Analytics

Deliverability analytics

Deliverability analytics tell you what happened to the email that you sent. You select metrics and a time window. Lettermint returns the totals, and can also return a time series and a breakdown by provider, domain, tag, or SMTP reply.

Deliverability analytics are available on the Pro plan.

Questions that analytics answer

Which mailbox provider defers or refuses your email?
Did the bounce rate change after your last release?
Which replies from receiving servers cause the most bounces?
How long does delivery take for most recipients, and for the slowest 1 percent?
Do people open your email, or do machines?
Which campaign tag has the most complaints?

Ways to use analytics

Method	Use it to	Status
API	Build reports, alerts, and dashboards in your own systems	Available
MCP server	Ask an AI assistant a question in plain language	Available
Dashboard	See charts in the Lettermint dashboard	Not available yet

The API and the MCP server run the same query on the same data. They return the same numbers.

How Lettermint counts

These rules explain most of the numbers that you see.

The unit is a recipient

One email to three addresses is three recipients. Each recipient has its own delivery result, so Lettermint counts recipients. The metric messages is the exception. It counts emails.

An event counts at the time that it happened

Lettermint does not count an event at the time that you sent the email. An email with this history adds to three different hours:

Time	Event	Metric that increases
09:58	Lettermint accepts the email	`accepted` in the 09:00 bucket
10:02	The receiving server accepts the email	`delivered` in the 10:00 bucket
14:30	The recipient opens the email	`human_opens` in the 14:00 bucket

This method shows you when something happened. A provider that starts to defer your email at 10:00 shows up in the 10:00 bucket, even for email that you sent the day before.

It has one side effect. In a short window, the opens can belong to email that was delivered before the window. A rate can then be higher than 1.

The buckets of a time series add up to the summary

Each event is in one bucket only. The sum of a count over all buckets is equal to the count in the summary. This applies to counts, not to rates or percentiles.

A recipient opens one time

human_opens counts a recipient one time, when the first open happens. The recipient stays counted in that bucket, and later opens do not count again. To count each open, use human_opens_events.

Test email and Sandbox email are not counted

Email to the test addresses at @lettermint.dev is not in analytics. Email from a Sandbox project is not in analytics.

Analytics are not real time

A new event appears in analytics after a short delay. Each response includes meta.last_ingested_at, the time of the newest data that the query read.

What analytics do not tell you

Inbox placement. delivered means that the receiving server accepted the email. The server can still put the email in the spam folder.

The history of one recipient. Analytics do not store recipient addresses. To find one email, use Email activity.

Inbound email. Analytics cover the email that you send.

Access

Method	You need
API	A Team API token with the `read:analytics` ability
MCP server	A role with the View analytics permission, in a team that allows MCP access

Two more rules apply:

A team member sees the analytics of the projects in their project access only. A Team API token sees all projects of the team.
The dimensions from_address and subject hold personal data. The token also needs read:messages. A team member also needs View messages.

How long Lettermint keeps analytics

A query can cover up to 90 days. The data that is available depends on the retention period. See Data retention.

Each response includes meta.available_since. Lettermint has no analytics before this time.