NOTE: It's easy to cause more harm than good when using this section incorrectly. Always try to start by editing your short description and extra prompt to reach desired results before editing this section.

You can fine tune the AI's training on objectives with the Objection training. This section should not be used unless you're fine tuning and have a solid understanding of how it works. If you use this section incorrectly, it can negatively impact your bot's accuracy.

The next section allows you to add up to seven examples of what the lead might say and how you would want the bot to respond. In this section you can also rate whether you want the bot to pass or fail this example and its satisfaction of your objective. Let's look at another example from the fitness bot objective Discover Motivator

The description of this objective is reason for reaching out to start fitness. You can see all of the examples with the ❌ are examples where we don't want the bot to satisfy the objective. The example with the ✅ is above our sensitivity rating so that would pass the objective! Users on the Basic or Free plans can click the icons to toggle satisfactory or unsatisfactory while higher tier plans can edit the fine-tuned rating number.

It’s a good idea when giving your passing example(s), you give at least one passing example that's vague to show the bot that it doesn't have to look solely for responses that CLEARLY satisfy the goal. It can accept gray area answers too.

Here the bot has to read between the nuances in speech to understand that the lead response “I’m not as happy as I’d like with how I look” means they are reaching out because they want help with improving their diet or exercise. On the contrary, if you used an example like “I’m reaching out to start fitness because I want to be more fit” the bot may be more strict on what kinds of responses it will accept as satisfactory.