Define the report before the send
Agree the reporting window, time zone, source systems, status mapping, recipient versus message counts, and deduplication rule. Preserve the source statuses so a summary can be traced to its records. A report design is not a promise that every system supplies every field.
Read the sample correctly
The synthetic report below contains 1,000 attempted messages to 1,000 unique recipients for one illustrative send. Delivered, failed, expired, pending, and unknown statuses form mutually exclusive buckets at the same snapshot. Accepted requests are an earlier lifecycle count; do not add them to those buckets.
Replies, opt-outs, and clicks need their own definitions
Count unique replying recipients separately from total reply messages. An opt-out may also be a reply, so those measures can overlap. Measured clicks require a documented tracking method and deduplication rule; automated link checks can affect the data. A click is not an SMS open or proof of a completed campaign action.
Specify the handoff
Your proposal should identify report cadence, export format, fields, access, and the responsible person. Ask whether the available export retains source event times, error categories, message-version references, unresolved outcomes, and suppression activity. The synthetic example demonstrates definitions, not a contracted export or live dashboard.
Website analytics is a separate system
Matomo measures use of the Ananasi website, such as page visits and sanitized interaction events. It is not the source of campaign-message delivery reports. Website event tracking excludes form answers, calculator inputs, and selected media; the privacy notice explains the website data flow.
Synthetic report · No customer data
An example campaign report
One illustrative send: 1,000 attempted messages to 1,000 unique recipients. All figures below are invented to demonstrate reporting definitions. The snapshot uses one reporting window and one status per attempt.
- Delivered
- 850Provider delivery reports
- Failed
- 80Terminal failure events
- Expired
- 20Source expiry events
- Pending
- 30Unresolved queue statuses
- Unknown
- 20No usable terminal report
850 + 80 + 20 + 30 + 20 = 1,000. Pending and unknown remain visible. A delivered event does not prove an observed open.
Response measures use explicit denominators
- Unique replying recipients
- 60 / 1,000 = 6%Deduplicated senders in the response log, within this example’s window.
- Unique opt-out recipients
- 12 / 1,000 = 1.2%Suppression events; these recipients can also be included among replies.
- Measured unique clicks
- 75 / 1,000 = 7.5%Illustrative linked measurement records after the chosen deduplication rule. Automated checks and attribution gaps still require review.
- Verified event responses
- 12 separate RSVP recordsIndependent outcome evidence. Do not attribute every RSVP to a text without a documented matching method.
Explore this topic
Choose the specific decision you need to work through.