Global agency networks are optimised for a world that is ending: Beerajaah Sswain, NeuGenM.AI

How do you test if India's biggest brands are visible on AI tools like ChatGPT?

Sswain describes a proprietary methodology built around NeuGenM's India AI Readiness Score™ — "a five-pillar framework that measures how well a brand shows up when consumers ask AI engines for advice, recommendations, or answers."

In practice, the team runs over 200 carefully designed prompts per brand across five AI platforms — ChatGPT, Gemini, Perplexity, Claude and Google AI Mode — phrased the way real Indian consumers actually talk to AI ("best cement for home construction in India," "which health insurance should I buy," "suggest a good shampoo for dry hair"), in English and in vernacular languages including Hindi, Tamil, Bengali and Marathi.

Each brand is then scored on five pillars:

• GEO Presence — does the brand get mentioned at all in AI-generated answers?
• Content Readiness — is the brand's website structured so LLMs can read and digest it?
• AI Share of Voice — when mentioned, how prominent is it against competitors?
• Vernacular Coverage — does it show up in non-English AI responses?
• Sentiment Alignment — is the description accurate and positive, or outdated and misleading?

Those roll up into a composite GEO Pulse Score, graded A through F. For the India AI Readiness Report, the exercise covered 100 brands across eight industry verticals — over 20,000 data points, verified by the Signal team.

Your report gave every top Indian brand a C grade or lower. Did any sector perform worse than the others?

BFSI was the best-performing vertical — and even it managed only a C+. The reason is structural: financial services firms are compelled by regulation to publish detailed, structured content (product disclosure documents, rate tables, compliance pages), which happens to be exactly what LLMs find easy to ingest. As Sswain puts it, those brands "got an accidental head start."

At the other end, legacy FMCG performed significantly worse. "D2C brands — the digitally-native direct-to-consumer players — outperformed legacy FMCG giants by a factor of 2.4 times on AI readiness." Not because legacy FMCG spends less — they spend orders of magnitude more — but because D2C brands were born with content architectures LLMs can digest: structured product pages, ingredient lists, comparison content, user-generated review ecosystems. Legacy FMCG built its digital presence around display ads and campaign microsites, formats that are "essentially invisible to AI engines."

Telecom, automotive, pharma and retail all clustered in the C-minus to D range.

What is the very first step brands must take to fix this?

"The first step is not technology. It's a content architecture audit."

Most Indian brands, Sswain argues, still run websites built for the Google search era — designed to rank through keywords, backlinks and meta tags. AI engines instead synthesise answers from structured data, schema markup, entity relationships and authoritative third-party sources. So the first question a brand must answer is whether an AI engine reading its entire digital footprint can actually understand what it sells, who it serves, and why it is better than its competitors.

In almost every audit NeuGenM has run, the answer is no — and the brand's own website contributes zero per cent of the citations appearing in non-branded AI category answers. Those citations come from review portals, editorial platforms, forums and listing sites the brand does not control. The fix begins with making owned content LLM-digestible: proper schema markup, structured FAQ and comparison content, authoritative long-form answers to real category questions, and vernacular availability. Under NeuGenM's NCS2M™ methodology, schema-level fixes are designed to show first measurable improvement within 48 hours, with full content architecture overhauls completing in 8 to 12 weeks.

Why would an established brand trust your Bangalore consulting firm over global agency networks they have worked with for decades?

"Because global agency networks are optimised for a world that is ending."

The holding-company model, Sswain argues, was built to buy media at scale — TV, print, outdoor, then programmatic display and paid search. That works when the consumer journey starts with a Google search and ends with a click. When a consumer instead asks ChatGPT which cement to use, or tells Gemini to find a family health insurance plan, the traditional media plan is irrelevant: no amount of display spend or keyword SEO buys a place in that generated answer.

He points to NeuGenM's founders — Amrit Thomas and Ashish Thukral — as C-suite veterans of those same global organisations who know the structural blind spot from the inside: none of the networks have built proprietary AI visibility products, because their revenue models depend on media buying commissions that AI engines do not generate.

Against that, he sets NeuGenM's own stack: Signal, a GEO platform purpose-built for India and South Asia, with brands audited across India, Vietnam, Malaysia and Singapore; Cortex.ai, the proprietary engine powering paid media optimisation; and NeuGenM Sigma, a unified data platform integrating over 300 source connectors with identity resolution built for DPDP Act compliance. "The question isn't whether a brand should leave their global agency. The question is whether their global agency can solve this specific problem."

How can old-school Indian businesses fight back before they lose customers to AI-savvy startups?

"Legacy companies have three structural advantages that no startup can replicate: brand trust built over decades, distribution networks that reach every pin code in India, and marketing budgets that can move markets overnight." What they lack is the connective tissue between those assets and the AI answer layer.

His three moves: own the category narrative in AI engines, since the brand publishing the most authoritative, structured, citation-worthy content on a topic wins the recommendation; activate the existing first-party data advantage — CRM, loyalty, dealer and service records that startups simply do not have — through Sigma and into LLM conversation environments via Signal Ads; and move at startup speed, adopting a "Think Big, Act Small, Scale Fast" philosophy. Rather than a twelve-month transformation programme: pick one category, audit in a week, deploy fixes in 48 hours, measure the Pulse Score in 30 days, then scale. That is the NeuGenM Springboard methodology — diagnose, immerse, consult, execute — compressed into sprints.

How does tracking brain responses actually help a brand sell more?

Traditional research asks people what they think, and "people are terrible at answering that question honestly" — they say they loved an ad but cannot recall it the next day, say price matters most but choose on packaging colour. Consumer neuroscience measures the reactions underneath: eye-tracking, facial expression analysis and biometric signals, all of which fire in milliseconds, before rationalisation.

"What consumer neuroscience gives a brand is the truth underneath the stated preference." It identifies which three seconds of a thirty-second ad create memory, and which product image on an e-commerce page is being skipped. At NeuGenM it informs both creative strategy and content architecture — so that content engineered for AI citation also lands when a human finally encounters it.

When you build an AI system for a client, do you use existing software or write code from scratch?

Both, on a clear rule. Where proven enterprise infrastructure exists, NeuGenM uses and whitelabels it — Sigma sits on enterprise-grade data infrastructure, customised for India with DPDP Act compliance, local data residency and 300-plus connectors. Where no tool solves the problem, the firm builds from zero: the Cortex.ai engine behind NeuGenM Surge, the GEO Pulse scoring methodology, the India AI Readiness Score™ framework, the NCS2M™ content methodology, and the Signal Ads product that places contextual advertising inside LLM conversations via the LLM Open Exchange network.

CTO Gautam Arora leads the architecture. The philosophy, in Sswain's words: "never rebuild what already works well, but never outsource the intelligence that creates competitive advantage."

How do physical events and experiential marketing connect to AI search optimisation?

Mandeep Malhotra, who brings nearly three decades in experiential marketing and brand activation through his practice TONIC, connects to a principle Sswain says pure-play digital agencies miss: AI engines do not only ingest web content — they ingest the digital footprint that real-world events generate.

A launch, pop-up or stadium activation throws off social posts, news coverage, blog reviews, video and UGC, all of which become part of the citation ecosystem AI engines draw on. The problem is that experiential and digital content strategy usually sit in separate silos, with nobody designing the event to produce structured, citable, AI-digestible output. "The physical event becomes a content engine for AI discoverability, and the AI visibility data informs where and how to design the next physical activation."

Can billboard campaigns really help a brand show up in AI recommendations?

"Offline media already influences AI recommendations — most brands just don't know how to engineer that connection."

When a Mumbai billboard campaign goes live and people photograph it, post about it, or journalists cover it as a creative story, that content joins the citation pool AI engines draw from. Most OOH, though, is designed for impressions and then simply ends. Sswain's fix: design offline campaigns to be inherently shareable and searchable; give every activation a structured digital shadow (a landing page with schema markup, an entity-rich press release, properly tagged social content); and run a GEO Pulse audit before and after to measure whether the offline activity actually moved the score. NeuGenM's audits show brands with strong editorial and PR ecosystems tend to score higher on non-branded AI recommendations.

Critics say AI recommendations are biased and unreliable. How do you ensure your data is accurate?

"We're not asking AI engines a single question and taking the first answer at face value." Over 200 prompts per brand, across five platforms, in multiple formulations, branded and non-branded, in English and vernacular languages — a multi-platform, multi-prompt, multi-language design intended to wash out any single engine's bias.

Second, NeuGenM separates measurement from advice: the GEO Pulse Score "is a measurement tool, not an oracle." Whether an AI's recommendations are fair is beside the point — consumers are receiving them and buying on them. Third, results are verified against real-world signals through Sigma, which connects AI visibility data to conversion rates, CPA trends and revenue attribution. NeuGenM's research found brands with strong GEO signals show lower paid media cost per acquisition — a correlation Sswain offers as evidence the score measures something real rather than a vanity metric.

What is the biggest mistake Indian marketing executives make when deploying AI?

"They treat AI as a productivity tool instead of a strategic capability."

The recurring pattern: a marketing head gets excited, asks the team to use ChatGPT or Copilot for emails and social captions, declares the organisation "AI-enabled" — and three months later nothing has changed. "The CMO who asks 'can AI write my ad copy faster' is asking the wrong question. The CMO who asks 'are consumers asking AI engines to recommend my competitor instead of me, and if so, why?' is asking the right one."

The second mistake is siloed adoption. AI visibility is not a digital marketing problem but a brand strategy problem touching content, product, PR, customer experience and data infrastructure at once. NeuGenM's Springboard methodology begins with cross-functional discovery across six functional areas before any technology is deployed.

Many AI startups burn through funding. What is your revenue model?

"We're not a venture-funded startup burning cash to chase growth." NeuGenM is founder-led with four distinct streams: consulting and advisory (GEO audits, AI readiness assessments, CXO advisory — profitable on its own); managed services through the sub-brands Shelf (retail media), Swarm (social and influencer) and Surge (paid search and social), generating recurring retainer revenue; product and platform licensing via Signal Ads on the LLM Open Exchange network and Sigma platform fees, which scale with LLM adoption across India and Southeast Asia; and training through the Marketing.AI Academy.

"The model works because no single stream has to carry the business… We're not dependent on raising the next round to survive the next quarter."

Are Indian marketers eager to learn these tools, or afraid of losing their jobs?

"The fear exists — we'd be dishonest to pretend otherwise." But through the Marketing.AI Academy and client workshops, Sswain says the fear is quickly overtaken by the realisation that AI makes good marketers dramatically more effective, and that those who learn first will lead rather than be replaced. The marketer who can run a GEO audit, interpret a Pulse Score and build LLM-optimised content is not competing with the AI — they are competing with every other marketer who cannot.

He notes a strong preference for applied over theoretical training: Indian marketers do not want a lecture on what generative AI is; they want to be shown how to check whether their brand is visible on ChatGPT, and how to fix it. "The eagerness is real, and it's growing faster than the fear."