Hi there,
Background to issue:
I have built two segments;
1. Users that have 31+ visits in the last 30 days and from USA
2. Users that have 1-30 visits in the last 30 days and from USA
Issue:
When in Workspace, I am trying to get the user ID's for each individual segment.
However, when I use the Visit metric and filter to the last 30 days, there are users that have visit counts outside the range of the specified segment.
For example, when I drag in the 31+ Segment, I can see that users who have less than 31 visits are also included. I thought the segment rules would have ensured that these people outside of the given range would be omitted? See screenshot below.
This was the case for both segments in Workspace i.e. The 1-30 Visits segment also returned results of people with over 30 visits in 30 days, which is not correct.
Can someone please advise? I have checked the Segment rules and they are set up correctly. The issue is in Workspace.
Thanks!
Kind regards,
Emmet
Solved! Go to Solution.
@emmetc2018 thanks for testing the segments, let me try to explain why you get higher or lower numbers...
lets start with the easy one: lower numbers. as I already explained, a user might log in not on the first visit, but on the 20th. that means, for visit 1-19 you don't have a UserID. but if the Visitor had a total of more than the lower condition, he will be in the segments since you are looking at the visitor and all his visits. and breaking down by UserID will just show the remaining numbers...
lets get to the other case: higher numbers. I assume that this is a visitor problem, maybe the user deleting the cookies. just an example: a user comes to your site for 40 visits (all with UserID), clears cookies and comes back again doing anothe bunch of visits. for Adobe Analytics this would mean "2 visitors" both matching your segment. and breaking down visits by UserID would show the total bisits of all visitors ... just drag in the visitor metric, I would bet it is above 1....
Knowing those cases I can't think of any segments who would solve your idea - at least it is not top of my mind. the problem is that your criteria is not based on visitors (or containers below), it is a global criteria.
last question: do you only need to see a number of how many UserID had a certain amount of visits or do you nead the specific UserIDs? If it's only the total number, there might be a solution with a calculated metric: "if visits greater that 30 and lower than 50 then 1 else 0". the row total in your table of UserID would be the number of UserIDs minus 1 (for "unspecified")
I can think of several problems:
don't know what the reason is without more information, but would bet for the "dimension"
Views
Replies
Total Likes
urs.boller thanks for your post.
1.
I have screenshotted two segments below which I have used. One segment tries to specifically omit users with less than 31 visits, but still those users show up in the Workspace query (e.g. A user with 10 visits). Screenshots below;
2.
The User ID is when the user logs in.
3.
The date range was 'In the last 30 days'. Top right corner of the screenshot in my previous post.
Thanks,
Emmet
Segment with exclusion included;
Views
Replies
Total Likes
Hi emmetc2018
just a last question begore answering: what is your "UserID" -> Prop or eVar? if eVar, what expiry? and is it possible a user can come to your site without any login (so no UserID is set)?
Views
Replies
Total Likes
I'd recommend using a visit container for both the include and exclude containers. So it'd look something like this:
Visit container where
- Visit number is greater than 30
And
Exclude visit container where
- Visit container is less than or equal to 30
Right now your include container is visitor and the exclude container is hit. visit number is a visit-based dimension, so if you're able to use visit containers, you'll get the most accurate data.
Views
Replies
Total Likes
Gigazelle while your segment might be ok, the problem with the low visit number still exists. And as long as we don't get more information from emmetc2018 we can't say what is ok as segmentstion.
I think there might be a problem with the userID, since it is not set in every visit. If it's a eVar with maximum visit expiration, we would see a picture as shown in the original question: visitors with more than 30 visits, but only in 10 visits a userID is set (that means it only counts the visits where a user has logged in, in all other visits there hasn't been a login)
but this is just "best guess" - would be happy to help with more information...
Views
Replies
Total Likes
Hi Emmet, were you able to get this issue resolved? Let us know if there's anything we can do to help!
Views
Replies
Total Likes
Re: Global Company ID? ursboller;
Yes, people can visit our site without logging in.
There is still an issue, as to get 31-99 visits in a month, people with higher visit numbers or lower vist numbers are included i.e.4 visits or 210 visits for example.
Views
Replies
Total Likes
Gigazelle Can you please see above answer
Views
Replies
Total Likes
I tried using the exclude container function in my segments, see screenshot.
When I run in workspace there is still the issue of people under 100 visits being included?
Views
Replies
Total Likes
Try changing the include container from visitor to visit, and also change the exclude container from hit to visit.
Views
Replies
Total Likes
Views
Replies
Total Likes
Hi emmetc2018
Sorry I haven't written earlier, missed the conversation. But let's try to solve. I think you have two issues, I will explain step by step.
1) Segment definition:
If you want to see visitors who had more than 30 visits in a certain time range, you can not use the dimension "visit number", This is due to the fact that the counter starts with the very first visit regardless the time range you look at. just take an example: a user visited your site 100 times this year but only once last month, he would have a visit with a number above 100 for last month (within your segment), but only 1 visit when you look at the data for last month!
Therefore you need to change the segment als follow:
{(visitor) where (visits) higher or equal to (100)}
this should return all Visitors who had more than 99 visits in the date range you are looking at.
2) workspace data
the second problem occurs by reading the data. even when you apply the segment above to your table, you will see numbers below 100! this is due to the fact that it shows you a single ID and how many visits this visitor had AFTER he catched the ID.
let me give an example: a user came to your side 105 times (visits) the last month (thus matching the segment), but only logged in after the 56 visit. looking at the corresponding user ID would inly show the remainig visits AFTER the user catched the ID, your report would show something about 50 (and not the full 105 as you expected.
I hope those explanations help you solve your question. feel free to ask for more details or other help if needed.
Hi Emmet, are there any follow-up questions you'd like to ask? Were you able to get everything working?
Views
Replies
Total Likes
Thanks for your explanation on this.
Using the above explanation, can we be certain when the segment is run in Workspace using a range (i.e. Between 30-99 visits in the last 30 days) that it is accurate?
For example when the above segment is run, values above and below that range are returned e.g. 120 visits and 10 visits in the last 30 days.
What is the explanation for this in this case?
Visit Number = Explained high count problem (Understood)
Cached ID = Explains low count problem (Understood), but only when I used incorrect segment.
When I apply correct segment, as outlined above in previous post, I still get user IDs with values outside the range specified? See screenshot below.
Can you please advise?
Thanks,
Emmet
Higher Visits than specified
Lower Visits than specified
Views
Replies
Total Likes
@emmetc2018 thanks for testing the segments, let me try to explain why you get higher or lower numbers...
lets start with the easy one: lower numbers. as I already explained, a user might log in not on the first visit, but on the 20th. that means, for visit 1-19 you don't have a UserID. but if the Visitor had a total of more than the lower condition, he will be in the segments since you are looking at the visitor and all his visits. and breaking down by UserID will just show the remaining numbers...
lets get to the other case: higher numbers. I assume that this is a visitor problem, maybe the user deleting the cookies. just an example: a user comes to your site for 40 visits (all with UserID), clears cookies and comes back again doing anothe bunch of visits. for Adobe Analytics this would mean "2 visitors" both matching your segment. and breaking down visits by UserID would show the total bisits of all visitors ... just drag in the visitor metric, I would bet it is above 1....
Knowing those cases I can't think of any segments who would solve your idea - at least it is not top of my mind. the problem is that your criteria is not based on visitors (or containers below), it is a global criteria.
last question: do you only need to see a number of how many UserID had a certain amount of visits or do you nead the specific UserIDs? If it's only the total number, there might be a solution with a calculated metric: "if visits greater that 30 and lower than 50 then 1 else 0". the row total in your table of UserID would be the number of UserIDs minus 1 (for "unspecified")
Apologies for the delay in responding. Thanks for your clarification on this.
To answer your ending question;
"last question: do you only need to see a number of how many UserID had a certain amount of visits or do you nead the specific UserIDs?"
We would like to know both for reporting purposes.
1. The specific User ID's of people who meet the criteria e.g. Above 100 visits or between 30-99 visits last 30 day sor last 60 days etc.
&
2. A count of how many people visited over 30 times last month, or how many visited over 100 times last month/last 2,6,12 months etc.
With that being said, is there any way on Adobe to have a query/segment run that would have exact accurate results for the above requirements, or will there always be this inaccuracy?
Thanks,Emmet
Views
Replies
Total Likes
Hi emmetc2018
if you want the perfect numbers go to the IT and ask for a login log file. Analytics might loose some of your users due to provacy settings on their end. therefore you will most like miss some userIDs in your reporting anyway...
However, here are some ways to get close to the question you are looking for:
List of userID's with certain amount of visits
create a table with dimension "userID". vor the first col use a calc. metric like this:
"if (greater than: visits | static number 30) then 1 else 0"
the table will show you a list of all UserID's with a "1" as result if the userID had more than the desired amount of visits. it will respect the reporting timeframe, so set it to any date range you want.
unfortunatelly the table will also include those userIDs with less visits, but having a 0 (zero) as number. please upvote Allow Filtering report columns/metrics for Limits & Ranges
and unfortunatelly you can create a segment based on the result. please upvote Use Calculated metrics inside segments
total userIDs
having the table above just look at the column total which will show you the total of userID matching the criteria. this will include "unspecified" as well, so don't forget to count -1 before showing the number.
(add a new calc. metric with "column total of above metric - 1")
one last idea: use classification
maybe not the best way, but having the desired userID you could upload a classification on those userIDs (to indicate visit volume) and start creating segments based on this classification. it is not flexible, needs work for setup and maintenance but maybe a way to go....
last remark:
I never used it but maybe you can upload customer data based on the userID? maybe data from login logs? I hope Gigazelle can tell you if it's an option and give you a link (thanks!)
Views
Replies
Total Likes
Views
Likes
Replies
Views
Like
Replies