Ajit Varghese and Vivek Das on why Madison Media is rewriting the agency playbook

It should come as no surprise that ad agencies are reinventing themselves. Like old trees whose roots once thrived in familiar soil, agencies are discovering that the land beneath them has shifted. What once sustained growth — stable retainers, media margins, and campaign-led scale — are no longer enough. To survive, roots must spread wider and deeper, drawing strength from new sources: data, technology, commerce, and culture.

The air around them has changed too, shaped by consulting-led competition and sharper client expectations that prioritise measurable outcomes and business impact. In this altered ecosystem, growth is no longer guaranteed by size alone; it depends on how well agencies adapt their foundations to a changing environment.

For Madison Media, this recalibration is no longer optional. The agency’s unveiling of MbrAIn and its proprietary Growth Planning System (GPS) earlier this month marks a decisive step in that direction, signalling the shift to Madison 3.0.

The move, Ajit Varghese, Partner & Group CEO of Madison Media Group & Hiveminds, explains, was prompted by structural changes in the ecosystem that made incremental change feel inadequate.

“Three shifts told us,” says Varghese, “that tweaks won’t cut it.” The first was the fragmentation of consumer attention. “Consumers now jump between TV, OTT, social, commerce and chat in minutes,” he says, adding that planning each channel separately makes little sense when consumers don’t think in channels at all.

The second was the growing overlap between media and commerce. With the rise of retail media, quick commerce and D2C, brand-building and selling have become deeply intertwined. “Brand and performance budgets can no longer live in separate worlds,” Varghese notes.

The third was the democratisation of AI. With tools widely available, advantage no longer lies in access to technology but in how it is systematised. “The advantage is having a system that uses AI in a disciplined, theory-led way.”

As a result, Madison consciously moved away from a campaign-first mindset to a system-first approach, where campaigns are treated as moments within a longer growth story.

“Madison 3.0 is our answer to an ecosystem where ‘more of the same’ would simply make us irrelevant,” Varghese says.

What Madison 3.0 changes for clients

For clients, the most immediate change is in how business problems are framed and solved.

“The most visible difference is that clients now experience one integrated growth system, not a set of disconnected campaigns,” Varghese says.

Earlier, clients were often presented with parallel plans — traditional media on one side, digital and performance on another, ecommerce ideas elsewhere, and strategy decks that didn’t always mirror execution. Under Madison 3.0, that fragmentation is intended to be replaced by what Varghese describes as ‘a single GPS-driven growth narrative’ that begins with business ambition and barriers, not channels.

Media, content, channels and commerce are planned as one system of effects, with a defined role for every rupee spent. Planning is supported by MbrAIn, which Varghese says allows teams to move faster without losing rigour and, crucially, to explain the logic behind decisions. Reviews and dashboards now link media inputs to brand health and revenue, rather than stopping at impressions or clicks.

“In simple terms,” he says, Madison 3.0 is “less noise, more clarity; fewer decks, more decisions; fewer silos, more outcomes.”

Components of Madison 3.0

Positioned as a pivot from services to next-generation products, Madison 3.0 reflects a move away from simply buying media to architecting growth. At its core is MbrAIn, designed to function as an always-on strategic intelligence layer enabling explainable, auditable, intelligence-led decision-making, supported by GPS, a structured growth framework built for India’s complexity. Together, they are intended to turn planning into a living operating system that absorbs every brief and grows stronger with each cycle.

Overall, Madison 3.0 rests on five interconnected pillars. At the centre is GPS, followed by MbrAIn. The third pillar is a Content Studio that integrates creative craft with data and AI to build scalable, platform-native content systems. Underpinning this is Catalyst OS, which serves as the agency’s unified technology and data backbone, while the Commerce Performance Engine grounds growth decisions at the physical and digital shelf.

“Together,” Varghese says, “these pillars are our blueprint for what a modern agency should be: a growth system, not a buying shop.”

Chief Digital Officer, Vivek Das, traces the beginning of the transition back to a core insight: "India needed an operating system for growth, not just tools and dashboards."

Madison identified three gaps in the market: an abundance of data with few decisions systematically linked to it; widespread discussion of AI with little built specifically for Indian brands and journeys; and strong individual strategists with no way to scale their thinking across teams.

“GPS was built to be that operating system,” Das explains. “MbrAIn was built to be the strategic brain that runs it.”

Internally, this required cultural and operational shifts. Framework thinking and problem-solving were elevated alongside buying skills. Teams were rewarded for business outcomes and learning, not just delivery. Experimentation replaced one-shot campaigns, while cross-functional squads brought together strategy, data, digital, commerce and tech. GPS became the default planning process, supported by training through the Madison Growth Academy.

Das is also clear that MbrAIn is not merely an optimisation layer. Unlike traditional tools that report past events, MbrAIn diagnoses brand context — ambition, barriers, job-to-be-done and growth stage, and applies established patterns to resolve conflicting signals, propose coherent strategies and provide reasoning for planners. While most agency AI tools act after strategy is set, MbrAIn operates at the decision-making level, orchestrating the full GPS planning cycle from brief to strategic plan.

Building for India’s complexity proved challenging. Das points to market heterogeneity, data patchiness and the coexistence of traditional retail, ecommerce and quick commerce as key hurdles. Equally important was building trust with senior strategists.

“The agent had to be good enough that they saw it as a partner, not a toy,” he says.

MbrAIn prioritises inputs through signal weighting, context matching and scenario testing, surfacing trade-offs when signals conflict rather than forcing a single answer.

From campaigns to accountable growth: What changes in practice? 

One of the more consequential shifts within Madison 3.0 is anchoring planning to a Commerce Performance Engine, directly tying media decisions to revenue outcomes.

“The commerce performance engine means we are no longer satisfied with saying, ‘Your awareness went up’ or ‘Your click-through rates are higher,’” Varghese says. “We are taking responsibility for what reaches the cash register.”

In practice, this involves connecting upper-funnel investments such as TV, OTT, creators and OOH to outcomes across ecommerce platforms, quick commerce, D2C and offline retail. GPS is used to determine when spend should build memory and when it should focus on closing a sale, with both measured together rather than in isolation.

The result is a single view of the funnel, where brand metrics, media delivery and sales data “sit in the same room, not in three different dashboards.”

“It changes the conversation with clients,” Varghese adds, “from ‘What did we run?’ to ‘What did we sell, who did we convert, and what did we learn about doing it better next time?’”

Such a shift, he acknowledges, comes with risks. The first risk was internal, requiring retraining teams, redesigning processes and changing long-held habits. The second was financial and strategic, with Madison choosing to invest in building its own system and AI agent rather than relying on off-the-shelf solutions. The third was accountability. 

“By tying ourselves to commerce outcomes, not just media metrics, we accept more scrutiny on performance. When things don’t work, the questions get harder. When things don’t work, the questions get harder,” ” Varghese tells Social Samosa. 

Still, he sees these risks as unavoidable. 

“The alternative — incremental change — would have left Madison stuck in the middle: too big to be fast, too slow to be strategic, and too tactical to be indispensable. Madison 3.0 is our way of choosing discomfort now over irrelevance later.”

Das draws a clear line between automation and judgement. MbrAIn replaces the mechanical parts of strategy — data gathering, initial diagnosis and first-draft narratives — while augmenting human judgement around risk, organisational dynamics, creativity and cultural instinct.

“If strategic planning is a 100-step journey,” he says, “MbrAIn can walk 70-80 of those steps reliably. The last 20-30, where relationships, courage and experience matter, still belong firmly to humans.”

For agency teams, planning becomes always-on, collaboration replaces silos, and learning compounds instead of resetting every quarter. Campaign thinking, Das argues, is limited by short-termism, fragmentation and attribution confusion.

Madison 3.0, Das says, is ultimately aimed at giving clients “a growth machine that can be tuned, explained and improved over time — not just great work that comes and goes.”