SERVICES
Data and Analytics
Harness the Power of Data and Analytics for better Business Outcomes
Highlights
- Organizations need to determine their technology and change management readiness when they begin their AI adoption journey.
- GenAI value realization helps business leaders understand the impact of technological initiatives within organization processes, whether business-specific or internal-focused.
- Organizations need to focus on a long-term AI strategy to ensure that the initiatives undertaken are not point solutions, rather they contribute to the overall business value delivery.
On this page
GenAI value realization
The use cases for generative AI (GenAI) adoption in companies are evolving rapidly.
Initial GenAI applications, such as summarization, translation, image generation, or co-pilot initiatives, often yield rapid benefits over a short timeframe. However, these pilot projects typically lack scalability and lose momentum as the novelty wears off.
As organizations advance in GenAI maturity, they transition to more complex and data-driven applications leveraging retrieval-augmented generation (RAG) techniques for conversational AI, question answering, synthetic data generation (artificial data that mimics real world data), and data visualization. The potential value derived from these initiatives is considerably high although the investment size significantly increases as well.
Yet, unlocking the full potential of GenAI requires considerable investment and continuous maintenance. High computing costs and risk mitigation expenses also imperil the return on investment (ROI). The TCS AI for Business Study Key Findings Report 2024 reveals that out of 1,300 senior leaders from 24 countries surveyed, 72% of business executives struggle to measure the success of their AI implementations effectively, making it challenging to secure funding for more advanced AI projects.
Factoring the cost
Realizing value from GenAI initiatives requires distinct evaluation metrics which are different from traditional KPIs.
This is because the nature of output moves from deterministic to more probabilistic. Proving the effectiveness of a solution takes more time and effort and hence higher investments.
The pricing models offered by GenAI providers are diverse, ranging from license or subscriptions, tokens, API calls, and per-user or feature-based frameworks. Add to it the price of infrastructure, performance, scalability, integration, data processing and transfer, and ongoing support and the cost shoots up considerably.
There are also indirect costs related to AI ethics and legal formalities for avoiding risks related to data privacy, copyright, and other regulatory requirements. Bias reduction and explainability are subjects an organization needs to have suitable roles appointed for the purpose. Such costs need to be incurred on a continuous basis and shared between initiatives.
Getting the value
Businesses tend to evolve over time and so does the value of outcome from an AI initiative.
In GenAI-driven solutions, the real value of outcome may fluctuate, requiring periodic re-evaluation due to the evolving nature of the underlying model. Future iterations of the model will necessitate retraining with new data. Assigning a monetary value to desired outcomes is crucial yet challenging and hence some kind of estimation model needs to be used.
Typical outcome measures that can be readily used for GenAI use cases are value chain KPIs (customer satisfaction, turnaround time), operational improvement KPIs related to productivity or efficiency, specific financial KPIs related to cost savings or EBIT impact, and risk mitigation KPIs (detection of bias, redaction of personal information) related to regulatory compliance.
To maintain parity, the outcome measure needs to be converted into monetary equivalent so that it is possible to calculate an expected return. This may not be obvious, and logical estimates need to be drawn with the help of business and finance teams for further scrutiny and auditing.
Peek into the future
As organizations start their journey towards GenAI adoption, they need to think through the readiness, both from technology and change management perspective.
GenAI use cases usually have people and business process impact, hence careful consideration is needed for selecting the right opportunity. Given the significant challenges from legal and regulatory perspective, GenAI adoption and integration is complicated and risk prone.
Value realization is beyond evaluating GenAI model performance against accuracy, error rate, latency, or safety. It is a tool in the hands of senior management to understand the effectiveness of a technological initiative within organization processes, whether business-specific or internal-focused. Delaying its incorporation within the operating model would only cause issues later in the transformation journey.
