General & Conceptual Questions

Q: What is the primary purpose of the Journey Path Optimizer?

A: The Journey Path Optimizer is a powerful tool designed to help marketers make data-driven decisions without manual efforts. It allows you to automate A/B test different paths or strategies within a single customer journey to identify which one most effectively leads to a desired conversion goal.

Q: How does the Path Optimizer work at a high level?

A: You can split users entering the path optimizer node into multiple (up to 5) different paths. Each path can contain a unique sequence of messages, channels, or time delays. The feature tracks how many users from each path complete a predefined goal. Based on this performance data, it determines the "currently winning" path in real-time and optimizes the predefined goal.

Q: What are the key benefits of using the Path Optimizer?

A: The main benefits include:

• Increased Conversions: By systematically identifying and scaling the most effective marketing sequence.
• Improved Customer Engagement: Discovering which channels, message copy, or timing resonates best with your audience.
• Reduced Guesswork: Replacing assumptions with concrete performance data to guide your strategy.
• Enhanced ROI: Ensuring your efforts are focused on the most impactful communication strategies.

Q: What is the difference between the Path Optimizer and a standard A/B split test node?

A: A standard A/B split test typically is manual testing by splitting the audience as per assumptions or hit & trial method. The Journey Path Optimizer is more advanced, allowing you to automate audience distribution by identifying the best path at that point in time and generating the highest conversion rate. For example, you can test a path with an Email followed by an SMS against a path with a Push Notification followed by an SMS.

Setup & Configuration

Q: How many different paths can I create in a single optimizer?

A: You can configure a minimum of 2 and a maximum of 5 paths to test against each other.

Q: How is the audience distributed among the different test paths?

A: JPO uses balance exploration and exploitation. Rather than applying a fixed split, the model evaluates each variant's real-time performance and dynamically routes more traffic toward the strongest-performing path, while continuing to explore others to account for changing user behavior. This distribution shifts continuously as data accumulates.

Q: What is a "Control Group" in the context of the Path Optimizer?

A: The Control Group is a small percentage of your audience that is intentionally held back from receiving any communication from the test paths. This allows you to measure the organic conversion rate (i.e., conversions that happen without any influence from your journey) and compare it against the performance of your test paths to understand the true uplift.

Q: What defines the "Goal" for the Path Optimizer?

A: The goal is the specific conversion event you want users to complete. This could be making a purchase, filling out a form, clicking a specific link, or any other custom event you are tracking. This goal is what the optimizer uses to measure the success of each path & it is defined while creating the journey.

Q: How do I define the test duration or sample size?

A: You just need to define the winning criteria (Clicks% or Conversion%). Path optimizer will do the heavy lifting and optimize the paths based on the winning criteria. As and when the users pass through the path optimizer node, it will do the distribution by determining the best performing path at that point in time.

Real-Time Optimization & Path Distribution

Q: How does the Path Optimizer decide which path to send a user down?

A: JPO continuously analyzes variant performance in real time. Based on live engagement data, it dynamically routes a larger share of incoming users toward the path currently showing the strongest performance — while still exploring other paths to adapt to shifts in user behavior over time.

Q: Is a single "winning path" ever permanently declared?

A: No, the system does not declare a single, final winner. User behavior can change over time (e.g., a path with a discount might work well during a sale, but not after). The optimizer is designed to continuously adapt to these changes, perpetually shifting traffic towards the most effective path at that specific point in time.

Q: Why do the traffic distribution percentages for my paths keep changing in the dashboard?

A: The changing distribution is the core of the optimization process. As the system gathers more performance data, the algorithm intelligently sends a larger percentage of new users down the path that is currently yielding the highest conversion rate. If another path starts to perform better later, the distribution will shift again to favor the new leader. This feature is working as intended.

Q: What if one path performs better on weekdays and another on weekends?

A: This is exactly the kind of scenario the Path Optimizer is built to handle. The algorithm will detect the shift in performance. For example, it might send more traffic to Path A during the week and then, as it sees conversion data change in real-time, automatically start sending more traffic to Path B over the weekend.

Q: How does JPO determine the best-performing variant?

A: JPO uses a statistical learning approach to continuously evaluate how each variant is performing. On the first run, all users are routed to Variant A to establish a baseline. The model then recalibrates continuously as engagement data accumulates, progressively directing more traffic toward the stronger-performing variant while continuing to explore others.

Visible distribution across variants develops over subsequent runs, once the model has sufficient data to make informed routing decisions.

📘

Tip

For faster and more dynamic optimization, use JPO with trigger-based journeys. Continuous user entry gives the model more frequent recalibration opportunities, leading to quicker variant exploration and clearer performance signals.

Q: Why does variant distribution shift more gradually in a dataset journey?

A: In dataset journeys, users enter simultaneously in large volumes. Since the model recalibrates on a rolling basis, users processed within the same window are routed using the current benchmark, with refined insights applying to the next scheduled run. The larger and more continuous the data flow, the faster the model evolves.

For the most dynamic optimization, trigger-based journeys are recommended. Because users enter continuously over time, the model recalibrates between entries — surfacing the best-performing variant faster as data matures.

Analytics & Reporting

Q: What specific metrics can I view for my Path Optimizer test?

A: The analytics dashboard provides a detailed breakdown for each path, including:
Impressions of users who entered the path.

• Number of conversions/clicks.
• Conversion/Click rate.

Q: Can I monitor the performance of the test in real-time?

A: Yes, the journey analytics are updated in near real-time, allowing you to track the performance of each path as users flow through the journey.

Q: Why is there a discrepancy between Path Optimizer conversion count and actual node-level conversions?

A: Path Optimizer tracks conversions only while the journey is active. However, due to the attribution window, conversions can still be recorded at the node level (e.g., WhatsApp) after the journey ends.

This can lead to higher conversion counts at the node compared to what's shown in the Path Optimizer.

📘

Use Path Optimizer for in-journey insights and node-level data for total attributed conversions.

Technical & Use Case Questions

Q: Can I use the Path Optimizer to test different channels (e.g., Email vs. Push Notification)?

A: Absolutely. This is a primary use case. You can create one path that sends an email and another path that sends a push notification to see which channel drives more conversions for that specific audience segment and context.

Q: Can I test different timings (e.g., wait 1 day vs. wait 3 days)?

A: Yes. You can configure one path with a 1-day time delay before a message and a second path with a 3-day delay to determine the optimal engagement timing.

Q. Once a journey with a Path Optimizer is live, can I modify the path configurations?

A: Yes, you can. You would need to pause the current journey and resume after making changes. Please note that once the journey is redeployed, path optimizer data will also get reset and it would start afresh.

Q: Is there a recommended audience size for running a reliable test?

A: While there is no fixed minimum, a larger and continuously flowing audience yields faster, more reliable optimization. For dataset journeys with smaller audiences, the model may require multiple runs before meaningful distribution shifts appear. A sustained audience of several hundred users or more per path provides the data density JPO needs to confidently identify and scale the best-performing variant.