the only solution now is to mention those two). its popular pending request
u can try Announcing the Universal Metadata Tool Beta for the time being
4 Likes
afaik inat doesnt reset annotations from votes as of now? i guess u only saw correlation where owner of observation/annotator explicitly reset (so the only solution now is to mention those two). its popular pending request
You might be right ! Maybe a small bug that makes the annotation visually disappear when downvoting, but still there. I upvoted the pending request 
u can try Announcing the Universal Metadata Tool Beta for the time being
Thank you @einsum for taking time sharing this @megachile tool, it seems to be a nice workaround. I am going to try it !
2 Likes