Skip to main content

AI Pricing Integration Guide

Botsi AI Pricing Model helps you dynamically select and display the most effective paywall for different user segments, aiming to enhance conversion rates. This guide walks you through the prerequisites and configuration steps to get your AI Pricing Model up and running.

Configuration Requirements​

Before the AI pricing model configuration, at least one Application should be configured in Botsi.

To start configuring the AI Pricing Model, you need to have several components already set up in the Botsi platform. Please ensure you have completed the following configurations.

Follow the guides below for detailed instructions on setting up each requirement:

AI Pricing Model Configuration​

Once all the prerequisites are met, you can proceed with configuring the AI Pricing Model itself. To configure your model, follow these steps:

  1. Navigate to the AI Pricing Models page and click the Create Model button. This will launch the configuration form.

    If you don’t have any models yet, you will see an informational page with a brief overview of the feature. To begin the configuration process, click the Get Started button.

    Screenshot of the Botsi AI Pricing Models page. It shows a table listing existing models with columns for Revenue and State. A prominent blue 'Create Model' button is at the top.

  2. Enter a descriptive name for your model in Step 1. This will help you identify it later.

  3. Click Next - Add Paywalls to proceed to Step 2.

    Screenshot of the 'Create AI Pricing Model' form, focused on Step 1: Model Name. The user is typing a name into the model name field.

  4. In Step2, select all the paywalls that will be part of the prediction model. The model will analyze the performance of these paywalls to find the best option for each user.

    Screenshot of Step 2: Add Paywalls in the model creation process. A list of available paywalls is shown with several checkboxes selected.

  5. Set Assumptions for Faster Training. Under Advanced Settings, you can provide initial assumptions or upload historical data. This information helps the model train faster by giving it a baseline. This step is recommended but not required. After setting these, click Next - Set Distribution.

    Screenshot showing the Advanced Settings section. It displays fields for setting assumptions like 'Weekly Renewal Rate' and an area to upload historical data.

  6. In Step 3, Set Distribution and Launch your model. This is the final configuration step where you define how the model will be shown to your users. To do so, follow these steps:

    1. Use an existing placement: Recommended if you already have a Botsi placement and either do not want or cannot make changes in your app. In this case, select an existing placement.
    Create a new placement

    If no suitable placement exists, or if you prefer to configure one specifically for this setup, select Create New and provide a name and ID for it.

    1. Select Best Paywall. It will be used as a comparison paywall. Also, Best Paywall is used as a fallback if the prediction API does not respond.
    Screenshot of Step 3: Set Distribution & Launch. The 'Use Existing Placement' option is selected and the 'Set Model to Live' toggle is off, indicating an A/B test setup.
    1. Set distribution: Use the slider to set the distribution percentage between the Best Paywall and the AI Pricing Model. The set percentage dictates the frequency with which each paywall—Best or Predicted (AI)—will be displayed. In the example shown below, 70% of users see the Best Paywall, and 30% see the AI-powered one.
    Screenshot of the Model Distribution section for an A/B test. A slider is set to a 70% and 30% traffic split between the 'Best Paywall' and the AI Pricing Model.
    Activate the Live Toggle

    The Set Model to Live for all new users, don’t test against another paywall toggle automatically sets the traffic distribution to 100% for the AI Pricing Model.

  7. Once your configuration is complete, you have two options:

  • Launch Model: Activates the model immediately according to your settings.

  • Save as Draft: Saves your configuration to be launched later.

    Screenshot of Step 3 showing the 'Create New Placement' option selected. Input fields for 'Placement Name' and 'Placement ID' are visible. The 'Set Model to Live for all new users' toggle is switched on, showing the traffic distribution.

Post-Launch: Editing and Managing Your Model​

After your model has been saved or launched, you can select it from the AI Pricing Models list to view and manage its settings. The Edit AI Pricing Model screen allows you to adjust distribution, manage paywalls, and view the model's details at any time.

Screenshot of the 'Edit AI Pricing Model' modal window. It displays a summary of the model's settings with buttons to 'Launch Model' or 'Save as Draft' at the bottom.

See also​

After completing these preparation steps, you are ready to integrate with our API.