The report is the last step.

A weekly report should help someone make a decision. First, they need to trust its comparisons, understand missing data, and know where delivery stands. In partnership with TMAI, we built an analytics reporting system around those responsibilities.

Separate the data from the document

n8n collects platform data and assembles it into a shared format. That format defines the reporting periods, metrics, sources, and limitations. A separate Python service validates the input and renders the branded PDF; n8n handles email and Teams delivery steps.

This lets integrations and page designs evolve separately, with a common definition of what the data means.

Give a missing number an explanation

If a social platform cannot return data, replacing the missing value with zero would turn an integration problem into an apparent performance collapse.

The contract distinguishes complete, partial, and unavailable coverage. Partial totals require a disclosure; comparisons are rejected when the current and previous totals represent different sets of sources. A cross-channel timeline also makes clear that events happening on the same day do not establish cause.

Make failures actionable

The cross-channel page raises a layout error when meaningful text cannot fit. A report request interrupted by an uncertain network response keeps its operation identity for reconciliation instead of blindly starting again. The workflow also includes a send ledger and delivery checks.

Give people control over the cycle

Herald, the reporting operations bot, offers report, resend, health-check, and status commands. It checks its recorded operation against n8n before describing whether work is still underway.

A successful run does not prove someone received or read the email. Distinct states help operators investigate the right step.

The business capability is connecting integrations, validation, document design, and controls so a recurring report stays understandable when requirements change.