The transformation of India’s FMCG sector is no longer a story of distribution muscle alone. It is increasingly a story of data, of signals, predictions, and decisions made in milliseconds rather than months. From the outside, it may still look like a business driven by scale, but inside companies like Kellanova, the operating fabric is being quietly rewritten by AI and machine learning.
Suchindra Khaidem, Head of Data & Analytics, AMEA and Director, IT, Kellanova South Asia places this shift in a longer continuum rather than as a sudden disruption. “In India, for close to a decade, we have been in the journey of secondary automation across our distributor network and field sales force, making digital adoption a core enabler of how we execute, scale, and make decisions,” he says.
What has changed now is not just the presence of data, but its quality and usability. “The availability of high-quality data across our internal systems and external platforms forms the foundation of our AI and machine-learning initiatives,” he explains, adding that the company is deliberate in how it approaches this shift. “We take a focused approach to AI, starting with small, business-led pilots. Confidence in AI outputs is critical, and once established, we scale rapidly with clear scope and value-driven KPIs.”
This philosophy, start small, prove value, scale fast, has shaped how AI is embedded into the company’s execution engine. At Kellanova, these capabilities are not standalone experiments; they are integrated into what Khaidem calls the “Every Day Great Execution (EDGE)” framework. “These AI and ML initiatives are embedded into our EDGE framework for sales teams, becoming part of how we work,” he notes.
Rewiring the sales engine
India’s FMCG market is notoriously complex, spanning fragmented general trade networks, organised modern retail, and fast-scaling digital commerce. Rather than forcing a one-size-fits-all model, Kellanova has chosen to tailor its AI interventions to each channel’s unique dynamics.
“The three channels, general trade, modern trade, and digital commerce, are the major channels for FMCG companies in India. And I am happy to say that we have AI models at play across all three, with varying levels of maturity and success,” Khaidem says.
In general trade, where human judgment has historically driven decisions, the company has deployed a “Suggestive Selling” engine that nudges sales representatives in real time. “This is a machine learning-based algorithm which guides the sales representatives with must-sell, cross-sell and focus SKUs based on historical transactions,” he explains.
The impact is not just operational, it is structural. “The ML output delivers consistent execution at scale and enables quicker onboarding by reducing dependence on individual judgement,” he adds.
What began as a pilot quickly demonstrated measurable gains. “Following strong results the initiative was scaled nationally… delivering a 200-basis-point incremental sales uplift in top outlets, alongside assortment expansion of 2% in urban towns and 5% in smaller towns.”
Meanwhile, in digital commerce, the emphasis shifts to precision. Trade promotion optimisation models analyse transaction-level data to guide investments, while in modern trade, computer vision is transforming store-level execution.
Seeing the shelf, differently
For decades, shelf visibility in retail stores depended on manual checks, merchandisers physically counting products and estimating share of shelf. That process is now being automated with image recognition.
“We have deployed image recognition at scale across our modern trade stores to enhance in-store execution and visibility,” says Khaidem.
The real breakthrough lies not just in automation, but in adoption. “This process has now been seamlessly automated without altering existing ways of working, which has been critical to driving rapid digital adoption on the ground,” he explains.
The system continuously learns from new images, including those of competitor products, creating a dynamic and increasingly accurate view of shelf performance. The results are tangible, he says “We scaled up improving productivity of 30 minutes per store… while we get real-time Share of Shelf and On-Shelf Availability. A 0.9% increase in OSA resulted in 1.8% sales uplift during the pilot.”
Rethinking trade spend
If sales execution is one side of the FMCG equation, trade promotions are the other, and often the more expensive one. Here too, AI is reshaping decision-making.
“Trade promotions are one of the largest and most complex investments in FMCG, making them a key focus for AI-led optimisation,” Khaidem says.
Instead of relying on retrospective analysis, Kellanova is moving toward predictive models that guide investments at a granular level. “Our models guide investments at a brand and SKU level, shifting from manual analysis to predictive, outcome-driven decisions,” he explains.
The gains are not just financial but operational. “This helps prioritise promotions that deliver sustainable value rather than short-term volume spikes… freeing up 20% of the team’s efforts,” he notes, adding that the company has already “redeployed 15% of our spending across 20 top SKUs using this model.”
The same intelligence is extending into marketing innovation. During the launch of a key product, “we used ML and transliteration to create over 6,000 personalised retailer video ads, reaching over 1 million consumers,” he says, an example of how AI is reshaping engagement at scale.
The factory floor awakens
While front-end transformation often grabs headlines, Kellanova’s AI journey is also unfolding on the factory floor, albeit at a different pace.
“We have initiated our smart factory journey in India with IoT deployments across our two manufacturing sites,Taloja and Sri City,” Khaidem says.
The focus so far has been foundational, connecting machines, capturing real-time data, and building visibility. “Our first priority has been to establish a strong IT-OT foundation, enabling machines with sensors. This data is surfaced through dashboards used by factory and supply chain teams,” he points out.
Unlike sales and marketing use cases, the transformation here is gradual, but deliberate. “The journey is inherently more gradual compared to front-end digital initiatives, [but] we are steadily scaling deployments… embedding it into decision-making cycles, and creating continuous feedback loops for operational improvement.”
Scaling AI beyond pilots
Across industries, one of the biggest challenges is not building AI pilots, but scaling them. Khaidem is clear that technology alone is not the answer.
“Scaling AI requires the right mix of cultural alignment, strong data foundations, and scalable architecture,” he says.
At Kellanova, trust is the starting point. “We build trust in AI outputs through close collaboration with business teams early on, which helps drive adoption and manage change,” he explains.
Equally important is focus. “We prioritise a few high-impact use cases to ensure focus and scale,” he adds, highlighting that some of these solutions are now expanding beyond India to global markets.
But perhaps the most telling insight is cultural. “Ultimately, scaling AI is as much about mindset and continuous learning as it is about technology,” Khaidem concludes.
A quiet reinvention
What emerges from Kellanova’s journey is not a story of disruption for its own sake, but of disciplined reinvention. AI is not replacing the fundamentals of FMCG, it is refining them, sharpening execution, and making decisions more precisely.
The industry may still be defined by scale, but increasingly, it is being differentiated by intelligence. And as companies like Kellanova demonstrate, the real competitive advantage lies not in adopting AI everywhere, but in deploying it where it matters most, and making it work.