Skip to main content
Gainsight Inc.

Tips For Working With MDA And S3

Fundamental information and workings of the Matrix Data Architecture (MDA) feature within Gainsight can be found here.

This article contains helpful tips and tricks within the MDA that aren't necessarily intuitive, but will end up saving you time.

Creating your Objects

Whether you're creating your MDA objects manually or automatically, make sure the name of your MDA objects are an exact match to the corresponding headers in your CSV file.

Helpful tips

  • Check for extra spaces both before and after your CSV headings
  • Double-check for any possible misspellings or odd characters
  • Ensure all of your columns in your CSV file have a title
  • Make sure each of your columns corresponds with one of the following data type categories when being mapped to MDA:
    • String
    • Boolean
    • Date
    • DateTime
    • Number
    • SFDC ID
    • Email
  • Once you've added all of your fields in the MDA, make sure you click SAVE! This is a commonly missed step and can be a source of frustration for Admins who then have to re-enter everything.


S3 Connector Tips and Tricks

For detailed information, please review the S3 documentation.

Helpful tips

  • Email to enable your S3 connector if you are unable to see your S3 connector information.
  • To upload files into the S3 Bucket or for troubleshooting, you can use your preferred S3 access tool. For the purpose of this document, we will focus on a very common tool called Cyberduck, which you can download for free here:
  • Once in Cyberduck, you will need to select S3 (Amazon Simple Storage Service). From here, you will need to fill in some data, including Username, Password and Path:
  • You will find the necessary data in Administration > Connectors > S3 Connector. From there, click View S3 Config in the top right corner to find the necessary information to load into Cyberduck.


  • Your screen will then look like below after clicking “View S3 Config”


  • In the Cyberduck tool, you'll map the information from Gainsight to Cyberduck.

Note: When using Cyberduck, you must copy the Bucket Access Path after s3:// (start with the text that reads gsext)


  • Once your CSV file is ready to upload, drag and drop it in the input directory of the S3 bucket.


  • Put your yet-to-be-uploaded files in input. (this is a step that is often forgotten)
  • Once the S3 Ingest Job runs from Administration -> Connectors -> S3 a few scenarios can happen with the file after ingestion:

    • Complete Pass or File: File will be moved from “input” folder or “archived” appended with a datetime stamp in S3 Bucket

    • Partial Fail: Success rows will be archived, while failed rows will go to “error” folder with appended datetime stamp in S3 bucket.

s3 data ingest setup.png

For Cyberduck version above 5.2.2, path parameter will not be available for certain users. However, you can still access the S3 Bucket by “Server” input with <bucketname>

amazon s3.png

Limitations with the MDA

  • Multiple files with different column formats require creation of corresponding projects for the desired Matrix Data-Object.
  • We recommend that the CSV file size not exceed 200MB.

Reporting on the MDA from the Rules Engine

  • Make sure your Account Lookup is correct
  • Do an upsert to make sure you can include in identifiers‚ for unique IDs
  • Click the gear icon to add the fields you need
  • Make sure you have the GS-Special license in order to have the correct permission to reference these MDA objects
  • Hit the back button on your browser rather than the back button in the Gainsight app. The back button in the Gainsight app will take you back to the rules list whereas

Creating an MDA join

  • Decide what needs to be put in your join table: typically the Account ID, Name, CSM information, other attributes, etc.
  • There must be at least one field that is an external identifier to your account data in Gainsight that will allow you to tie together your external data and your Gainsight data
  • In the object listing screen, the object existing in the Redshift data store will have the "Joins allowed" property set to true.
  • The lookup relation between two objects can be defined only on "String" data type fields.
  • Currently, we support creation of lookup relation only on MDA objects for which joins are allowed.

MDA Storage Limits

General data storage limitations:

  • 30 GB limit of total MDA storage

  • 10 GB or 100 million records (either) is the limit per MDA object (Subject area)
1 million events per day

If you are concerned about exceeding any of these limits, please contact

Data purges: This can occur if the maximum data storage is reached, but it occurs after a Gainsight team member informs you that you have reached the limit, and Gainsight team member discusses with you on how the purge could occur.


  • Was this article helpful?