Skip to main content

How to Report a WallaPredict Error

A quick guide to understanding and reporting unexpected predictions

Updated over 9 months ago

WallaPredict uses AI to surface early signals of client churn based on patterns—not just single actions. Because predictions are built from multiple combined data points, you may not always be able to "prove" something is off with a report. And that’s okay. If a prediction feels inaccurate or doesn’t match what you know about a client, your feedback is still valuable. Flagging potential issues helps us improve the model and make WallaPredict smarter over time.


What to Know When Reporting

AI predictions work differently than reports. The model looks at patterns and trends, not single data points. That means:

  • You might not be able to "prove" the prediction is wrong using one report

  • The score is influenced by multiple factors combined over time

  • Not every signal is visible in the UI—some inputs come from behind-the-scenes data

Example: A client may be flagged for churn risk based on declining pass usage, longer gaps between bookings, and past cancellations—even if they booked a class yesterday.


What Counts as a Potential Error?

Because WallaPredict is powered by AI, it may occasionally flag something that doesn’t quite line up with your expectations. Here are a few examples of what we consider potential errors worth reporting:

  • Outdated or missing data
    Ex: The client visited recently, but WallaPredict shows no recent activity.

  • Inconsistent scoring
    Ex: Two clients with nearly identical behavior have very different churn risks.

  • Unexpected prediction
    Ex: A loyal client is flagged as at risk, even though they’re actively booking.

  • Missing signals
    Ex: A referral or plan purchase doesn’t appear to be factored in.

  • Confusing logic
    Ex: The prediction doesn’t make sense based on what you know about the client.

You don’t need to be 100% certain something is wrong—if something feels off, it’s helpful to let us know.


How to Report a Potential Error

  1. Open the Client Panel

  2. Select the clients Churn Score

  3. Choose Report a potential error

  4. Send a WallaPredict Ticket to the team

  5. Add details—be as specific as possible

  6. Submit your ticket. You’ll receive a confirmation reply once we receive it.


💡 What Happens Next

  • Our team will review your ticket.

  • If it’s an issue we can fix right away (like missing data), we’ll follow up.

  • If your feedback relates to how the AI model makes predictions, we’ll use it to improve future results and keep you in the loop.

  • You may get a confirmation message if more information is needed.


You don’t need to wait for a resolution to keep using WallaPredict. This process is designed to be lightweight!


🙌 Thank you for helping improve WallaPredict
Your experience and observations are essential in making churn predictions more reliable. If you ever have questions about how predictions work or what to expect, start a chat with us anytime.

Did this answer your question?