How to Measure the ROI of GenAI Investments?

Introduction

Generative AI is experiencing an incredible boom, and it’s no longer just a tech-centric topic. It has caught the eye of top business leaders and is now a tool in the C-suite’s arsenal. As organizations deploy Generative AI in their workflows, it is crucial for them to evaluate if this technology is delivering the promised results. In this article, we’ll understand how organizations can go about calculating the return on investment (ROI) that GenAI can bring to their businesses. We’ll explore the key factors to consider when measuring GenAI impact, the steps to calculate ROI, and the challenges companies might face along the way. 

Overview

• Explore why is it important to measure ROI of GenAI Projects
• Get to know the factors to consider while measuring the ROI of GenAI investments.
• Learn how to measure the ROI of GenAI Projects.

GenAI’s Role in Business Transformation

Why is it important to measure the ROI of GenAI investments? Before answering this question, Let me first take you through the findings from some recent surveys:

• According to a McKinsey report, “The state of AI in 2023: Generative AI’s breakout year“, one-third of surveyed companies use generative AI tools regularly in at least one area of their business. Additionally, 40% of companies are planning to boost their overall AI investment.
• In another survey of Lucidworks, over 2,500 business leaders are involved in generative AI decision-making, indicating the rapid growth of generative AI adoption in 2024. However, despite this growth, the survey found that only 63% of companies plan to increase their AI investments in 2024, down from 93% in 2023.

These reports show that while some companies rapidly adopted Generative AI in 2023, the pace slowed down in 2024. There are several reasons behind the slower pace such as high implementation costs, data security, and doubts on the accuracy of AI-generated content.

In particular, high implementation costs have made companies evaluate whether their Gen AI investments are yielding the expected results. This is why measuring the impact of generative AI on business is crucial—and the simplest way to do so is by assessing the return on investment (ROI).

Alos Read: Beyond the Buzz: Exploring the Practical Applications of Generative AI in Industries

What is the ROI of GenAI Investments?

Return on investment (ROI) is a method to calculate the financial benefit of a business from its projects. In case of GenAI projects, you can find it by just subtracting the cost of setting up and maintaining the GenAI systems from the revenue generated.

By measuring ROI, Businesses can clearly understand how effective and valuable their Gen AI investments are. This insight helps them evaluate whether their AI initiatives are worth the effort and money.

Factors Affecting ROI of Generative AI Investments

Generative AI is not a cost -effective solution for all your problems. Its successful implementation depends on various factors. It’s important to consider these factors to calculate the ROI of your GenAI investments:  

1. Determining specific goal
2. Measuring Key Metrics
3. Recognizing the required investment 
4. Evaluate the current scenario
5. Analyzing the possible returns

Let’s explore each one of these in detail.

1. Determining Specific Goal

The first step is to have a clear vision of how your generative AI project can help align with a company’s overall goals and business strategy. 

For example:

• In the finance sector, a clear goal could be to improve fraud detection by 15% through analysing transaction patterns.
• In the fashion industry, GenAI can accelerate the product design process and streamline new product launches. Here, a specific goal could be to increase product launch frequency to one product per month.
• Similarly for the customer support function, a specific goal could be to address 10% more customer queries.

2. Measure Key Metrics

Why is it important to measure key metrics? Metrics help companies stay aligned with their goals and objectives. 

For example: 

• Some key metrics to improve fraud detection by 15% include fraud detection rate and operational efficiency (measured by time saved in investigating fraud cases) 
• For product design and launch in the fashion industry, key metrics are revenue from new products, design and development costs, and an increase in the number of products launched in a year.
• Some key metrics for a specific goal of addressing 10% more customer queries are costs, an increase in the number of queries handled (efficiency gain), customer satisfaction rating, and repeat sales.

3. Recognize the investment required

It’s essential to calculate the expenses involved in setting up and operating generative AI. These costs include several smaller components, such as:

• Tools usage cost: This covers expenses related to tools such as cloud infrastructure costs, license, hardware, maintenance, data collection and model training.
• Learning & Development: It includes costs associated with employee training to ensure the effective use of GenAI tools.
• Third-Party Advisory Costs: This includes fees paid to external consultants for advisory services related to GenAI tools.

For example: Suppose an e-commerce company wants to set up a GenAI-powered chatbot to enhance customer support. The estimated set-up costs would include: 

• Cloud infrastructure: Renting cloud-based servers from AWS or Azure. Let’s assume the cost is $6.76/hour, amounting $20,000/month for a server with sufficient resources.
• GenAI tool licensing: Subscribing to GenAI platforms like OpenAI API or Hugging Face might cost around $15,000/month.
• Data and model costs: Assume data collection, data cleaning and model retraining might cost approximately $10,000/month.
• Learning and Development: Let employee hiring and training overall expenses be around $15,000/month.

All these would combine to have an overall expense to be $60,000/month ,or ($60,000/month*12) $720,000/year.

4. Evaluate the Current Scenario

Let’s continue to analyse the problem statement to build a GenAI-powered chatbot to enhance customer support. Analyzing the current situation is a critical part of business. Before the use of GenAI, when the company handled customer queries manually, the costs were as follows:

• Hiring staff: This includes recruitment expenses, internal training, annual salaries, and employee benefits such as health insurance. The total cost was estimated at around $50,000/month, or $600,000/year.
• Customer Service infrastructure and software: This covers hardware costs, office rent for employees, live chat software, and CRM licenses (e.g., Zendesk or Salesforce). The total investment for infrastructure and CRM amounted to $50,000/year.

The total cost for handling customer queries manually was $650,000 per year.

Suppose in the earlier setup, 21,000 queries per month were handled.

Assuming 10% of these queries result in repeat sales, with an average sale value of $350, this generates an estimated revenue of: 21000*10%*$350 = $735,000

5. Analyze the Possible Returns

The next step is to consider all the possible advantages of adopting generative AI in your business. These may include increased revenue through automation, enhanced productivity, reduced errors, and improved customer satisfaction and interaction.

For example: After implementing a GenAI-powered chatbot, the e-commerce company foresaw that GenAI chatbot can handle 19% more customer inquiries. As a result, the total queries managed monthly would increase from 21,000 to 25,000.

With the use of Gen-AI powered chatbot, the customer inquiries can be resolved instantly, which would significantly enhance the customer experience. 

Let’s assume that the repeat business stays at 10%. The total revenue would increase to = 25000*10%*350 = $875,000

Now that you know the investments and the possible returns, you’re ready to calculate ROI for your GenAI investments.

Measure ROI for GenAI Investment

To calculate the ROI, you can use the ROI formula:

Here, in the e-commerce example, GenAI setup and operational cost is $720,000 and Revenue generated from GenAI is $875,000.

As you can see, the ROI has improved from 13.07% to  21.52% after using Gen AI in this scenario. This significant improvement in ROI is due to costs and more efficient query resolution rates.

Challenges while implementing ROI of GenAI

With the high ROI using GenAI, you can think about implementing it in your business but still it comes with some challenges. It’s important to consider these when it comes to implementing GenAI projects:

• Unreliable Results: GenAI models may produce false or hallucinated results, leading to rework, which adds cost, time, and effort—ultimately affecting ROI.
• Lack of Skills: Generative AI is a rapidly evolving field, and your workforce must keep pace with the latest developments. However, finding employees with the necessary skills can be challenging.
• Complex technology:Customising AI models is complex and requires technical expertise. Data preparation, algorithm design, and deployments require costly hardware, resulting in high computational expenses.
• Maintenance: GenAI models need constant maintenance, continuous monitoring, and updates. These ongoing costs are often ignored, causing incomplete ROI calculations.

Also Read: How Generative AI Is Reshaping Business, Healthcare, and the Arts?

Conclusion

Generative AI is becoming increasingly popular in businesses. However, from a business perspective, it is essential to measure the return on investment (ROI) of their GenAI projects to understand their impact. ROI analysis helps businesses see how effective GenAI is for their business and compare its costs and benefits. 

Frequently Asked Questions