Returns aggregate delivery and latency stats grouped by tag (`name:value`) for the requested period. Rows are ranked by the `sort` metric (default `accepted`) descending and capped at the requested `limit` (default 50, hard maximum 200). Use this to compare delivery performance across the campaigns, experiments, or segments you tag your sends with. Only tagged messages appear. A message carrying several tags is counted once under each of them, so rows do not sum to the period total. Rows are attributed to send time, not event time: a delivery confirmed during the period for a message accepted earlier counts against the earlier period, not this one. A recent period therefore under-reports `delivered` while its delivery reports are still arriving, and its counts only ever grow toward the truth. The maximum window is 365 days; requesting a longer range returns 422. Start date (inclusive) in YYYY-MM-DD, interpreted as a calendar day in `timezone` (a UTC day when `timezone` is omitted). Defaults to 30 days before `to` when omitted; with `include_trend=true` and `trend_grain=hourly` the default tightens to keep the window within the 720-hour trend cap. End date (inclusive) in YYYY-MM-DD, interpreted as a calendar day in `timezone` (a UTC day when `timezone` is omitted). Defaults to today in that timezone when omitted. Window may not exceed 365 days. IANA timezone identifier (for example `Asia/Kathmandu`) to report in; defaults to UTC. Day and hour boundaries and the default window when `from` and `to` are omitted both follow it, so a calendar-day `from` or `to` names a local day. A `from` or `to` carrying its own UTC offset is rejected while this is set: pass a calendar day or a `Z` instant. Metric to rank rows by, applied descending. Any lifecycle count or derived rate in the response may be used; rows whose rate is undefined (zero denominator) sort last. Defaults to `accepted`. Possible values: `accepted`, `sent`, `delivered`, `undelivered`, `failed`, `rejected`, `expired`, `delivery_rate`, `failure_rate` Maximum number of tag rows to return, ranked by the `sort` field descending. When true, each row also carries a `trend` array: a short per-bucket lifecycle-count series for that row over the window. Returned only when `limit` is 50 or fewer and the window is at most 90 days (trend_grain=daily) or 720 hours (trend_grain=hourly); a larger request returns 422. Bucket grain for the `trend` series. Has no effect unless `include_trend=true`. Possible values: `daily`, `hourly` The date range the response covers (echoed back from the request), plus `data_as_of`, the freshness boundary the data is current to. Inclusive start of the window the response covers -- a calendar day (YYYY-MM-DD) for day windows, or an hour boundary (RFC 3339) for hour windows. Inclusive end of the window the response covers -- a calendar day (YYYY-MM-DD) for day windows, or an hour boundary (RFC 3339) for hour windows. The instant the statistics in this response are current to: events recorded up to roughly this time are reflected, while more recent events may not be yet. Statistics are served from a rolling aggregation that refreshes every few seconds, so a response is near-real-time but not live; use this field to label data freshness (for example "as of 14:03") rather than assuming the numbers are to-the-second. Null when the freshness boundary is not being reported. Tag breakdown rows, ranked by the `sort` metric (default `accepted`) descending. Empty when no tagged messages were sent in the period. The tag this row aggregates, in `name:value` form. Each distinct name-and-value pair is its own row, and a message carrying several tags is counted once under each of them, so rows do not sum to the period total. Distinct messages accepted for sending after admission checks. This is the denominator for `delivery_rate` and `failure_rate`. Distinct messages handed off to the carrier for delivery. Distinct messages the carrier confirmed as delivered to the handset. Distinct messages the carrier reported as not delivered. Distinct messages that failed during sending. Distinct messages rejected before any send attempt, for example by sending policy or a message-generation failure. Distinct messages that could not be delivered within their validity window and expired. Share of accepted messages that were delivered, computed as `delivered / accepted`. Null when no messages were accepted in scope. Share of accepted messages that ultimately failed, computed as `(undelivered + failed + expired) / accepted`. Null when no messages were accepted in scope. p50, p95, and p99 latency percentiles in milliseconds for one latency family over the scope. Percentiles are approximate (computed from a high-volume aggregation pipeline). All three are null together when no qualifying event contributed a latency measurement in scope. Median (50th percentile) latency in milliseconds. Null when no qualifying event contributed a measurement. 95th percentile latency in milliseconds. Null when no qualifying event contributed a measurement. 99th percentile latency in milliseconds. Null when no qualifying event contributed a measurement. p50, p95, and p99 latency percentiles in milliseconds for one latency family over the scope. Percentiles are approximate (computed from a high-volume aggregation pipeline). All three are null together when no qualifying event contributed a latency measurement in scope. Median (50th percentile) latency in milliseconds. Null when no qualifying event contributed a measurement. 95th percentile latency in milliseconds. Null when no qualifying event contributed a measurement. 99th percentile latency in milliseconds. Null when no qualifying event contributed a measurement. p50, p95, and p99 latency percentiles in milliseconds for one latency family over the scope. Percentiles are approximate (computed from a high-volume aggregation pipeline). All three are null together when no qualifying event contributed a latency measurement in scope. Median (50th percentile) latency in milliseconds. Null when no qualifying event contributed a measurement. 95th percentile latency in milliseconds. Null when no qualifying event contributed a measurement. 99th percentile latency in milliseconds. Null when no qualifying event contributed a measurement. Per-bucket lifecycle-count series for this tag over the window, bucketed by `trend_grain`. Sparse, so only buckets with activity are present rather than zero-filled, unlike the daily and hourly series. Present only when `include_trend=true`. The day (YYYY-MM-DD) or hour (RFC 3339, on the hour) this point covers, matching the period's grain. Distinct messages accepted for sending after admission checks. Distinct messages handed off to the carrier for delivery. Distinct messages the carrier confirmed as delivered to the handset. Distinct messages the carrier reported as not delivered. Distinct messages that failed during sending. Distinct messages rejected before any send attempt, for example by sending policy or a message-generation failure. Distinct messages that could not be delivered within their validity window and expired. Total number of distinct tags with activity in the period, regardless of `limit`. When it exceeds the number of rows returned, the ranking was capped; raise `limit` (up to 200) or narrow the window to see more.