The Review System

AI generates, humans approve — that's the core principle of the review system. No matter how good your AI Team gets, having a human check on important outputs is essential for brand safety and quality control.

The review system exists to prevent embarrassing mistakes, off-brand messaging, or inappropriate engagement. It's the safety net that lets you benefit from AI automation while maintaining full control over your brand's public presence.

• Prevents off-brand content from being published
• Catches tone or context errors the AI might miss
• Gives you final say on all outgoing engagement
• Builds trust in the AI Team's output over time
• Provides feedback that helps the Orchestrator improve

The Review tab in your AI Team side panel shows all pending items that need your attention. A badge on the tab shows the count of pending items.

Queue Organization

Pending items are organized by type and can be filtered for easier navigation:

• Posts — Social media posts, carousels, and other content generated by Content Gen and AI Clone
• Comments — AI-drafted replies to comments on your posts or in relevant threads
• DMs — Personalized direct messages drafted by the Auto DM bot
• Emails — Follow-up email sequences created by the Follow-Up bot
• Ads — Ad creatives, copy, and campaign suggestions from the Ad Campaign bot
• SEO — Blog posts and articles created by the SEO & Blog bot

Status Filters

• Pending — Items waiting for your review
• Approved — Items you've approved (for reference)
• Rejected — Items you've rejected (for reference)

When reviewing content pieces (posts, articles, carousels), each item shows the full content preview along with metadata about why it was created.

What You See

• The full content text, including any hashtags and formatting
• Target platform(s) for publishing
• Which bot created it and what directive it was fulfilling
• Any attached media (images, video thumbnails)
• Scheduled publish time (if applicable)

Your Actions

• Approve — Accept the content for publishing
• Edit — Modify the content inline (text, hashtags, etc.) before approving
• Reject — Discard the content entirely
• Publish — Publish immediately to the target platform
• Schedule — Schedule the content for a specific date and time
• Skip — Skip this item without making a decision

Comments and DMs require extra care because they're direct conversations with real people. Each pending engagement item shows the full conversation context.

Context View

• The original comment or conversation thread
• Who wrote the original message (profile info if available)
• The AI-drafted response
• Relevance score explaining why this engagement was prioritized
• Which bot drafted the response

Review Tips for Engagement

• Check that the response actually addresses the original comment's point
• Verify the tone matches the context (supportive for complaints, enthusiastic for praise, helpful for questions)
• Make sure no sensitive information is disclosed
• Ensure the response doesn't make promises your team can't keep
• Confirm the response sounds natural, not robotic

Your automation level determines how much goes through manual review vs. auto-approval. You can change this anytime in the Settings tab.

Conservative

Everything goes through review. Every post, comment, DM, and ad creative requires your explicit approval before being published or sent. Best for new teams, sensitive brands, or when you want full control.

Balanced (Recommended)

Routine, low-risk actions are auto-approved (like simple comment replies that match established patterns). Important decisions — publishing new content, sending DMs to high-value leads, ad campaigns — still require your review.

Aggressive

Most actions are auto-approved. Only high-stakes items (large ad budgets, sensitive topic content, outreach to high-profile contacts) require review. Only use this after you've verified the team consistently produces quality output over several weeks.

Effective reviewing is about finding the right balance between speed and thoroughness. Here are practices that successful AI Team operators follow:

Timing

• Set fixed review times (e.g., 9am and 3pm) rather than checking constantly
• Prioritize time-sensitive items: comments and DMs first, then content, then ads
• Don't let items sit in the queue for more than 24 hours — engagement loses value over time

Prioritization

• Review DMs and comments first — they have the shortest effective lifespan
• Check content posts next, especially those with upcoming publish times
• Ad creatives can be reviewed last — they typically run on longer timescales
• Pay extra attention to content about sensitive topics, competitors, or customer complaints

Quality Signals

After a few weeks, you'll notice the AI Team's output quality improving. Good signs that you can increase automation:

• Your approval rate exceeds 85% — most generated content is on-brand
• You're mostly making minor edits rather than major changes
• The Orchestrator's strategies align with your marketing vision
• Engagement bot responses consistently match your brand's tone