---
title: What best practices are recommended for naming metrics and tags?
description: Datadog, the leading service for cloud-scale monitoring.
breadcrumbs: >-
  Docs > Extend Datadog > Extending Datadog Guides > What best practices are
  recommended for naming metrics and tags?
---

# What best practices are recommended for naming metrics and tags?

Naming convention is an art and possibly one of the most difficult decisions to agree on. Defining a naming convention for your metrics, tags, and services is crucial to have a clean, readable, and maintainable telemetry data. Here are some recommendations:

- Provide descriptive and meaningful names: the metrics or tags clearly describe the purpose or meaning of the value.
- Adhere to the format and limitations described below.
- Avoid abbreviations that might have multiple meanings
- Maintain consistency across all teams, apps, and services.
- Avoid reserved keywords that might cause clashes with the other tags or metrics.
- In the case of metrics, prefix them with a namespace depicting the application or service generating the data.
- Avoid the use of personal and sensitive data of any kind.

## Rules and best practices for naming metrics{% #rules-and-best-practices-for-naming-metrics %}

- Metric names must start with a letter.
- Can only contain ASCII alphanumerics, underscores, and periods. Other characters are converted to underscores.
- Should not exceed 200 characters (though less than 100 is generally preferred from a UI perspective)
- Unicode is not supported.

Metrics reported by the Agent are in a pseudo-hierarchical dotted format, for example: `http.nginx.response_time`. This is described as pseudo-hierarchical because a hierarchy is not actually enforced, but the structure is used to infer certain relationships, for example: *"I see hostA and hostB are reporting `http.nginx.*`, those must be web frontends"*).

**Note**: Metric names are case sensitive in Datadog.

## Rules and best practices for naming tags{% #rules-and-best-practices-for-naming-tags %}

As a best practice, Datadog recommends using unified service tagging when assigning tags. Unified service tagging ties Datadog telemetry together through the use of three standard tags: `env`, `service`, and `version`. To learn how to configure your environment with unified tagging, see [Unified service tagging](https://docs.datadoghq.com/getting_started/tagging/unified_service_tagging.md).

For a full list of tag requirements, including allowed characters, length limits, and normalization behavior, see [Define tags](https://docs.datadoghq.com/getting_started/tagging.md#define-tags).

Examples of commonly used metric tag keys are `instance`, `name`, and `role`.

## Further Reading{% #further-reading %}

- [Learn more about Datadog metrics](https://docs.datadoghq.com/metrics.md)
- [Getting started with tags](https://docs.datadoghq.com/getting_started/tagging.md)