As a Marketplace partner, you’ll need to regularly upload your data into Adobe Audience Manager (AAM). This is a guide to help you troubleshoot those times when you try to upload a data file and yet no data shows up in your traits (or less data comes in than you expected).
Wait until the data is finished processing
The first thing to do is wait a little while—if you uploaded the file recently, it may just be that your file hasn’t finished processing. From the documentation:
File processing is executed twice per day. These procedures ingest data and prepare it for delivery.
File delivery times vary because they are affected by the total amount of customer data that needs to be processed. You should expect a maximum latency of 48 hours between the moment the file is uploaded in Audience Manager and until the data is available for reporting and activation.
Check out the onboarding reports
If you’ve waited more than 48 hours and the data still isn’t there, your first stop should be the Onboarding Status reports:
Select the data source that you’re trying to upload data to, select a date range, and click okay. If there's a problem with the files you've been sending in, you'll get something that looks like this:
Those red bars show all of the data that failed to upload. Scroll down and this report will tell you why the records failed. There are five reasons that AAM will give you:
- Format Error
- Invalid AAM ID
- Invalid Device ID
- No Trait Realized
- No Matching AAM ID
Let’s look at each error:
Format Error
You get these errors when the records you uploaded did not match the syntax or formatting requirements. See Inbound Data File Contents: Syntax, Invalid Characters, Variables, and Examples for information on how to format your data.
Invalid Device ID
If you uploaded a file with mobile IDs (Apple IDFA or Android GAID), AAM checks to see if the IDs conform to the correct format (the right number of digits and hyphen placement, the capitalization of letters etc.). Here are the standards that it's checking against:
Device Advertising ID - IDFA - iOS devices IDFA IDs are mobile device identifiers, provided by the device manufacturer. These IDs represent devices that run the iOS operating system. |
The format strictly consists of 32 uppercase hexadecimal digits, displayed in five groups and separated by hyphens, in the form 8-4-4-4-12, for a total of 36 characters. Example: AEBE52E7-03EE-455A-B3C4-E57283966239 |
Device Advertising ID - GAID - Android devices GAID IDs are mobile device identifiers, provided by the device manufacturer. These IDs represent devices that run the Android operating system. |
The format strictly consists of 32 lowercase hexadecimal digits, displayed in five groups and separated by hyphens, in the form 8-4-4-4-12, for a total of 36 characters. Example: e4fe9bde-caa0-47b6-908d-ffba3fa184f2 |
AAM will reject IDs that don't conform to those standards and those rejected IDs (Note, your filenames will contain 20914 as the DPID for GAIDs or 20915 as the DPID for IDFAs. See this documentation information on how filenames should be constructed to tell AAM what ID type you've uploaded in your file.)
Invalid AAM ID
You'll only see this error if you are uploading AAM IDs (not your own cookie IDs). It indicates the IDs:
- Did not match the expected 38-digit format.
- Contain alphabetical characters. IDs should be numbers only.
No Trait Realized
You’ll see this error when Audience Manager cannot match the trait IDs in your file to a trait in AAM. There are a few reasons this might happen:
- The trait taxonomy and associated trait rules have not yet been created in AAM. You will need to create the taxonomy and rules and re-upload the data. See this documentation for details on how to create the taxonomy.
- The trait taxonomy has been created but the trait rules are incorrect. See this documentation for details on what the rules ought to look like given the format of your data. You will need to change the rule setting and re-upload the data.
- You created the traits and the rules are correct, but you didn’t select “Onboarded” when creating the trait. You can confirm the trait type by going to Audience Data > Traits and checking in the trait type column:
No Matching AAM ID
You’ll see this error when Audience Manager cannot match the IDs in the file to an existing ID in AAM. This can happen because the Data Source configuration is wrong, the filename is incorrect, or there are problems with the IDs themselves in the file. It can also happen when Audience Manager has not yet performed an ID sync.
Data Source configuration problem:
If the IDs inside the file you are sending into AAM do not match the configuration that you’ve set up in AAM, AAM will not be able to match the IDs. For example, if you tell AAM that you will be sending in AAM user IDs but you actually send in your own user IDs you will see this error.
You'll need to go back into your data source and select the correct ID.
Incorrect filename:
When you send a file into AAM, there’s a specific naming scheme you need to use to tell AAM which IDs you are using and what data source to send the data to. Generally, your file name will contain both a DPID and a DPID_TARGET_DATA_OWNER. See this documentation on file naming for more details.
For cookie IDs, this DPID usually refers to a user ID data source that you set up before sending in your data. (See this documentation on setting up the user ID data source for more details.) The configuration of this data source tells AAM what ID type you are using in the files you send in. So you should check the configuration of this data source to make sure that the settings match the IDs you are sending in.
File formatting problem:
If your configuration matches the ID type your are sending in, check to make sure there’s no formatting issue with your file. Sometimes partners send in data files where the ID and the trait ID have been concatenated accidentally or the incorrect delimiter is used.
Onboarding report—error sampling
Even with the error reporting, it can sometimes be difficult to track down the erroneous data in your files. AAM has a feature called error sampling that can show you specific lines from your files that triggered errors. You'll need to activate the error sampling feature manually. Go to this link to learn more about error sampling and how to activate it.
What if no data shows up in the onboarding report?
If you don’t see any data at all in the onboarding report, then the file either didn’t make it into the S3 bucket, it was sent to the wrong data source, or it didn’t pass the initial checks.
Log into your S3 bucket and look for the file to make sure it was uploaded properly. If you’ve waited a bit before checking, the file should be there and should be appended with either .processed or .invalid. If it’s appended with .invalid, that means that it didn’t pass AAM’s initial processing checks:
- The file was too big. Compressed files cannot be larger than 3 GB and uncompressed files cannot be larger than 5 GB. You’ll need to split your data into multiple files if they’re larger than either of those limits.
- The filename didn’t conform to the file naming conventions. Check your file name to make sure it’s not badly formatted.
- The file itself is unreadable or malformed in some way.
If the file is there and is appended with .processed and yet you still can’t see any data in the onboarding report, make sure that the file was sent to the correct data source. You can check your data source ID by going to Audience Data > Data Sources:
In the example above, you’d expect the filename to look something like this:
ftp_dpm_2345_30651_1527632600.sync.gz
Where the 2345
identifies the table that holds the ID sync, and 30651
identifies the data source the data is going to. If you didn't see the 30651 data source ID in the filename, then you could assume that the data was being uploaded to the wrong place. You would need to correct the filename and re-upload the data.
I’ve checked everything and I still can’t find the problem
If none of these troubleshooting steps work, then reach out to support.