How to Set Up and Train AI-Powered Automatic Conversation Categorization

Papo's (https://papo.global/) automatic conversation categorization uses artificial intelligence to analyze and assign a score to each conversation, allowing your team to prioritize leads with the highest conversion potential. Setup is done by requesting an initial evaluation of your inbox, and the AI model is continuously trained through re-evaluations and the addition of contextual information, such as lead origin and the use of tags.

Papo's AI categorization feature is designed to bring efficiency and strategic focus to your sales and customer service teams. By analyzing conversation content, the AI assigns a score to each lead, helping to quickly identify which contacts are "hot" and deserve immediate attention.

Feature	Description	Main Benefit
Lead Scoring	The AI assigns a star rating (score) to each contact based on conversation analysis.	Focus on leads with the highest conversion potential.
Batch Evaluation	Allows analyzing all conversations in an inbox at once.	Time savings and an overview of contact base quality.
Re-evaluation	Option to request a new AI analysis for a contact after new interactions.	Keeps the lead score always updated and accurate.
Origin Definition	Allows manually recording the lead's origin (e.g., campaign, video, event).	Improves analysis of the return from specific marketing channels.

How AI Categorization and Scoring Works

Papo's Artificial Intelligence system reads the history of interactions with a customer to identify signs of interest, urgency, and purchase potential. Based on this analysis, it performs two main actions:

Categorization: Groups conversations and contacts based on their potential.
Scoring: Assigns a rating, represented by stars, indicating how "hot" a lead is. A higher number of stars means a lead with a greater likelihood of closing a deal.

This process transforms your inbox into a priority list, allowing agents to direct their efforts where returns are most likely, optimizing time and increasing conversion rates.

Step-by-step to activate automatic evaluation

To start using automatic categorization, you need to request that the AI analyze conversations in one of your inboxes. Each analyzed conversation consumes an AI credit.

Step	Action	Details
1	Access the feature	In the main panel, click on the "Evaluate open conversations" option, usually located in the top left corner.
2	Select the Inbox	Choose which inbox (WhatsApp, Instagram, etc.) you want the AI to analyze.
3	Start Evaluation	Click the "Start evaluation" button for the AI to begin processing conversations and assigning scores.
4	Analyze Results	After processing, you will see the star rating next to each contact in your conversation list.

How to train and refine AI accuracy

An AI model becomes more accurate over time and with more information. While Papo's system automatically learns from new interactions, you can accelerate and refine this process by providing more context.

1. Re-evaluating contacts

A lead may become more or less qualified after a new conversation. If a previously cold customer interacts again showing great interest, it's useful to update their score.

Select the desired contact from the conversation list.
In the contact information panel on the right, click the "Re-evaluate" button.
The AI will perform a new analysis considering the most recent interactions and update the lead's score.

2. Adding context with traffic origin

Knowing where your best leads come from is crucial for optimizing marketing investments. By informing a contact's origin, you enrich the data for future analysis and train the system to recognize patterns from high-converting channels.

In the same contact information panel, locate the origin field (it may appear as "Unknown origin").
Click to edit and define where the lead came from (e.g., "Black Friday Campaign", "YouTube Video", "Instagram Ad").

3. Using tags for manual categorization

While AI offers automatic scoring, you can create your own manual categorization system using tags. This helps organize contacts granularly and can be used in conjunction with AI scoring to refine segmentation.

Create relevant tags for your business, such as "Quote requested", "Technical issue", "Interest in Gold Plan", or "Follow-up in 7 days".
Apply tags to contacts during or after conversations.
Use these tags to filter and create segmented lists, combining your criteria with AI-driven prioritization.

Frequently Asked Questions

How does the AI decide what score to give a lead?

The AI analyzes the content of exchanged messages, looking for keywords, questions about price, demonstrations of urgency, general sentiment of the conversation, and other indicators to determine the lead's level of interest.

Is a lead's score updated automatically?

No. The initial score is based on the history up to the time of evaluation. To update the analysis based on important new interactions, you must use the "Re-evaluate" function in the contact's profile.

What are "AI credits"?

Each conversation analyzed or re-evaluated by the categorization feature consumes an AI credit. For details on the number of credits included in your plan and additional costs, please refer to the information on our website.

Can I define my own categories for the AI to use?

The AI assigns a potential score, not custom textual categories. To create your own categories, such as "VIP Customer" or "Marketing Lead", you should use Papo's tagging system.

What if I don't agree with the AI's score?

The AI's score is a powerful suggestion to help prioritize, but human judgment remains essential. Use the score as a guide and complement the analysis with your own tags and notes in the contact's profile.

Is conversation analysis instantaneous?

Processing time depends on the volume of conversations in the selected inbox. Generally, it is a quick process, and scores begin to appear in your conversation list within minutes.

The first user on your account is free for you to validate the complete flow.