saas-sales-conversations

Dataset Description

This is a synthetic dataset of sales conversations for SaaS (Software as a Service) companies, designed for training sales conversion prediction models. The dataset was created following the methodology presented in "SalesRLAgent: A Reinforcement Learning Approach for Real-Time Sales Conversion Prediction and Optimization" (Nandakishor M, 2025).

The dataset contains realistic dialogues between sales representatives and potential customers, along with conversation outcomes and engagement metrics. It is specifically designed to support the development of reinforcement learning approaches for sales conversion prediction, as demonstrated in the SalesRLAgent framework.

Dataset Summary

• Language: English
• Task: Sales conversion prediction, conversation analysis
• Format: CSV with embeddings and metadata

Dataset Structure

Data Fields

• company_id: Unique identifier for the SaaS company
• company_name: Name of the SaaS company
• product_name: Name of the SaaS product
• product_type: Industry/category of the product
• conversation_id: Unique identifier for each conversation
• scenario: JSON containing conversation scenario details
• conversation: JSON array of conversation messages
• full_text: Complete conversation text
• outcome: Binary outcome (0: no conversion, 1: conversion)
• conversation_length: Number of messages in the conversation
• customer_engagement: Customer engagement score (0-1)
• sales_effectiveness: Sales rep effectiveness score (0-1)
• probability_trajectory: JSON showing conversion probability at each turn
• conversation_style: Style of the conversation (e.g., casual_friendly, direct_professional)
• conversation_flow: Flow pattern of the conversation
• communication_channel: Channel used (e.g., email, phone, chat)
• embedding_0 to embedding_3071: 3072-dimensional embedding vector

Data Splits

This dataset is provided as a single train set. Users should create their own train/validation/test splits as needed.

Uses

Direct Use

• Training sales conversation prediction models
• Analyzing sales conversation patterns
• Training reinforcement learning agents for sales optimization
• Fine-tuning language models for sales-specific tasks

Out-of-Scope Use

• Real customer data analysis (this is synthetic data)
• Making business decisions without validation on real data
• Direct deployment without domain adaptation

Dataset Creation

Curation Rationale

This dataset was created to provide a large-scale resource for training machine learning models on sales conversation analysis, particularly for SaaS companies. The synthetic nature allows for diverse scenarios while protecting privacy.

Source Data

The dataset is entirely synthetic, generated using Azure OpenAI's GPT-4 model with carefully crafted prompts to ensure realistic and diverse conversations. The methodology is based on the approach described in "SalesRLAgent: A Reinforcement Learning Approach for Real-Time Sales Conversion Prediction and Optimization" (arXiv:2503.23303).

Annotations

The dataset includes automatic annotations for:

• Conversation outcomes
• Engagement metrics
• Sales effectiveness scores
• Conversion probability trajectories

Considerations for Using the Data

Social Impact

This dataset is intended for research and development of sales assistance tools. Users should be aware that:

• The conversations are synthetic and may not fully capture real-world nuances
• Models trained on this data should be evaluated on real conversations before deployment
• The dataset should not be used to replace human judgment in sales processes

Known Limitations

• Synthetic nature may not capture all real-world conversation patterns
• Limited to English language
• Focused on B2B SaaS scenarios
• May not generalize to all industries or sales contexts

Additional Information

Citation

If you use this dataset, please cite:

@article{nandakishor2025salesrlagent,
  title={SalesRLAgent: A Reinforcement Learning Approach for Real-Time Sales Conversion Prediction and Optimization},
  author={Nandakishor M},
  journal={arXiv preprint arXiv:2503.23303},
  year={2025},
  url={https://arxiv.org/abs/2503.23303}
}

Contact

For questions or issues with the dataset, please open an issue in the dataset repository.