# Evaluate and define your generative AI business use case

This document helps you define an AI business use case by following a business
value-driven decision approach.

> [!NOTE]
> **Note:** In this document, *traditional AI* refers to AI capabilities and use cases that might not require employing generative AI capabilities, like some [classification](https://developers.google.com/machine-learning/crash-course/classification/video-lecture) and [predictive](https://cloud.google.com/learn/what-is-predictive-analytics) AI use cases. For information about AI capabilities, see [When to use generative AI or traditional AI](https://docs.cloud.google.com/docs/ai-ml/generative-ai/generative-ai-or-traditional-ai.md.txt).

Generative AI and traditional AI solutions are powerful tools, but they should
always support your business goals, and they shouldn't exist in isolation. To
create successful generative AI or traditional AI solutions, begin by clearly
identifying the specific measurable business goals or needs that you want to
address. Then work backward from the business outcomes that you want--such as
increased employee efficiency or enhanced customer satisfaction--to make sure
that the solution directly contributes to your business goals.

To define your generative AI or traditional AI use case with a business-value
focus, use the following simplified decision process:

1. **Business goal and success criteria** : Identify measurable business goals.
   - Focus on the [business goal](https://developers.google.com/machine-learning/problem-framing/problem#state_the_goal) and value to be achieved, such as increasing efficiency and productivity, cost reduction, enhancing customer experiences, and competitive advantage.
   - Clarify how the business plans to measure the success of the identified goals and objectives. Return on Investment (ROI) is one of the key measures of AI project success. ROI can be measured through several metrics like the following:
     - **Direct financial gains**: Increased revenue or reduced costs.
     - **Operational efficiency**: Faster time-to-market or quicker issue resolution.
     - **Customer experience**: Increased satisfaction scores or improved retention.
   - Identify any potential business constraints and considerations, such as ensuring that security and privacy aspects meet specific industry compliances or country regulatory requirements.
2. **Type of AI/ML**: Determine whether AI/ML is the right approach for
   solving your business problem or achieving the identified goal.

   Decide whether the identified business expectation requires generative AI,
   other types of AI, or whether it doesn't require AI to achieve it. For
   more information, see
   [Identify the output you need](https://developers.google.com/machine-learning/problem-framing/ml-framing#identify_the_output_you_need)
   in "Framing an ML problem."
3. **User experience expectation**: Identify the end users of the use case
   and how they might interact with the generative AI- or traditional
   AI-powered application or service. Consider what the user expectations or
   preferences might be.

4. **Business driven and user-centric AI solution**: Connect the optimal
   generative AI or traditional AI technology use case with measurable
   business requirements, the organization's executive priorities, and user
   expectations. Consider the following:

   - How the business can drive optimized efficiency and productivity by using generative AI or traditional AI to achieve more outcomes at a faster pace, and with less operational complexity or with reduced efforts (and potentially with cost savings).
   - How the business can drive enhanced customer or product experience by using generative AI or traditional AI.
   - How you can create business value in an innovative way by using generative AI or traditional AI:
     - Analyze your existing business offerings and capabilities to identify areas where generative AI or traditional AI can improve your existing solutions, enhance creativity, or enable you to explore new possibilities.
     - Understand how AI can enable innovative enhancements that set your business apart. Generative AI can help create differentiated capabilities and value, help you go beyond solving immediate business pain points, and explore ways to boost your existing offerings.
     - Prioritize using technology to enhance business capabilities that align with the organization's priority goals.
5. **Business process change**: Identify the changes that the business has
   to make to existing processes or workflows to adapt to the generative AI or
   traditional AI use case.

   Consider how the AI solution will change the way that employees
   or customers interact with the company's systems and workflows, such as
   through a mobile app or customer support chatbot. These interactions
   might require backend processes to be changed or reinvented in order to
   leverage AI capabilities like workflow automation and to help the
   business realize the benefits of AI.

## Generative AI business use case example

The following sections provide a simplified example that demonstrates how to
identify and connect measurable business needs and expectations to impactful
generative AI business use cases.

### Business problem statement

In this scenario, customer support service teams are overloaded with a high
volume of repetitive inquiries, manual tickets management, and constant support
emails communication. The overload strains resources, increases agents' working
hours, and slows resolution times, which results in decreased customer
satisfaction and retention.

### Potential areas of optimization with measurable business value

The following are examples of the possible measurable business values that can
be achieved by using a technology solution (a chatbot) that's powered by
generative AI capabilities to address the preceding business challenges. Based
on their business model and priorities, the business might consider some or all
of these measurable targets.

- **Enhance customer support efficiency** : Reduce support costs and streamline agent workflows. Measurable success criteria include the following:
  - Percentage decrease in customer support operational costs over a defined period (such as quarterly).
  - Percentage increase in the volume of customer inquiries handled by the chatbot.
  - Average reduction in agent working hours for repetitive tasks.
- **Optimize ticket resolution** : Improve resolution speed and increase the percentage of issues that are resolved directly by the chatbot. Measurable success criteria include the following:
  - Average decrease in time-to-resolution for inquiries that the chatbot handles.
  - Percentage of tickets resolved without human intervention.
  - Percentage decrease in the volume of tickets that must be escalated to the technical support team due to complexity.
  - Increase in first-contact resolution rate (issues solved in a single interaction).
  - Percentage increase in the volume of customer inquiries that the chatbot handles and resolves.
- **Enhance customer experience** : Boost customer satisfaction by offering responsiveness and personalized support that's available 24 hours per day. Measurable success criteria include the following:
  - Increase in customer satisfaction (CSAT) scores in surveys tied to chatbot usage.
  - Reduced average customer wait times for initial interaction.
  - Increase of issues solved in a single interaction.
  - Percentage of positive sentiment detected in chatbot conversations and feedback surveys.
  - Improved customer retention rate.
- **Support business operations growth** : Handle increase in customer demand without incurring linearly increasing costs or increase in wait times for initial customer interaction. Measurable success criteria include the following:
  - Ability to handle a specified percentage increase in support request volume without human intervention.
  - Maintain consistent CSAT scores and time-to-resolution during periods of high demand.
  - Maintain consistent customer wait times for initial interaction.

### Generative AI-powered solutions

**Conversational chatbot**: Generative AI-powered chatbots or virtual agents
offer a significant enhancement in personalization and natural, human-like
conversation. This is due to the ability of generative AI to understand complex
context, sentiment, and relationships within language. This ability leads to
more natural interaction, asking relevant questions, and providing tailored
recommendations for an improved user experience.

Generative AI abilities also help organizations to drive more work efficiencies
and productivity. In contrast, a traditional rule-based chatbot is commonly
limited to predefined keywords and intent patterns. Therefore, as conversational
patterns evolve or new questions arise, a rule-based chatbot requires additional
operational effort, for rule updates and refinements and intent training. For
this use case, generative AI chatbots provide the following benefits compared to
traditional rule-based chatbots:

- Generative AI-powered chatbot answers aren't limited to frequently asked questions (FAQs). The chatbot can find answers within large datasets from different sources like historical data of support cases, websites, product documentation, inventory, emails, and old chat conversations with resolution. It can also understand conversational queries and summarize complex information.
- Generative AI virtual agents synthesize information from all your data sources. This synthesis enables them to provide specific, reasoned, and actionable responses that are based on the data that you have provided and that are aligned with your business expectations.
- Generative AI interprets the complex language and nuances within a ticket. It can understand the full context of a customer's issue; a traditional AI chatbot primarily focuses on specific keywords.
- Generative AI chatbots provide the flexibility for customers to express themselves using their preferred method (text, voice, image), while the chatbot leverages all input to improve issue resolution. For example, customers can share photos of a damaged product during the chat conversation, and generative AI can combine the customer's description with the photo in order to help enhance diagnostic and troubleshooting accuracy.

**Case management and insight-generation workflow**: A Generative AI-powered
chatbot can automatically generate tickets from every interaction. The chatbot
utilizes generative AI capabilities to understand the urgency, sentiment
analysis, and complexity of the issue. These capabilities ensure that tickets
are prioritized effectively. The chatbot can interact with your ticketing system
in these ways:

- The generative AI chatbot interfaces directly with your support ticketing system to create and populate the support ticket with required information like the following:
  - Customer details
  - Technical issue categorization and priority
  - A full transcript of the conversation for context
  - Summarization of the core issue
- For new, complex issues, the chatbot can assign the ticket to the correct team with supporting context such as a summary of the issue and conversation.

## What's next

- Learn about how generative AI might apply to your use case in [Generative AI examples](https://cloud.google.com/use-cases/generative-ai).
- Learn more about the stages of developing a generative AI application and choose the best products and tools for your use case in [Build a generative AI application on Google Cloud](https://docs.cloud.google.com/docs/ai-ml/generative-ai).
- Assess your AI capabilities and create a roadmap to harness its potential with the [AI Readiness Workshop](https://docs.cloud.google.com/consulting/ai-readiness-workshop).