Reporting

The headline numbers for one send. Start here; the other reporting endpoints break these down.

Rates are over total_recipients — the intended audience — not over what the provider accepted. That keeps them comparable with the list page and means a send with bounces reads slightly lower than a provider-side dashboard would show.

Three open counts, deliberately. apparent_opened is every contact with any open event, which is the number other tools report. verified_opened is the headline: non-bot openers plus anyone who clicked, since a person cannot click without opening — this is what guarantees open rate ≥ click rate. confirmed_opened is the strict high-confidence floor. total_opened and total_clicked are high-confidence only, so they already exclude bots.

pending_opened and pending_clicked are engagement the bot classifier has not judged yet, and they are forced to 0 once the send is no longer live — a late open on a finished campaign is organic tail, not a scanner flood, so do not read a zero as "nothing waiting". classifier_last_run_at and classifier_interval_ms are likewise only populated while the send is live; they are null otherwise.

Every tracked destination in the email with its click counts, busiest first.

Counts are settled and bot-filtered, so they will be lower than a raw click log.

click_rate is unique_clicks over the broadcast's recipients, so the link rates do not sum to the broadcast's click rate — one reader clicking three links counts once in each row.

Links that were never clicked still appear, with zeroes and null timestamps: the universe comes from what was sent, not from what was clicked. That is what makes this usable as a "which call to action failed" view.

Headline metrics for all of the account's broadcasts at once — the endpoint to build a dashboard on, instead of calling /report in a loop.

Served from a cache with a long TTL. For a send that just started, or after an action you expect to have moved the numbers, pass ?refresh=true to invalidate and recompute before reading. Without it a freshly-started send can read stale for as long as the TTL.

Leaner than /report: opened and clicked here are the verified (bot-filtered) counts, there is no apparent/confirmed split, and there is no total_failed. sparkline is the first 6 hourly open counts from the broadcast's first open, for a trend glyph.

refresh is not declared as a parameter on purpose — this route has never had a querystring schema, and adding one would change what Fastify does with the request. It must be exactly true; any other value, 1 included, is ignored.