14 years ago, OGS’s India operation began with five people and a narrow mandate around data and digital marketing. It has since grown to about 7,000 people across India and the Philippines. The more difficult question now is what that scale means as AI begins to change the economics of knowledge work.
For Kiran Guruswami, Chief Operating Officer, OGS India and Philippines, the GCC model is moving beyond its traditional emphasis on execution and capacity. The expectation from India is increasingly about what it can contribute to the business.
“The expectation is far beyond just execution,” Guruswami says. “It’s no longer just about scale, but also about innovation and ideation.”
From data to a wider mandate
OGS began under the Annalect brand, with a five-member team working within Omnicom’s media agencies. Its initial mandate was to use data to make marketing decisions more evidence-based.
The early work centred on data pipelines, reporting, visualisation and insights. As more agencies and markets began using the operation, its remit expanded into technology and product development. A significant part of the engineering talent behind Omnicom’s Omni platform also came from the group Guruswami represents.
The scope subsequently widened beyond data and technology into advertising, public relations, healthcare advertising, branding, precision marketing and production.
“That five-member team over the years has now become 7,000 people across a wide variety of capabilities that continue to integrate,” Guruswami says.
India remains the largest of OGS’s three major hubs, alongside Colombia and the Philippines. The India operation has seven offices. Before the pandemic, it was concentrated in Bengaluru and Gurugram. The movement of employees back to their home cities during and after the pandemic led OGS to expand its presence into additional locations.
The distribution is not uniform. Bengaluru has a significant production presence, Mumbai has concentrations of strategy and creative talent, while Hyderabad is strong in technology, data and media. A business support services unit of about 500 people also operates from India.
The change is less about adding functions for their own sake than about broadening the work that can be undertaken from India. Guruswami points to employees seeking to move across disciplines, including data analysts interested in advertising strategy, as evidence of that shift.
Cost is no longer the whole argument
India’s cost advantage remains part of the GCC equation, but Guruswami does not see it as sufficient on its own.
“If it’s just cost arbitrage, it has always been in play and I think it will continue to be,” he points out. “But the expectations have really risen.”
The change is in what parent organisations expect their Indian operations to contribute. Execution remains part of the remit, but innovation, ideation and the ability to move beyond the original brief are increasingly being factored into the relationship.
Guruswami also points to the breadth of India’s talent pool. People entering advertising, media and related functions do not necessarily come from conventional backgrounds. He cites his own engineering background as an example of how technical skills can find applications outside their original discipline.
The implication for the GCC model is straightforward. An operation cannot rely indefinitely on the availability of people at scale if the work itself is becoming more complex.
AI is being applied selectively
Artificial intelligence adds another variable to that equation.
Guruswami’s approach is notably cautious. He does not describe AI as a way to bypass inefficient processes. Instead, he sees it as a layer that can be applied once the underlying workflow is sufficiently mature.
“You throw AI to an inefficient process or an inefficient framework into it, or even an ill-trained data set, it will still be garbage,” he says.
At OGS, AI is being used in areas including audience discovery, content recommendations, concept testing and media optimisation. The technology is intended to accelerate parts of the process, while human judgement remains necessary.
“You still need a proper person who understands the channel,” Guruswami says, adding that it helps augment and fast-track that journey.
There is some evidence of efficiency gains, although the numbers are limited. In certain production-related use cases, Guruswami says OGS has seen improvements of around 30% in cycle time. He does not describe this as an organisation-wide productivity figure.
He is equally clear about the cost and effort involved in getting AI systems to work reliably. Building the networks, workflows and agents required for these applications takes time, and Guruswami avers the organisation is cautious about making large bets before the economics become clearer.
“We are also conservative about making bigger bets and bigger stories coming out of it,” he says. “We are careful about this.”
That caution extends to consumption. OGS has a governance framework around AI usage, reflecting the concern that widespread access without discipline could create costs without corresponding value.
“Just because you can use AI for everything doesn’t mean you use AI for everything,” Guruswami quips.
The headcount question
The more fundamental issue is what AI means for the scale-led GCC model itself.
Guruswami does not expect the change to translate simply into smaller GCCs. But he acknowledges that the traditional equation is being tested.
“You can’t throw people to a problem or a solution anymore,” he says.
For GCC leaders, that makes workforce transition as important as technology adoption. Employees will need to acquire new skills and use technology to generate more value from existing resources.
It also changes the basis on which GCC performance is judged.
“The way how the GCCs are measured in performance is no longer just going to be on headcount. It’s going to be on impact,” Guruswami remarks.
That distinction is increasingly important as the traditional growth narrative of India’s GCC sector comes under pressure. A larger workforce is still useful where demand requires it, but headcount alone says little about the value being created.
OGS’s own trajectory provides the numerical contrast. The operation has moved from five people working largely on data and digital marketing to about 7,000 people across a much broader set of capabilities in 14 years.
The next phase will require a different measure of progress. The question is no longer simply how many people a GCC can employ or how much work it can absorb. It is what those teams can deliver as technology changes the amount and nature of human effort required.
For Guruswami, AI will eventually have to become part of that everyday equation rather than remain the preserve of a specialist function. OGS has an AI centre of excellence, but he sees adoption across the wider workforce as the more important objective.
“To me, AI is like Excel,” he says. “AI is like a tool that all of us, no matter what trade or what we do, everyone needs to start using it.”
For India’s GCC model, that could be the more consequential change. Scale built the first phase. The next phase will have to demonstrate what that scale delivers.