ref(alerts): Add docs for dynamic alerts (#11158)

* feat(alerts): Add anomaly detection docs

* update screenshot

* Add plan gate banner and note about alert types

* update text for options

* force push ate some stuff

* just trying to get linting to run

* test

Author: ceorourke

docs/organization/early-adopter-features/index.mdx

@@ -22,3 +22,4 @@ Limitations:
 - [Span Summary](/product/performance/transaction-summary/#span-summary)
 - [Investigation Mode](/product/performance/retention-priorities/#investigation-mode) for retention priorities in Tracing
 - [Uptime Monitoring](/product/alerts/uptime-monitoring/)
+- [Dynamic Alerts](/product/alerts/create-alerts/metric-alert-config/#dynamic-alerts)

docs/product/alerts/alert-types.mdx

@@ -36,7 +36,7 @@ The **Alert Details** page also includes a list of issues that triggered the ale
 
 <Include name="feature-available-for-plan-team.mdx" />
 
-Metric alerts tell you when a [metric](/product/performance/metrics/) crosses a threshold set by you, like a spike in the number of errors in a project, or a change in a performance metric, such as [transaction duration](/product/performance/metrics/#latency), [Apdex](/product/performance/metrics/#apdex), [failure rate](/product/performance/metrics/#failure-rate), or [throughput](/product/performance/metrics/#throughput-total-tpm-tps).
+Metric alerts tell you when a [metric](/product/performance/metrics/) crosses a threshold set by you, like a spike in the number of errors in a project, or a change in a performance metric, such as [transaction duration](/product/performance/metrics/#latency), [Apdex](/product/performance/metrics/#apdex), [failure rate](/product/performance/metrics/#failure-rate), or [throughput](/product/performance/metrics/#throughput-total-tpm-tps). You can use [dynamic alerts](/product/alerts/create-alerts/metric-alert-config/#dynamic-alerts) to let Sentry define the threshold for you.
 
 Metric alerts monitor macro-level metrics for both error and transaction events. A metric takes a set of events and computes an aggregate value using a function, such as `count()` or `avg()`, applied to the event properties over a period of time. When you create a metric alert, you can filter events by attributes and <PlatformLink to="/enriching-events/tags/">tags</PlatformLink>, which is particularly useful for aggregating across events that aren't grouped into single issues. Sentry allows a maximum of 1000 metric alerts for an organization.
 

docs/product/alerts/create-alerts/img/dynamic-threshold.png

@@ -111,10 +111,11 @@ While Sentry won’t allow you to create new alerts with invalid or unavailable
 
 ## Thresholds
 
-There are two threshold types:
+There are three threshold types:
 
-- **Static**: A fixed threshold, such as when there are 100 errors in a period of time.
-- **Percent change**: A dynamic threshold, such as when there are 10% more errors in a time period compared to a previous period. These are also referred to as [Change Alerts](#change-alerts-percent-change).
+- **Static**: A fixed numerical threshold set by you. (For example, if there are 100 errors in a set period of time.)
+- **Percent change**: A percent based threshold, such as when there are 10% more errors in a time period compared to a previous period. These are also referred to as [Change Alerts](#change-alerts-percent-change).
+- **Dynamic**: A dynamic threshold set by Sentry that detects anomalies whenever values fall outside of expected bounds.
 
 By default, metric alerts use a fixed threshold.
 
@@ -126,6 +127,19 @@ Change alerts, or alerts that use a percent change threshold, are useful when yo
 
 ![When the percent change option is selected.](./img/percent-change.png)
 
+### Dynamic Alerts
+
+<Include name="feature-available-for-plan-trial-business.mdx" />
+<Include name="feature-available-for-user-group-early-adopter.mdx" />
+
+Dynamic alerts can be used when you don't know what threshold to set, but you know you want to be alerted when something is far outside the bounds of normalcy. Sentry will look at the historical data for the given metric and determine if the current data is anomalous. You have three options for the responsiveness of the alert: low, medium, and high. Low responsiveness means the alert will fire less frequently with each instance being higher signal. High responsiveness will capture more instances of the alert, which creates more coverage with a greater likelihood for false positives. Our recommendation is to start with our default (medium), and iterate based on the results you see.
+Another setting is to set the direction of the alert: above and/or below the expected bounds. This setting can help increase the signal of your alert rule.
+
+Dynamic alerts do not use the Critical, Warning, and Resolved thresholds like other alert threshold types as you are not setting these yourself. When no more anomalies are detected, the alert will resolve. Dynamic alerts are not available for all metric alert types.
+
+
+![When the dynamic threshold is selected.](./img/dynamic-threshold.png)
+
 ### Set Threshold to Trigger Alert
 
 You can set the status of an alert rule when a threshold is met using the labels: