GCC isn't Gen AI capability center

India is home to nearly 50% of the GCCs of the world, with a market size expected to reach USD 110 billion by 2030.

It is for the first time in the history of technology, that a technological innovation itself is being used to sell it. Let me simplify it, we never used cloud to position cloud in front of customers, but Gen AI is currently leveraged aggressively to compose, and present appealing messages in front of the key personas about the value, benefit and impact of Gen AI.

It is natural for product and services companies, to embrace a new technological wave ahead of time, and a make the most of it by positioning it nicely across their prospects. Considering the growth and use cases presented by GCCs, a similar trend is visible.

There is no doubt about the fact that Gen AI will create a lot of value, for specific use cases. But if product companies, services firms try to force-fit it across use cases, where the path to that value is extremely long, or where the risk implementing Gen AI is higher, than the potential benefits it can deliver- is certainly an unethical and unfavourable approach. Unfortunately, its visible the myopic sales focused companies have not learned their lessons from the past technological waves and hypes.

Value creation or value capture?

Gen AI identifies patterns in data, across an extremely large data set to create new content- text, images, video, code, etc. This could be extremely valuable from a GCC perspective, considering the scope of activities GCCs support like marketing, finance, procurement, IT services, HRMS, etc. A lot will require back-and-forth interactions, a lot will be repetitive in nature.

However, just because one is able to create value, doesn't necessarily mean - value is getting captured as well. Value creation refers to the process of generating additional benefits or utility for the business or its customers. Whereas, Value capture is more about converting the created value into profit for the company.

For example, a GCC supporting finance function of a large enterprise leverages generative AI for vendor helpdesk function, and ultimately reduces time responding to messages, from vendors who want to know the status of payment for their pending invoices. This can help optimize the costs associated with putting the first line helpdesk resources, which creates value. But if everytime Gen AI is used to frame a response, that response and the associated prompts, feed the Gen AI data set as well, making it richer but at the same time losing the same comeptitive advantage it was expected to create.

Metaphorically, think about a scenario you find a lamp, you rub it and genie pops out. You believe you have acquired a superpower, but what if - a similar lamp is discovered at the same time by everyone in town, do you have an advantage, or its getting more complex.

The first-mover advantage might be real temporarily, but the real test will happen when one Gen AI deployment, is able to capture value better than another deployment of the same Gen AI model. In a Gen AI led ecosystem of competing players, the benefit doesn't come from one's own efforts to advance, but from the prior efforts of others to do the same.

The original mandate against the latest hype

The original mandate for GCC's was cost reduction, which over a period time transformed into a center of excellence providing strategic value to different functions of the enterprise, as it has become a 'one-place' repository of data and intelligence housed in a business. A strong push to make everything 'Gen AI' and 'AI First', without any real consideration to fitment across different use cases, will not only burn resources, but will also give Gen AI a bad name.

It is important to make an assessment of the different use cases where Gen AI can help, and where one must not think about putting Gen AI. A much better understanding of how chained prompts work, and how the token economy functions across LLM products, will help the executives in determining what use case is best for them. Putting Gen AI investment, where a small RPA was best, would make the business case negative and putting it across a process, where the scale is not commensurate, or the process is not standardized enough would mean that the implementation will fail to generate the benefit.

Of course, if put through some propreitary database, Generative AI can be lever of driving tremendous value, but how many products you see around- committing to that promise?

Trying to push Gen AI to an extent a Global Capability Center starts sounding like a Gen AI capability center is not going to help with either value generation or value capture.

When you look at actual use cases, and the kind of disclaimers Gen AI product companies are giving after securing billions of dollars in investments, you would agree with me - that ultimately Gen AI is a god of small things.

The push towards a certain technological innovation has to balance with value creation and capture. The goals and objectives of GCCs are different and important.

I love the technology marvel Gen AI is, and I really want it to mature to a stage where the real value across the landscape is visible and achieved.

Do let me know what you think across this line of thought.

-Mohit Sharma

Written with absolute respect and reverence to all innovators and absolutely zero-prejudice for any product or companies.