Overview
Your Outreach Agent doesn’t just run on autopilot—it learns as it goes. Even without manual input, it continuously tweaks how and when it reaches out based on what’s working. This ensures your outreach stays smart, responsive, and optimized for success.
Below, we break down the specific areas the Agent adjusts dynamically.
1. Sequencing: How the Agent Adjusts Cadence
When you create a Outreach Agent, it comes with a suggested cadence: a timeline for sending messages across different channels.
Before launch, you can edit this cadence manually. But once live, the Agent takes over—and starts testing different timing patterns to see what works best.
How it works:
The cadence you see on the Strategy page is a baseline—not a fixed schedule.
For example: If the cadence says to send a LinkedIn message on Day 5, the Agent may test sending it on Day 3, 4, or 6.
If LinkedIn messages on Day 3 are getting more responses, the Agent will shift more messages to Day 3.
Important: The Agent isn’t guessing—it’s learning from performance and prioritizing sequences that convert.
2. Task Prioritization: Deciding What to Send When
Every Agent has daily send limits—based on your number of inboxes and LinkedIn constraints. If there are more messages to send than it has capacity for, the Agent needs to prioritize.
Example scenario:
Daily email send limit: 50 messages
Net new prospects to email: 30
Follow-ups due that day: 40
Total tasks: 70, but only 50 can go out
How the Agent decides:
It will typically prioritize net new outreach
But it tests different mixes (e.g. more follow-ups vs. new sends) to learn what’s most effective
If follow-ups on Day 8 are converting well, the Agent will make sure to prioritize those—even if it means delaying less effective messages
3. Channel Testing: Finding the Right Order
Your Strategy page might show a sequence like:
Email → LinkedIn → Email → LinkedIn
But the Agent doesn’t treat that as gospel.
What it does:
Tests different orders (e.g. LinkedIn → LinkedIn → Email)
Monitors which sequence gets the most replies
Shifts more prospects to high-performing sequences
Note: This is how the Agent builds channel-level intelligence over time—ensuring your outreach adapts to what actually works.
4. Content Optimization: Learning What Messages Perform
Every message your Agent sends is generated uniquely—and tracked for performance.
What it analyzes:
Tone and language:
Does a casual CTA get more clicks in SaaS?
Do enterprise leads respond better to formal phrasing?
Call-to-action effectiveness:
Which lines are getting clicks or replies?
Which subject lines are getting opens?
Industry preferences:
What works for Founders in Fintech might flop for VPs in Manufacturing
The Agent doesn’t just personalize—it learns from how personalization performs.
5. Strategy Matching: Assigning the Right Messaging Framework
When your Agent is created, it develops several Content Strategies—each a unique structure and flow for messaging.
What it does:
Applies these strategies at random across leads
Tracks which strategies get responses
Learns what works best for different segments:
Strategy A performing with CEOs → more CEOs get it
Strategy B converting in Texas → more Texas leads get that flow