As the festival season in India draws near, e-commerce and retail companies are expected to prioritise artificial intelligence (AI) in their strategies. As pilot programmes transition to full-scale implementation, the next three months will test whether autonomous AI agents can effectively manage demand on a large scale.

This festive period also comes against the backdrop of goods and service tax (GST) rationalisation on consumer goods and easing income tax, measures expected to boost consumer spending on durables, electronics and personal care items. To capture this demand, companies are rushing to adjust supply chains, ramp up production, tweak prices, and roll out new products while relying heavily on AI to deliver real-time results.

According to a survey by Shopify, 73% of consumers say AI improves their shopping experience, and the global AI-powered ecommerce market is expected to reach $8.65 billion in 2025. Another study by US-based SellersCommerce showed that 33% of ecommerce enterprises will include agentic AI by 2024, up from only 1% today.

Indian players warming up to the idea

Beauty and fashion retailer Nykaa said it has more than 40 GenAI initiatives active across customer experience, technology and internal productivity.

“This festive season we are doubling down on the transformative potential of AI,” a Nykaa spokesperson told ET. With these initiatives, the company expects a 30% gain in productivity this year, the person said, adding that the company expects AI to write nearly half its software code by the end of FY26.

“The results so far have been encouraging. We’re seeing measurable improvements in search relevance, personalisation, and product recommendations, which directly translate into higher conversion, average order value, and engagement metrics,” the Nykaa spokesperson said.

On the front-end, Nykaa is reimagining product discovery by levelling-up search algorithms. The company will soon launch Nykaa Muse, an AI-powered personal stylist.

Fashion and lifestyle ecommerce firm Myntra is similarly using AI to fight “choice overload” during its Big Fashion Festival (BFF). “Fashion is deeply personal and celebratory, and with BFF around the corner, AI helps us bring that individuality alive,” said Lakshminarayan Swaminathan, vice president, product management & design, Myntra. Swaminathan said GenAI also powers tasks like auto-tagging catalogues, styling suggestions and trend mapping, “significantly improving both speed and accuracy”.

Myntra is also using AI to push the most relevant content as influencer-led marketing scales up on the platform, he said.

Supply chain is often the biggest stress point during festive season peaks. Logistics provider Delhivery said it has invested in real-time tracking and agent-driven demand forecasting that anticipates spikes at the pin-code level. “Customers today demand visibility platforms. Live tracking and proactive communication are no longer just desired; they are essential requirements,” said Arun Bagavathi, senior vice president at Delhivery.

“AI in retail used to mean time-series forecasting, essentially looking at historical data and predicting what will happen in three months,” said Prem Bhatia, co-founder and CEO of analytics company Graas.ai.

That, he said, “is outdated, especially in quick commerce, where platforms are buying inventory twice a week and consumer demand shifts daily at the PIN code level”.

Bhatia said AI agents ingest live data and act upon it in real time.

Gurgaon-based Hostbooks said an ice-cream brand serving 4,500 retailers was able to cut reporting errors by 80%, track stock movement in real time, and close distributor accounts faster.

“The result was quicker replenishment during the festive rush, fewer lost sales due to stock-outs, and more satisfied customers, all while saving valuable time, manpower and working capital,” explained Kumar Somya Pandey, chief strategy officer at HostBooks.

However, companies are treading the automation wave cautiously as risks of hallucinations and errors persist. Generative AI systems have been known to produce misleading outputs or make up information. In an e-commerce context, this could mean AI-generated product descriptions that exaggerate claims, or shopping assistants recommending out-of-stock or irrelevant items. In 2023, for instance, Amazon faced backlash when its AI-generated review summaries misrepresented customer feedback.

Over-reliance on automation during a period of extreme demand could expose companies to glitches, supply chain bottlenecks, or misaligned recommendations.

Another risk is algorithmic bias. Personalisation engines trained on skewed data may underserve rural buyers or over-prioritise high-margin products, alienating sections of consumers. Studies have shown how biased AI recommendation systems could disproportionately promote private labels over popular third-party products.