I have been told that the approval or rejection of road edit can take up to 90 days before they are timed out.
When I did my 72-road edits test I found that all edits got processed (either rejected or approved) within a week. As nobody had reported this before, I was afraid I made some mistake in keeping track of my road edits.
But after @ErmesT and now @MarcoDavoli have confirmed my findings in their replies here and here, I’m confident to announce that the process to approve road edits has for sure been changed and the max processing time has been reduced from 90 days to only 5 days.
Here is my proof:
Since I did the 72-roads test I have continued to make road edits to learn more and focus on how I could up my approval rate of road edits. Between February 26th and March 12th, I made 1434 new road edits following all the tips and tricks mentioned in the new manual for road editor. Of these 353 have been approved. That makes the approval rate about 25%. Which is still too low to my liking.
If we only look at the 353 approved edits I have analyzed how many days it takes for them to get approved. Here are the results:
1/4 of my edits got approved on the same day the road edit was submitted.
40% on the next day
18% on the next day.
After 5 days practically all road edits are resolved.
If you wonder what happened to all the rejected edits, I can refer you to the 72-roads test where I daily revisited every edit to see if they were still pending. After 1 week all pending road edits could no longer be found when opening the road editor. Also, this tabulation of my rejected road edits supports the notion that pending road edits do not stay pending for long.
I speculate that the low approval rate and faster processing time could be linked to how Google Maps had made changes and maybe delegated all the approvals to the AI system and no longer depend on human operators to check our road edits.
If the much lower approval rate is a consequence of the higher processing speed, this is “bad” news.
Maybe older road edits from before the change was made will still be processed within 90 days. But I’m quite confident that if I stop making road edits today the drippeling of approved road edits will end in 5 days.
@MortenCopenhagen good to know. I added roads in my neighborhood last year, and they just stayed in “Pending” for so long that eventually the roads were added to the map the natural way. My road edits never got accepted, and I never got credit for them. This actually caused me to ignore the feature because I felt it wasn’t worth the time.
Although my neighborhood is now complete, I’ll have to keep an eye out for new roads I can add in the future.
Thank you, @MortenCopenhagen for tagging me in your valuable post. I appreciate your in-depth analysis and sharing the results with us promptly. It’s very helpful. I agree with your findings that there is lot of improvement in the approval process recently. This surely encourage our TRAC team to do more.
Thank you as always for your thoughtful analysis. Before the speed up in approvals started, I was tracking how many edits were approved in the same month they were submitted. After considering your posts, I added tracking for <=5 days, 6-14 days, and >=15 days. So far this year it appears about 47% of my edits are being approved in <= 5days. Here is the current summary of my road editing this year.
I have spent more time analyzing than editing the last few days. I have a propensity to submit a shape edit with a missing road edit. In my reviewing I discovered that the emails (both reported and published) do not capture the shape edit, I still have about 90 pending dual edits submitted “pending (2)” but I also found about 25 shape edits that had been approved. At least now I have them all tracked and it won’t be so difficult going forward.
One of the elements I have been concerned with is tracking rejected (not approved) edits. So far I have 24 of them. I have pasted the notice in the edits section into a spreadsheet. To insure that I am not double counting any, I have also pasted the location coordinates (15°53’33.0"N 97°04’45.9"W) into the same row and set up conditional formatting to notify me if I add a rejection a second time. I’d love to know how others are tracking rejections and other stats that have to accounted manually.
Spending more time on tracking and analysing our road edits is a clear sign that the system is broken. Google Maps has all the data - they ought to make it easy for us to track and learn from our road edits.
I was wondering how you identify your rejected edits. I expect you check the contribution list for edits. In my 72-road edits test I went the extra mile to visit the placed in the Road Editor to see if the orange dotted line is still there or it had disappeared (= rejected).
Would you be able to make this test for some or all your still pending 2024 road edits?
Next, I wanted to study your data to maybe figure out when the system was changed to processing roads more quickly.
I’m unsure why the March 93 is different from 85. Is that roads submitted before March?
If so having 84 out of 85 approved within 5 days confirms my findings. But then I do not understand where the 15 and 18 comes from.
If I just look at column 3 and 4 in the bottom table it looks like the shorter processing time has been in place since the beginning of 2024.
I was wondering how you identify your rejected edits. I expect you check the contribution list for edits. In my 72-road edits test I went the extra mile to visit the placed in the Road Editor to see if the orange dotted line is still there or it had disappeared (= rejected).
Would you be able to make this test for some or all your still pending 2024 road edits?
I search my edits for “not accepted (1)” and “not accepted (2)”
I will go back and look for the orange dotted line and get back to you.
I’m unsure why the March 93 is different from 85. Is that roads submitted before March?
Yes, the difference is due to edits from January and February being accepted in March. I am still trying to understand the numbers in the 6-14 and =>15 columns.
As to using plus code vs GPS; I never thought of it and appreciate the suggestion. The plus code has the advantage of giving the location. I will switch to this going forward.
As you suggested I did go in to some not accepted edits and open them in map editing mode. I found the results confusing. In some areas there is no orange line. IS a small dotted line an indication of a road that was not accepted? I admit I had trouble keeping track of what my screen shots represented. I will post a screen recording and some photos for you to comment on.
I believe that the approval time depends on many factors, and that we don’t know them all, @MortenCopenhagen
Your test is interesting, and I’m trying to replicate it and above all to understand if it is scalable. For this reason, on March 11 I tried to do a test that was an order of magnitude larger: 619 streets in one day. Despite the excellent approval rate, after 9 days only 60.58% of the edits have been evaluated, while 39.42% (244 edits) are still pending.
So on March 16th I did a test on a scale 10 times smaller: 59 edits. This time the results are much more similar to yours in terms of approval times. In fact, 78% of the edits have already been approved, in just three days.
On March 17th I launched an intermediate test (267 roads) but it is too early to make an evaluation.
However the overall daily approval data shows a tendency to have a high percentage of approvals in the first 7 days, and then dilute for another couple of weeks. However, the overall daily approval data shows a tendency to have a high percentage of approvals in the first 7 days, and then dilute for another couple of weeks. Rare cases are approved near the 90 day mark
On your first screenshot under the video you are showing both of the two lines:
Orange dots (with noting in between the dots)
and
The Orange rectangles mixed with white ditto.
Here is my interpretation (I could be wrong):
Orange dots indicate new pending roads we have added.
The orange+white lines indicate a circular road that we have added a new road to or that we reshaped. I have no idea why circular roads are treated differently.
When an orange or orange+white line no longer visible I interpret this as the road edit was rejected. This is what you showed in your last screenshot from
On March 19th I received one approved notification about a road submitted on March 10th. That makes the delay 9 days. But the picture is still clear that approximately 2 out of 3 road edits get approved within the first 2 days. Keep in mind that this diagram is based on approved edits only.
This is based on your roads submitted on March 11th and March 16th. I conclude that 68% (2 out of 3) of your approved roads are approved within the first 4 days. This can be compared to my 2 out of 3 in the first 2 days. See my updated graph in the reply just above this reply.
It seems your roads are approved a bit slower than mine and it keeps dripping for longer. But we are in the same ballpark.
Things might change when we have more data, but I doubt the conclusions will change much.
Now to the approval rates.
OH BOY! it would be difficult to say that we are anywhere close to being comparable.
Here are my latest data:
3/4/2024
32
66
48%
3/5/2024
42
181
23%
3/6/2024
76
249
31%
3/7/2024
48
158
30%
3/8/2024
28
111
25%
3/9/2024
28
109
26%
3/10/2024
54
175
31%
3/11/2024
31
101
31%
3/14/2024
55
213
26%
3/15/2024
57
191
30%
3/16/2024
18
112
16%
3/17/2024
33
154
21%
3/18/2024
13
74
18%
3/19/2024
4
18
22%
Columns:
1: Date of submission
2: # approved edits
3: # submitted edits
4: Approval rate
As you can see all my efforts to learn and improve my approval rate only increased from 24% to maybe 30%! So 7 out of 10 of my read edits are rejected.
As I already alerted to here, this is far below what I hoped for.
I would love to get to the bottom of why I’m unable to push the approval rate further up. But before we start the guessing game and hopefully come up with some relevant tests we can do, I think we need to reflect more and come up with some possible overall explanations.
There is still some time to wait for the final numbers to come in, but I will not reach anything comparable to what you guys @ErmesT@Rednewt74 and @MarcoDavoli consistently report on approval percentages.
Yeah, @MortenCopenhagen , in some way the data are comparable. Regarding your last table, do you mean that ALL the edits were already processed, or you still have pending edits for some day?
What I can see from my data is that the amount of edits submitted by a user can makes a difference in the approval time. I will have to collect more data to have a consistency, but at the moment what I can see is that more edits/day = longer process time
Regarding the 90 days as a time limit for the approval of a pending edit, the time frame is the same, even if there is a big “empty area” between the approved edits and the ones rejected for overtime, even if in some rare case I have seen an edit approved on the very last minute
Regarding your approval rate, it can only increase. Don’t forget that I started to add roads from the very first beginning of the program, and I had the possibility to see all the changes in the approval process and to adapt my way to edit to the improvement of the AI. But acting by experience can help a lot.
In two separate tests I have found that there is no need to search for rejected edits as after 7 days no edits are pending. You know that my edits contribution list is full of other stuff. And Maps for some reason decided not to send email notifications when a road edit is not approved. So I systematically surveyed all my edits in the 72-road edits test by daily visiting the places in the road editor to see if the orange dotted lines were still there = still pending or gone = rejected. All 72 were gone after 7 days. Also my sporadic checks in the past 2-3 weeks confirm that pending roads do not linger. And I did a huge test harvesting the edits contribution list notifications. Again this proved that after a week all are either approved or rejected . So it is a waste if time trying to keep close tabs of the rejections. Let me know if I should share the link to where these data are shared
So my method to calculate rejections is submitted edits minus approved edits after 7 days. Very easy.
Regarding the meager 30 percent approvals you are correct that it can only go up. Haha. But the demotivation from following all your wonderful tips and tricks and not getting results is preventing me from being excited.
