Retail Media emerges as AI-driven 'commerce intelligence,' marginalizing traditional digital spend

In 2025, artificial intelligence has moved from the margins of advertising operations into the core of how campaigns are designed, deployed, and evaluated. What was once positioned as a productivity enhancer or creative assistant is now functioning as an operational layer across creative production, media buying, measurement, and agency economics. The shift is not cosmetic. It reflects a deeper reordering of incentives, accountability frameworks, and budget governance within the advertising ecosystem.

Across India’s fast-scaling digital market, AI adoption is being driven less by regulation and more by measurable efficiency, cost compression, and platform consolidation. The result is an industry increasingly governed by data discipline rather than intuition alone.

From Creative Assistance to Algorithmic Execution

In 2025, AI-driven creative capabilities are no longer confined to ideation or mockups. They are embedded directly into live campaigns, producing and optimising creative assets at scale. According to Kalyan Ram Challapalli, Founder and Head of Strategy at WolfzHowl Global, AI is now powering fully personalised advertising across formats.

“AI is no longer just a helper for mockups or ideation. It is the engine behind fully personalised campaigns: generative video, micro-segment-specific creative, programmatic DCO, and 1:1 product personalisation,” he says. “The question is not what AI can do, It is whether agencies and brands have the courage to hand over the creative reins to algorithms.”

In India, this transition is most visible in retail media networks and short-video platforms, where controlled experimentation and attribution allow AI-led creative to be tested and scaled rapidly. These environments reduce the risk traditionally associated with delegating creative decisions to automated systems.

Vaishal Dalal, Co founder and Director at Excellent Publicity, notes that generative video and imagery tools are now routinely producing complete ad films, social assets, and display creatives from text prompts or reference images. AI-powered dynamic creative systems generate variations in real time, adapting visuals, messaging, and calls to action across segments without manual intervention. Automated localisation, mass customisation, and even AI-enabled outdoor formats with responsive billboards have moved into mainstream use.

These capabilities are no longer pilot-stage technologies. They are shaping how campaigns are executed across categories.

Creative Accountability Becomes a Data Governance Issue

The expansion of AI-generated creative has introduced a new accountability challenge. While output volume has increased dramatically, performance scrutiny has intensified.

“AI doesn’t just increase output; it exposes your creative intelligence,” Challapalli says. “Brands now face a choice: either test strategically at scale or drown in vanity variants.”

In the Indian market, he observes that many advertisers are producing large volumes of AI-generated permutations without corresponding measurement frameworks. Others, however, are pairing AI-driven creative with rigorous performance tracking, effectively running thousands of experiments per week and accelerating learning cycles that previously took years.

Dalal frames this evolution as a shift from content production to validation efficiency. AI tools are increasingly used for forecasting, pre-testing, and real-time optimisation, allowing creative performance to be evaluated against clearly defined KPIs. While human strategy and brand oversight remain central, AI is enforcing tighter feedback loops between creative decisions and measurable outcomes.

Challapalli cautions that this data-driven definition of creative accountability may not be permanent. “The provocative insight: creative accountability will soon be defined more by data discipline than by the quality of the initial idea. But this will fail too, and then brands will revert to a mix of the imperfection but raw attraction of instinct-led initial ideas and data.”

As A/B testing becomes inexpensive and ubiquitous, he suggests that brands may eventually rebalance toward fewer, more differentiated ideas refined through selective optimisation rather than exhaustive testing.

Media Budgets Shift Toward Measurable and Platform-Native Channels

AI-led targeting has also reshaped media allocation decisions. Unlike the EU or US, where privacy regulation has been the primary driver of change, India’s budget shifts are being led by performance measurability and platform efficiency.

“Budgets are moving faster than most marketers realise,” Challapalli says. “Privacy isn’t a constraint in India like in the EU/US — it’s an opportunity to consolidate spend into first-party and platform-native ecosystems.”

Retail media, connected TV, short-video platforms, and programmatic AI bidding are absorbing spend that once flowed to the open web. Channels that cannot provide deterministic attribution are increasingly marginalised.

“The real insight: the more measurable the channel, the more dollars will flow — and that means traditional TV and non-attributed digital are quietly becoming ‘luxury’ media rather than default,” he adds.

Dalal notes a similar trend, with budgets shifting toward connected TV, in-app environments, contextual targeting, and first-party data ecosystems where AI can deliver privacy-compliant optimisation. Rather than relying on third-party identifiers, advertisers are prioritising environments where deterministic or contextual signals can be combined with AI-driven budget allocation.

Both executives expect privacy-conscious budgeting to strengthen over time, driven by consumer awareness rather than immediate regulatory enforcement. Challapalli anticipates that Gen Z and Gen Alpha will play a key role in normalising privacy-led expectations across broader cohorts.

Retail Media Emerges as an AI-Led Commerce Layer

Retail media is one of the clearest beneficiaries of AI integration. However, its value proposition in 2025 extends beyond bid optimisation.

“AI in retail media isn’t just bid optimisation. It’s commerce intelligence,” Challapalli says. Inventory-aware creative, next-best-offer personalisation, and incrementality modelling are allowing brands to measure real sales lift with greater precision than traditional digital channels.

India’s major retail platforms offer deterministic purchase data, enabling AI systems to optimise for outcomes rather than proxies. Challapalli argues that brands not actively testing AI-optimised retail media campaigns are effectively transferring competitive advantage to rivals measuring performance in real time.

Dalal highlights similar gains from predictive bidding, automated budget recommendations, personalised creative generation, and improved attribution. AI-powered analytics are helping advertisers isolate incremental impact, strengthening confidence in retail media ROI.

At the same time, Challapalli flags emerging behavioural effects such as heightened cart parking, impulse-driven purchases, post-purchase dissonance, and higher return rates. While he expects measurable lift to continue for the next three to five years, particularly across Bharat markets, he believes corrective consumer behaviour will eventually emerge.

Agency Economics and the Repricing of Intelligence

AI is also forcing a reassessment of agency billing structures. Execution hours are increasingly commoditised, while strategic and technical intelligence is becoming the primary value driver.

“Execution hours are commoditised; intelligence hours — model building, prompt engineering, AI ops, DCO orchestration — are now premium,” Challapalli says.

In India, many agencies continue to rely on manpower-based retainers, but client expectations are shifting. Brands adopting AI-led in-house models are demanding outcome-linked and intelligence-driven pricing. Smaller agencies are using AI to reduce production costs, prompting clients to question traditional billing logic.

Dalal sees a gradual move toward hybrid and value-based models, where agencies are compensated for judgment, domain expertise, and measurable impact rather than time spent. While retainers persist, the definition of value is being renegotiated.

Performance, Storytelling, and AI-Governed Allocation

AI has not eliminated the distinction between performance and brand investment, but it has changed how the two are balanced. Challapalli notes significant budget shifts toward performance channels and connected TV, with DCO-driven personalisation revitalising performance marketing.

At the same time, leading brands continue to invest in storytelling on CTV and social video to build attention and equity, using AI to optimise timing, hooks, and regional relevance rather than to replace narrative.

The broader implication is structural. AI is transforming budget allocation from intuition-driven decision-making into a data-structured process. In India, this transition is occurring faster than in Western markets, driven by speed, cost sensitivity, and competitive pressure.

The risk, as Challapalli concludes, is that AI adoption without governance may generate more noise than impact. As AI becomes ubiquitous, creative governance and strategic restraint are likely to become as important as automation itself.