MCDM Data Normalization

Updated on

Oct 19, 2023

{% docs-informer info title="Important" %}

To better understand the Improvado Marketing Common Data Model and how it works, you can check the Improvado MCDM Overview.

{% docs-informer-end %}

Normalization Algorithm

The mapping mechanism allows you to normalize the data coming from the data sources in a common way if the values are missing.  
MCDM data normalization works by devices, gender, age, and geo data, if any, applied to a specific recipe.
Below in this doc are the sets of values for the specific data fields.

Normalization rules are applied in priority of order:

  1. The empty value (e.g., empty string or NULL) will be transformed to ```Unknown```.
  2. The ```other``` and ```unknown``` values will be transformed to ```Other```.
    E.g., the device_campaign (Facebook) may return ```other``` for the device field.
    The geo_by_ads report type (Facebook) may return the ```unknown``` for the country field.
  3. The value out of range of MCDM predefined range (if it is not empty, ```other``` or ```unknown``` and contains meaningful data) will be transformed to ```not identified```.
Priority Source value Normalized value
  • Empty string
  • ```NULL```
  • ```other```
  • ```unknown```
3rd Out of predefined range ```not identified```

Device Normalized Values


  • Connected TV
  • Mobile
  • Tablet
  • PC
  • Other
  • Unknown
  • Not identified
Platform normalization algorithm


  • iPhone
  • iPad
  • Android Smartphone
  • Android Tablet
  • Other
  • Unknown
  • Not identified
Device normalization algorithm

Operating system

  • iOS
  • Android
  • Desktop
  • Other
  • Unknown
  • Not identified
Operating system normalization algorithm

Age Normalized Values

  • <18
  • 18-24
  • 25-34
  • 35-44
  • 45-54
  • 55-64
  • >65
  • Other
  • Unknown
  • Not identified

Age normalization algorithm

Gender Normalized Values

  • Male
  • Female
  • Other
  • Unknown
  • Not identified

Geo Normalized Values

MCDM models meet the ISO 3166-1 alpha-2 standard developed and maintained by Looker Studio. Learn more here.

All geo values are converted into country and region codes.
We use country_region_mapping, a spreadsheet with all country names/codes and region names/codes specified in a unified format.

Normalization Logic for Geo Data

  1. A normalized country/region value will be added to MCDM if joined with the country_region_mapping spreadsheet values.
  2. MCDM will contain raw data if the join with the spreadsheet fails. 
  3. The empty or unknown value will be included in MCDM without changes.

Geo Data Normalization Example Table

Raw data MCDM Mapping Spreadsheet MCDM Comment
Country Region Country Name Country Code Region Name Region Code Country Region
United States California United States US California US-CA US US-CA Normalized (joined successfully)
US Illinois - - - - US Illinois Not normalized (Unsuccessful join with region mapping)
US - - - - US Not normalized (Unsuccessful join with region mapping)
Denver Pavillion Theater - - - - Denver Pavillion Theater Not normalized (Unsuccessful join)
- - - - Empty raw columns
Unknown Unknown - - - - Unknown Unknown Unknown in raw data

Schema information

Setup guide


No items found.


Troubleshooting guides

Check out troubleshooting guides for
MCDM Data Normalization


Frequently asked questions

No items found.
☶ On this page
Related articles
No items found.
No items found.


Improvado team is always happy to help with any other questions you might have! Send us an email.

Contact your Customer Success Manager or raise a request in Improvado Service Desk.