← All docs
Using Merit
Commands reference
How commands work, what they can do, and how to write effective ones.
A command is any message sent to @Mavrick in Slack. Commands can be questions, requests for data, requests to take action, or requests to generate a report. Merit interprets them in natural language — no syntax to memorize.
Types of commands
Data queries
Example
@Mavrick what's our Meta ROAS for the last 30 days, broken down by campaign?
Analysis commands
Example
@Mavrick compare this week's Google Ads performance to last week. Flag anything that changed by more than 20% in either direction.
Report generation
Example
@Mavrick generate a weekly brief for leadership: spend, ROAS, revenue, top performers, and three recommended actions. Format as a structured report.
Action commands
Example
@Mavrick pause all Meta campaigns where 3-day ROAS is below 2.0 — show me the list first before doing anything
Writing effective commands
- →Specify date ranges explicitly: 'last 7 days' beats 'recently'
- →Set thresholds when asking for flags: 'below 2.5x' is better than 'underperforming'
- →Ask for what you want to do with the data, not just the data itself
- →Use follow-up questions in the same thread — Merit keeps context
- →If a command returns too much, ask to narrow it: 'just the top 5'
Ready to start?
Free to start. Connects in 5 minutes.