(Low Traffic) - Seriously, what is this im simple terms?

[New2Adobe3]

I've read different articles a number of times and just absolutely fail to understand how and why this populates.

Low-traffic value in Adobe Analytics | Adobe Analytics

It says:

Adobe Analytics uses two thresholds to determine which unique values are displayed in reports each month: A low threshold and a high threshold. These thresholds can be adjusted by Adobe from time to time. The current threshold limits are:

Low threshold: 2,000,000 unique values during the month.
High threshold: 2,100,000 unique values during the month.
Reporting is not affected if a variable does not reach the low threshold in a given month.

When a variable reaches the low threshold, data begins to be bucketed under a dimension item labeled Low-Traffic. Each value beyond this threshold goes through the following logic:

If a value is already seen in reports, add to that value as usual.
If a value is not yet seen in reports, initially add it to the Low-Traffic dimension item

•  

Now I have a table showing visitors to a page, and I breakdown the monthly visits by my customer ID evar:

Month	Unique Visitors
Jan 2025	5,152
CUST ID
1. (Low Traffic)	95
2. cust_id_1234567	1
3. cust_id_7645213	1
4. cust_id_abcdef	1
Page: 1/99 > Rows 1-5 of 500

Why are some cust_ID visits showing, and others are not? All of these cust IDs only have a visit count of 1. I just don't get how the above explanation links or explains this artefact in the data.

Is there no way to see which customer IDs fall in the (Low Traffic) segment? 

[EurosIMS]

5 million people visit your site in a month. Each of them is assigned a random, unique number. The report will show the first 2million values. After that, each one gets added to the (low traffic) bucket. At months end, the report could show 5million unique values, but instead it shows 2million, with 3 million in the (low traffic) bucket. It goes by date seen from memory, so if you checked the report one week into the month, there wouldn't be any (low traffic) as the threshold wouldn't have been reached yet. Once the threshold is reached, every new item seen goes to (low traffic). Where it gets a bit more complicated is on additional values. So if a value comes in, & by chance it happens to be the same as an already seen value, that will increment to 2. As that's now more than every other value, it gets moved from the (low traffic) bucket into 'normal' reporting.

You can extract all of the values from a dimension, including the (low traffic) ones, via a Data Warehouse extract (possibly via data feeds or other methods too, but definitely using DW). You can also ask client care to expand a dimensions parameters to show more values (or at least you could 7 or 8 years ago), but there was a cost associated. If you have a large site with many, many, URLs for example, where some URLs will get a rare value added, then that’s another report that could very quickly hit the (low traffic) limit.

[Jennifer_Dungan]

This ^

The "low traffic" threshold is to help improve performance..  since most uses in Adobe are about seeing what is performing the most (and not the least), the lower traffic in a dimension is just bucketed. Adobe uses smart logic to shift around what is in low traffic... so if you hit the "unique value threshold" 3 weeks into the reporting period, then a new value spiked to high traffic, it would be included in your report, and another value (which previously showed) would then be moved to the low traffic bucket.

In the case of a customer id, where there is one value per user, there is no real variations to speak of, then new traffic as of hitting the threshold will just automatically be bucketed.

Any values that are part of "low traffic" also will not be found by segment definitions...

The current threshold limit is 2 million (it was updated a few years ago, the previous limit was 50K)

And yes, you can make a request to increase the bucket size of specific dimensions, I've done it in the past... a few years ago I had a number of dimensions increased to 2 M from 30K.... as long as your request is reasonable (as in you don't ask for 50 B), you shouldn't be charged.

But also as @EurosIMS said, you can get this information via Data Warehouse and Raw Data Feeds. I believe that the API would have the same limitations since Workspaces/Projects use the API to build the report.. so it makes sense that they would have the same constraints.

[Jennifer_Dungan]

Oh, I should add, I've seen this before, where the total seems to be low (like you showing approximately 5K), but still having data bucketed under low traffic, despite the limit being 2 million... I can't explain that one....

[Gigazelle]

Great response - just wanted to elaborate and clarify a couple things:

• Values require hundreds to thousands of instances before they can "escape" Low-Traffic. It isn't the top 2 million values included in reporting, but rather, it's the first 2 million values plus whatever has had enough instances to be prominent enough to be its own line item.
• Both Data Warehouse and data feeds do not have unique limits.
• You can request a specific dimension has its unique limit raised, but there might be a reporting performance impact and there's also a possibility that the request can't be accommodated (either due to too many variables having the limit increased, the requested limit is too high, or something else).

[Gigazelle]

Worth noting that I gave this article a complete refresh this last week. Please read the revised content over and let me know if there's anything that the article does not adequately cover!

https://experienceleague.adobe.com/en/docs/analytics/technotes/low-traffic