How to use the Sentiment Analysis Agent to route contacts by engagement intent

The Sentiment Analysis Agent reads how your contacts engaged with LinkedIn posts, including their comments and reactions, and automatically routes them into separate workflow paths based on whether their engagement signals positive, neutral, or negative intent.

Instead of treating everyone who engaged with a post the same, the agent separates the genuinely interested from the passive and the skeptical, so your outreach reaches the right people with the right message at the right time.

In the Start campaign setup node, go to the Audience tab. Select Source from LinkedIn, then choose LinkedIn post. Enter a list name and up to three post URLs, and select which engagement types to include: Reacted, Commented, or Reposted.

You can also turn on Auto re-sync to keep pulling in new engagers over time.

This step is required. The agent has no engagement data to analyze if your source is not a LinkedIn post.

Click the + button on the canvas and under Agents, select Sentiment analysis agent.

The agent panel shows three segments, each with its own toggle:

Toggle on the segments you want to create paths for. Each active segment becomes its own output path on the canvas.

Use the Merge toggle if you want two segments to rejoin into a single workflow after the analysis instead of running separate sequences. Three merge scenarios are available, but only one can be active at a time:

Select the scenario that fits your campaign. The Overview tab on the node shows your active segments and merge configuration at a glance.

Click Save when your segments are configured.

Add the next steps under each output path. A common structure:

When the campaign runs, the agent analyzes each contact's engagement with your source posts: what they commented, how they reacted, and the tone of that engagement. It then assigns the contact to one of your active sentiment segments and sends them down the matching path automatically. No manual review is needed.

For every contact the agent analyzes, an AI Insight is added to their Activity history explaining what the analysis found.

To view it, open the campaign's Inbox tab, click a contact, and select the Activity tab on the right. The AI Insight appears alongside the engagement entry (for example, "commented on a post about...") and summarizes:

This is useful before jumping into a conversation. The insight gives you the context behind the contact's sentiment so you can reference the post and their reaction naturally in your reply.

Note: AI Insights are generated automatically and can make mistakes, for example misreading a contact's intent. Use them as context, not as a verdict.

Run lead scoring first. Add the Profile analysis & Lead scoring agent before the Sentiment Analysis Agent, with a Conditions node filtering for score 75 and above in between. That way you only analyze sentiment for contacts who fit your ICP, and your outreach list is filtered for both fit and intent.

Merge Positive & Neutral for most campaigns. Neutral engagers are often just quieter prospects. Merging them with positive contacts keeps your reachable audience larger while still excluding negative sentiment.

Reference the post in your outreach. Contacts routed by this agent engaged with a specific post. Messages that mention the topic (or use the AI Ice Breaker variable) convert noticeably better than generic openers.

Let AI write from their engagement. Contacts routed by this agent engaged with a specific post, and messages that reference it convert noticeably better than generic openers. You do not have to write these yourself: in any messaging node, click AI assist and select Write from engagement. The AI drafts a message based on the contact's actual engagement, such as what they commented and which post it was on. This option is available when your campaign uses a LinkedIn post source.

Choose relevant source posts. The agent is only as useful as the posts it analyzes. Competitor posts and industry discussion posts where your ICP is active work best.

How do I know why a contact was routed to a particular segment? Open the contact from the campaign Inbox and check the Activity tab. The AI Insight entry explains what the contact engaged with and what their engagement signaled.

Which source types does the agent support? Only LinkedIn post sources. If your audience comes from a CSV, saved list, search, group, event, newsletter, or profile visitors, the agent has no post engagement to analyze.

What if a contact engaged with more than one of my source posts? The agent combines all of their engagement. If a contact commented on one post and reacted to another, their sentiment is based on their overall engagement across all your selected posts, not on any single interaction.

Can I change my segments after publishing? No. Sentiment segments cannot be changed once the campaign is published. Configure your segments and merge settings carefully before going live. If you need a different segment setup, create a new campaign.

Do I have to create a path for every sentiment? No. Toggle on only the segments you want. Contacts falling into a disabled segment will not continue in the workflow.

If you have any questions or need help, click the chat button at the bottom of your screen and our support team will be happy to help.