Add organizational taxonomy

Introduction

AWS customers use various methods to allocate costs of AWS resources, ranging from a single level of business units to complex, multi-dimensional organizational taxonomy.

This document presents the architectural framework for resource ownership attribution and cost allocation in AWS environment, including automated integration with Cloud Intelligence Dashboards.

The following video introduces the concept of taxonomy in the context of AWS resource management. It will guide you through the main concepts using the example of Unicorn Rental Corporation.

Architectural Foundations

Resource allocation in AWS typically follows two primary architectural patterns:

Cloud Intelligence Dashboards provide a comprehensive way to integrate organizational taxonomy into your dashboards, enabling precise filtering and cost allocation tracking. You can define a set of taxonomy fields to be used as Filters and GroupBy dimensions within the CID dashboards. Additionally, you can configure Row-Level Security (RLS) to restrict access based on one of these taxonomy fields, ensuring that users see only the data relevant to their scope of responsibility.

Components for building Organizational Taxonomy

It is essential to establish reporting that accurately reflects the current structure of your organization while also accounting for potential future changes. Once the organizational taxonomies are clear, the next step is to map them to the appropriate technical data.

Additional recommendations:

Account Level Cost Allocation

Using Account Tags as Cost Allocation Tags in CUR2

Since December 2025, AWS added support of Account Tags in AWS Organizations to be enabled as Cost Allocation Tags.

The main benefit of Account Tags as Cost Allocation Tags is simplicity: all required information comes directly from CUR2, eliminating the need for additional data sources or collection mechanisms.

Implementation Steps:

When to Use CID Data Collection Method for organizational data Instead:

While Account Tags as Cost Allocation Tags in CUR2 are recommended for simple account level taxonomy use cases, you should use the Advanced account map method if:

Using Account Map View

To implement account level mapping, CID uses a special View in Amazon Athena called account_map. This Athena view is specifically designed to be customizable by customers. It can be generated using cid-cmd map (recommended) or created manually using any of the following data sources:

The following sections describe different ways to use the account_map view to build dynamic or static account mapping to organizational taxonomy dimensions.

Account map based on AWS Organizations data (Recommended)

The cid-cmd map command provides an interactive workflow for creating enriched account_map views with custom taxonomy dimensions. It supports multiple data sources and can be used with or without an organization_data table in Athena.

The cid-cmd map command automatically generates an account_map Athena view with your organizational taxonomy columns. It supports three data source modes:

Prerequisites:

Basic usage:

# Interactive mode — prompts for data source selection
cid-cmd map

# (Optional) With a CSV file for additional static mapping taxonomy
cid-cmd map --file accounts_mapping.csv

Available taxonomy dimension sources (organization_data mode):

Example workflow:

$ cid-cmd map --file accounts_mapping.csv

? Select account data source for taxonomy:
❯ organization_data table (collected with CID Data Collection)
  CSV file only
  Both (organization_data table + CSV file for additional taxonomy)

? Select data sources for taxonomy dimensions:
  ☑ Tags from source table
  ☑ OU hierarchy level
  ☑ Additional file (CSV)
  ☐ Split account name column

The command will guide you through selecting specific tags, OU levels, and file columns to include as taxonomy dimensions, then preview the resulting SQL before creating the view.

CREATE OR REPLACE VIEW "cid_cur"."account_map" AS
SELECT
    -- Mandatory
      id AS account_id
    , name AS account_name
    , managementaccountid AS parent_account_id

    -- Friendly name for management account(s)
    , CASE managementaccountid
        WHEN '111111111111' THEN 'My Management Org'
        WHEN '222222222222' THEN 'My Test Org'
        ELSE managementaccountid
    END AS parent_account_name

    -- OU hierarchy levels
    , TRY(hierarchy[1].name) AS o_u_level1
    , TRY(hierarchy[2].name) AS business_unit

    -- Hierarchical Tags (set on OU level, override on lower OU or Account level)
    , element_at(filter(hierarchytags, x - x.key = 'environment'), 1).value AS environment
    , element_at(filter(hierarchytags, x - x.key = 'cost_center'), 1).value AS cost_center

    -- Part of Account Name (when naming convention allows)
    , TRY(split_part(name, '-', 1)) AS account_prefix

    -- Additional columns from CSV file (joined by account_id)
    , file.team AS team
FROM
    "optimization_data"."organization_data" org
LEFT JOIN "cid_cur"."account_map_file_source" file ON org.id = file.account_id

You can also create a similar view manually if you want to tailor it to your specific needs not supported by cid-cmd map capabilities. Please note that you do not need to use all the fields from this example — adjust it to your specific business and organizational requirements.

Static Account Map from CSV

As an alternative to the dynamic account map based on AWS Organizations data, you can create a static account map from a CSV file. This approach is useful when you have custom mapping maintained in a CMDB or spreadsheet and you want to bring it to account map.

Using cid-cmd map with CSV file:

cid-cmd map --file my_accounts.csv
# Select "CSV file only" when prompted

The CSV file should contain at minimum an account_id column. All other columns can be selected as taxonomy dimensions during the interactive workflow.

CREATE OR REPLACE VIEW account_map AS
SELECT *
FROM
  (
 VALUES
     ROW ('111111111111', 'Account1', 'Company 1', 'Business Unit 11')
   , ROW ('222222222222', 'Account2', 'Company 1', 'Business Unit 11')
   , ROW ('333333333333', 'Account3', 'Company 2', 'Business Unit 21')
   , ROW ('444444444444', 'Account4', 'Company 2', 'Business Unit 22')
)  ignored_table_name (
    account_id,   --mandatory
    account_name, --mandatory
    company,      --custom
    business_unit --custom
)

Resource Level Cost Allocation (Tags and Cost Categories)

Resource-level cost allocation uses tags and cost categories embedded in the Cost and Usage Report (CUR) to attribute costs to specific resources, teams, or applications. The cid-cmd tool automatically discovers available tags and allows you to select which ones to include in your dashboards.

Cost Allocation Tags

Cost Allocation Tags are the most common resource-level taxonomy source. They come from the resource_tags column in CUR and include both user-defined tags and AWS-generated tags.

IAM Principal Tags, IAM Principal ARN, and User Attributes

CUR2 introduces additional tag types in the tags MAP column that enable attribution based on who performed an action rather than just what resource was used:

Prerequisites:

Cost Categories

AWS Cost Categories allow you to define rules that categorize costs into meaningful groups. They appear in CUR as the cost_category column and are automatically discovered by cid-cmd.

How cid-cmd discovers tags and cost categories

When you run cid-cmd update --force --recursive, the tool scans the CUR2 tags column for each prefix separately:

Scanning resource_tags in cur2_data.
 resource_tags   | Distinct Values (Top 20)
 Environment     | 3
 Project         | 18
 Owner           | 42

Scanning cost_category in cur2_data.
 cost_category   | Distinct Values (Top 20)
 Team            | 6
 Department      | 4

Scanning Account Tags (tags column) in cur2_data.
 Account Tags    | Distinct Values (Top 20)
 BusinessUnit    | 5
 Environment     | 3

Scanning User Attributes (tags column) in cur2_data.
 User Attributes         | Distinct Values (Top 20)
 PipelineName            | 12
 DeveloperAlias          | 45

Scanning IAM Principal Tags (tags column) in cur2_data.
 IAM Principal Tags      | Distinct Values (Top 20)
 Team                    | 8
 ServiceName             | 15

Scanning line_item_iam_principal in cur2_data.
 line_item_iam_principal | 234 distinct values

You can then select which of these to include as taxonomy dimensions in your dashboards.

Performance Considerations

Please note that introducing tags with many unique values can significantly expand the dataset. For example, using tag_a with 10 unique values and tag_b with 100 independent values could result in up to a 1,000-fold increase in dataset size due to combinatorial growth. Avoid using high-cardinality tags, such as Name, especially when working with large datasets based on the Cost and Usage Report (CUR).

If your Quick datasets takes too much time refreshing after configuring the tags, please remove all tags (cid-cmd update --force --recursive -y) and try re-adding them one by one.

Adding Taxonomy to the Dashboards

The CID framework allows you to seamlessly add taxonomy fields to the dashboards including attributes from the Account Map as well as resource-level tags.

Once configured, taxonomy fields are added as Filter Controls across all sheets in the respective dashboard. They can also be used in Calculated Fields and Parameter Controls to support "Group By" aggregations.

To modify the taxonomy fields or update the set of Cost Allocation Tags, run the following commands:

After update Amazon Quick datasets will be refreshed automatically. During the refresh process you may see Dataset changed too much error which should disappear once datasets are fully refreshed.

See more in update documentation.

cid-cmd update --force --recursive  --resource-tags 'tag_environment' --taxonomy 'company,business_unit,tag_environment'

Once dashboard is installed AND all datasets are updated, you can use filters and Group By elements in dashboards.

See more details on the live interactive dashboard demo.

FAQ

Do all dashboards support taxonomy?

The following dashboards support adding organizational taxonomy with cid-cmd tool: CID, KPI, CUDOS, FOCUS, Extended Support Cost Projection Dashboard, and Graviton Savings Dashboard. We plan to support this in all CID dashboards in the future.

I need to append a taxonomy from AWS Organization Tags

I just need to add a taxonomy based on AWS Cost Allocation Tag

Run update via command line. No other actions needed.

I want to use IAM Principal Tags or User Attributes for cost allocation

After Update and Adding Tags the refresh of SPICE datasets times out or shows a error about resources limitation

Try to re-update using tags with less cardinality (less unique values).

How to restore the default Account Map?

The default account map view is based on CUR 2.0 data. You can use this query to restore the view to its default state:

CREATE OR REPLACE VIEW account_map AS
SELECT DISTINCT
    line_item_usage_account_id                                        account_id,
    MAX_BY(line_item_usage_account_name, line_item_usage_start_date)  account_name,
    MAX_BY(bill_payer_account_id,        line_item_usage_start_date)  parent_account_id,
    MAX_BY(bill_payer_account_name,      line_item_usage_start_date)  parent_account_name
FROM
    "${cur_database}"."${cur_table_name}"
GROUP BY
    line_item_usage_account_id

How to update Account Map manually?

Can I change taxonomy?

Yes. Just re-update.

Can CloudFormation Template manage this?

As of now only you can use CID-CMD command line tool to manage taxonomy in as it allows choosing values from existing configuration in interactive way.