Improve How "Unknown" Observations Are Treated and Displayed

The problem as I see it is that not everyone wants help with their IDs, there is currently no way to tell which observations belong to those people, and it’s not fun being at the receiving end of people’s annoyance when a coarse ID is added. As a result, I tend to avoid them. My suggestion was an attempt to find a way to let the observer decide if they want anyone contributing IDs yet or not. I get that the notifications could annoy people, too, which is why I was going to edit or withdraw this post while it was pending approval, but there is apparently no way to do that.

1 Like

I agree. Maybe that would be effective.

1 Like

Fixing the ‘vanishing placeholder’ issue might also help with this. I’ve noticed people getting annoyed when their placeholder gets deleted, especially considering sometimes the plaeholder is not intended to be one (if the app doesn’t ‘find’ the right name in the database it goes in as a placeholder which is deleted if someone adds a real ID)


I’m not really sure what to do here. I had wanted to withdraw this Feature Request while it was pending approval. I agree that the notifications it suggests may be at least as annoying as the current situation, so I’m not voting for the Feature Request myself as written, and I don’t want to edit it now after others have responded to it. So if the moderators want to do something to substitute topics about kueda’s suggestion of “better onboarding that teaches people about adding coarse IDs to their own observations, and encourages everyone to help out by adding coarse IDs to other people’s observations” and charlie’s suggestion of “Fixing the ‘vanishing placeholder’ issue,” that’s fine with me.

1 Like

I don’t think it’s a big deal. People can vote on it, if you think it’s a bad idea now, you can say why which will reduce its votes. Maybe Tony can close it.

I’m with you 100% here, and this is exactly why I don’t think an automated reminder would be super effective. In my opinion, the recipients of your comments appreciate your input because it is sincere and hand-written (or at least hand copied-and-pasted). If the system is the one providing the reminder, that feels less sincere, e.g. “This can’t be that important if they assigned this task to a robot.” Remember Clippy, or Jimmy Wales’ heartfelt midwinter pleas.

Honestly, if anything on iNat really wears on you like this, just don’t do it! I think welcoming new users is valuable, but it is really time consuming. IMO, it’s totally fine to just add identifications and follow up with an explanation if someone asks for one.

That is annoying, and also largely indicative of a failure on our part to communicate what iNat is about, i.e. people who don’t want IDs don’t seem to realize they’re on the wrong website, because all new observations need IDs on iNat, and if they don’t want external input they should just enter data in a spreadsheet or scream their findings into the void or something. It does suck to be on the receiving end of someone’s occasionally irate confusion (those of us on staff kind of get a lot of it), but while a measured response is always ideal, I think it’s also ok to just point people toward


The particular change I would find helpful for unknowns/uploads is something @charlie has mentioned more than once - some sort of draft mode and/or delay before posting.

On a few occasions I have done personal big days - it takes too much time in the field to enter names as I go, and even when sitting at home, it’s quite tedious to enter the names via mobile device when there’s the much faster option to upload them without names and go through them via the ‘identify’ interface on the website with filters set so only my unknown observations are showing up.

Uploading several hundred observations takes a while, and I try to get them in batches as they upload. Mostly I have names on the observations within 5 minutes of uploading, but even so, sometimes people add identifications in that time.

For me personally, coarse identifications aren’t a big deal, but they do seem like a bit of a waste of time/attention of the community when I will typically be adding much more refined IDs shortly, and there are plenty of other observations that could use the attention.

Another scenario that occasionally happens is someone will add a fairly specific ID that I know is wrong (this seems to happen relatively less often when there is already an ID there). Of course I disagree with it, but most often the original ID remains, so unless I reject community ID, the observation will show up as the common taxa between the two IDs. It’s also nothing more than a minor annoyance (and from a community stand point, less of an issue than the allocation of attention to observations that won’t really need it) but it would be easily avoided by having some ability to have a draft status or small delay (10-15 minutes?) before having unknown observations show up in the community Needs ID pool.




agreed except with my ‘big days’ on my slow home internet connection, i won’t be fully uploaded even in 15 minutes. I am trying to remember which site it was that had a draft mode after uploading, i want to say flickr, but it worked really well. I review all of my observations anyway to look at photos and range maps, add fields and annotations, etc.


It doesn’t “really wear on” me. Since we don’t know each other and tone can’t be read through written word all the time, I’ll just state for the record that I’m not actually complaining. My instincts and the way my brain work cause me to typically try to find solutions for problems or improvements for things that might be working okay but could work better as a habit. That’s what I was doing here by adding on to paloma’s original post; offering my relevant experience and suggestions if they are useful. I’m not annoyed but I do see potential for things to work better, improve productivity, engagement and a usable, searchable pool of data, and to clear up some confusing messaging around “unknowns.” My intention was not to complain but to be honest about what energy it takes to do these tasks, and some of the perceived challenges, in the thought that perhaps my experience is also that of others and that we all may benefit from reflecting on it.

Likewise, since I don’t know you (although I am grateful for your creation of iNaturalist!!!) I can’t interpret the tone of your messages but really don’t want to be perceived as a whiner here when that is not my intention at all. Rather, I am an overthinker with good intentions and don’t want to detract but rather add what I can when I can to the conversation. Nothing needs changing and I will gladly continue my IDing and welcoming to the extent that I enjoy it without annoyance.

Please use whatever of my comments are useful and discard the rest.


1 Like

Have you encountered this often? Just curious, I feel like I maybe have come across it once or twice.

I wouldn’t say often, but more than once or twice. At least one very angry person, one person who needed to discuss me and my coarse IDs on the iNaturalist group, some people responding that they are dealing with an expert they know so please butt out, some “Duh”-type comments. As someone whose profile says “Amateur” I’m not surprised that someone whose profile says “Admin” or “Developer” would get less of the negative feedback.


I find the amount of Unknown observations really frustrating. Last winter I went through all of the Unknown observations from Canada and got it down to just a few of stuff I had no idea what to do with (random slime). 90% of the observations were plants. I checked back a couple days ago, and there are now over 100 pages of Unknown observations for Canada again (I think a greater proportion are fungi now, but still mostly plants).

Most Unknown observations are from new users with only a few observations. Probably mostly from school projects, but I think from other people exploring as well. Given the huge backlog of Unknown observations, and the fact that most identifiers filter for specific taxa, these observations are unlikely to be interacted with and the user won’t be drawn in to be more active. I feel like at some point people figure out that adding an ID increases the chances of getting help, if they stick around long enough.

I think there are issues on both the website and the app that contribute to the amount of Unknown observations:

Website: New users not knowing how the identification system works (this also causes a bunch of other issues like agreeing etc.). When you sign up for the website, you aren’t pushed through any walkthrough or tutorial, so people just do whatever makes sense to them.
App: When I reinstall the app (Android), as I make my first observation, it says this under the ID text field: “Name what you saw if you can, but it’s OK if you can’t. Others may help you identify it!” I disagree with this; Unknown observations are less likely to be seen, and 99% of observations can be identified to Animals, Plants, or Fungi. Why not strongly encourage that? Assuming most observations are submitted with the app, I think this is a big problem.


I think it helps if you consider that an observation is a valid thing already, even before we see it. When we add an ID we are adding value, or detail, to that observation. By applying DQAs we are improving and adding value to the observation. There might be some that question the value we add because they see value differently to ourselves. As long as you are doing so in good faith, following the guidelines or at least the spirit of and reply politely, let them rant and complain… It is they who are on the wrong side of the guidelines! For example, to tell you to butt out because they have an expert looking at it is against the guidelines, although I would add that simply advising you of that is a different situation, as they might be thinking they are saving you some time. You are perfectly entitled to make IDs as you see fit, and they can opt out if they want full control of the id…


I was just responding to Tony’s direct question to me when I gave the examples of people’s responses.

I’ve learnt from hard experience to not post coarse level IDs on new ‘unknown’ observations. A polite welcome (assuming it was a new user) and comment instead.

When trawling the ‘unknowns’ I sort them in ascending date order, so oldest first. But this is where becomes annoying - the number of not actually ‘unkown’ observations that populate the ‘unknown’ group.



I agree, a Draft DQA (or separate checkbox) might best get at the use case that seems to cause the most consternation – folks who upload huge batches, want to get them onto the web site first, but then have a chance to use the web tools to complete their intended data set before making them visible for community input. It’s just a workflow thing as I see it. Seems reasonable, and I think still fits under the topic of this Feature Request as an alternate solution.

I do think that any Draft mode should come with an automated, regular (weekly?), single notification, something like “Reminder, you have draft observations pending” and provide the link to view them all in Identify mode.

[Edit to add:] also, date/time posted should be updated to the date taken out draft mode, and once out of draft mode, it can never go back.

Definitely agree here too, and think that should also be part of the solution to this topic. If it involves some kind of automated pop-up reminder, that should only be for non-“draft” observations coming in as “Unknown,” and no more than say 5 times per user, so it doesn’t become a continuing annoyance if someone wants to ignore it. (Or maybe just a “don’t show this again” option.)


How about a ‘group’ designation when uploading obs. These could be radio buttons or a dropdown of the coarsest ‘groups’ - something like: vertebrate, invertebrate, plant, fungus, bacterium, unknown - which would just need a click/tap to select without the need for typing in any letters. The group(s) must be chosen before the obs can be saved (if there is no other ID added), and they can be done in batches like one can with the location, fields and projects.

This would lump the obs into more helpful groups (for the IDers) while removing the need to type anything (which seems to be the major issue mentioned by several people above (especially when in the field on the app)). It would also mean that the coarsest IDs are already on the ob so that users cannot get upset by other users adding them.


For what it’s worth, the Korean citizen science site 네이처링 (Naturing) has a similar system (along with biome categories):



some folks here have mentioned implementing some sort of draft mode.

in one suggested implementation, there would be a DQA flag that would indicate draft. i had started thinking myself that this would be a logical way to do it, but i think there would be items that would never get taken out of draft mode, and that would cause a different kind of consternation.

in another suggested implementation, there would be a more distinct publishing workflow. because of forgetfulness or unfamiliarity with workflows or tech unsavviness, i bet this would cause a lot of observations to get stuck in a draft zombie state, too, though no one but maybe the iNaturalist development team might ever be able to see the problem. so a publishing workflow might create a hidden problem.

i think a compromise solution, which might carry other benefits, is to provide a “give me [x time] before you ID this observation” option. in this workflow, you could add some time, say, up to 1 week, to an observation or group of observations (only) at upload time. after upload, you would not be able to add time, though you could reduce it to 0. these observations would be excluded from Identify until time expired. in a timeline view of observations, there could be a box checked that allows you to still see (but not act on) observations with unexpired time. if the box was checked, the timeline would be sorted based on the upload date. if the box was unchecked, the timeline would be sorted based on the date at expiration of this time. (or something like that.) the extra solo time could provide people something like a draft mode, or it could provide the observer first dibs at trying to ID an observation. (sometimes i want to go ahead and upload interesting stuff for people to see, but i want first dibs at trying to use my ID skills before other jump all over it.) and because the solo time would expire at some point, that would prevent observations from getting stuck in draft mode as a hidden problem.


Ah, rather than a “draft”, an “embargo” (like you can do with GenBank uploads).

1 Like