> For the complete documentation index, see [llms.txt](https://docs.meetalto.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.meetalto.ai/using-meet-alto/playbooks/weekly-review.md).

# Make one useful improvement each week

A repeatable routine for improving answers, finding missed work, and measuring useful outcomes.

Choose a consistent date range and a person responsible for the review. Keep a short record of the problem, change, and result so the team learns what actually helped.

## Review the experience

Read a sample of successful conversations, unanswered items, handoffs, and corrected drafts. Include more than one account or workflow when they serve different audiences.

Ask: Did the customer get a direct answer? Was the information correct? Was the next step appropriate? Did the conversation stop or hand off at the right time?

## Review the system

* Check account and integration health.
* Review knowledge gaps, conflicts, processing failures, and outdated policies.
* Check the Inbox for unresolved work and unclear ownership.
* Inspect campaign and sequence status, destinations, and overlapping follow-ups.
* Review AI Credits, Active Contacts, and current allowances.

## Review the outcome

Use [Dashboard metrics](/using-meet-alto/analytics.md) with the same accounts and date range. Separate reach, replies, clicks, and collected details. A larger count may simply reflect more activity, so compare the customer experience too.

## Choose one change

Fix the highest-impact repeatable issue: an unclear collection prompt, a stale answer, an incorrect destination, or a confusing handoff. Update the owning source, save it, and retest the affected path.

Write down the expected improvement and when you will check it. Avoid changing several parts of a working flow at once. Share the learning with the teammate who owns the next review.