Join our Community in its new Home - The Datorama Trailblazer Community Group!

It's been an amazing 3 years coming together in this forum to collaborate, innovate, support, and inspire each other about our shared usage of Datorama. While this is not quite a goodbye, we are excited to announce that we are getting a fresh start in our new home within the Salesforce Trailblazer Community. We have a ton of fun new content planned and you may even see the revival of some of our most popular posts from the past few years.

We’ll be keeping this group around for a bit for you to peruse, but as of November 15, we will no longer be allowing new posts or comments. Be sure to join our new group at to keep the conversation going.

We can’t wait to see you there!

Data_Governance_Error_Identification_And_Correction - FinalVersion

Nidene_PillayNidene_Pillay LondonSYS_ADMIN IMG42
edited August 2021 in Tips & Tricks
Are you making the most of Datorama for your Data Governance?

We know that no client’s data will be perfect - there are countless ways for data inputs to be incorrect or otherwise undesired.
A strong client adherence to naming conventions and trafficking best practices as well as Datorama’s productized approach to error identification and correction can help you along the way to address some of the issues.

Disclaimer : To learn more about the importance of naming conventions, the benefits of it or how to build the perfect naming convention using a tool built by Datorama, please leave a comment down below and we will reach out to you!

In this post, we will cover the error identification and correction aspects taking into consideration that you already have a naming convention in place.

Error Identification

If you are already using a naming convention, you might ask how can Datorama help you identify and correct placements, campaigns, etc. with erroneous names. There are a number of features that can help accomplish this task.

1. Patterns + Harmonization Center
The patterns tool will help parse naming conventions based on rules established by the customer. Here is a video that can help you with the setup!

2. Activation Center
If you are not keen on using patterns, the activation center can be used to send alerts to users when invalid names/values are loaded into the platform. See the process st. below for how to implement such use cases:

Activation Center- “Default” or “Other” Use Case

It is worth noting that this functionality extends to measurements as well. Activation Center is a great way to send notifications for anomalous data.

3. Reports
In the Activation Center use case above, the platform is sending alerts based on incorrect/undesired dimensional values. You can also use the reports module to send files to individuals on a recurring basis. All that’s required is to filter on incorrect values and add a measurement and relevant dimensions to the report.

4. QA Dashboard
In the Activation Center use case above, the platform is sending alerts based on incorrect/undesired dimensional values. You can also use the dashbaords to have a “live” look at all incorrect values. All that’s required is to filter a dashboard page on incorrect values and add widgets to display values that require correction. This page can be shared to users with login accounts or via share/embed link if users to not have access to a Datorama login.

Error Correction

Now that you are more comfortable with identifying errors in naming conventions, you’ll want guidance on the best way to correct these values. It is important to note that collaboration from non-analytics teams is crucial to optimizing the correction process. Below are the best ways to correct incorrectly named placements, campaigns, etc.

Option 1: Correct at the Source

This is the preferred option and should always be Datorama’s first recommendation. This includes changes in your ad servers and billing systems. Most Datorama setups are designed to accommodate changes in naming conventions by virtue of the “Update Entities’ Attributes” setting. This option is not viable if media is done running (and subsequently cannot be updated in system of origin) or if volume of necessary changes exceeds media team’s bandwidth.

Option 2: Correct in Data Stream Mapping

Another option is to use mapping rules to make changes. Let’s say Placement ID 123456 below is misnamed:

The correct name is actually Media Buy 99_728x90. An admin can write an “if” statement to correct this name directly in the source data stream mapping for the Media Buy Name dimension:

After processing the data with this mapping change, the Media Buy Name is now corrected:

Option 3: VLOOKUP to Comprehensive Naming Correction Document

This solution involves creating a running list of all placements that require name changes. The list should include the Placement ID and its corrected name (original/incorrect name can also be included for reference but is not required.) The ID value is helpful as IDs are generally less error prone than names when used in a VLOOKUP.

Using the Placement Names from the previous scenario, let’s say an important value was left off the front of each name:

Each of the placements above should have the text “Corrected Name_” appended to the front. A table is created that includes Placement ID and corrected name. Please note- this file should always be comprehensive. In other words, if 1- new placements need correcting after the file below is updated, the next uploaded file should have 14 total rows. This will make QA significantly easier as your list grows.

You can load the file into a LiteConnect or TotalConnect (with generic bucket) depending on preference. In this demo I’ll use a generic bucket. I like using a generic bucket because I can view values in dimension explorer later if there are QA questions.

Mapping looks like this (Placement ID to main key, corrected name to attribute value):

From here, navigate to the relevant stream where placement names are being loaded. Insert a formula like the one below in the mapping for Media Buy Name. The formula below searches for the media buy key in the generic bucket stream created in the previous step via VLOOKUP. It stores the outcome in a variable called “namelookup.” If namelookup is empty, this means the ID was not present in the lookup file and subsequently does not need to be corrected. If namelookup is not empty, you should return the result of the VLOOKUP (corrected name from lookup file.)

After running the data stream, you’ll see names corrected:


Happy naming and correcting your data! 
Let us know if we can be of any help :smile:

Content put together by Success Architects:
- Nidene Pillay
- Matthew Dykeman
- Michael Barilli 

Sign In or Register to comment.