AI Search
As artificial intelligence rapidly reshapes the way consumers discover, evaluate, and choose brands, marketers are facing a fundamental shift in visibility and engagement. Traditional search rankings and social media presence are no longer enough in a world where AI assistants increasingly determine which brands enter the consideration set. Recognising this emerging challenge, Pulp Strategy launched NeuroRank, a platform designed to help brands understand, govern, and optimise how large language models perceive and recommend them.
In conversation with Adgully, Ambika Sharma, Product Architect, NeuroRank and Chief Strategist, Pulp Strategy, discusses the rise of Large Language Model Optimization (LLMO), the changing dynamics of consumer engagement, common misconceptions around AI discoverability, and why governing brand visibility in AI-driven conversations is set to become a critical business priority for organisations in the years ahead.
What has been the biggest shift you have witnessed in consumer engagement over the years?
The biggest shift has been in who controls the conversation.
For nearly two decades, brands controlled the surface. They owned the campaign, the channel and the message, while consumers engaged on the brand’s terms. Search disrupted that model first, followed by social media.
Today, however, we are witnessing an even bigger transformation. Consumers are no longer engaging with brands at the first moment of consideration—they are engaging with AI. Increasingly, it is AI that determines which brands enter the conversation.
By the time a consumer reaches a brand’s website or platform, AI has already shaped their expectations. The brand is no longer the primary storyteller; it has become the subject of someone else’s summary.
Consumer engagement has effectively moved upstream of the brand itself, while most marketers continue to optimise the downstream experience.
That is precisely why we built NeuroRank. The upstream decision-making layer is where purchase intent now begins, yet there was no solution available to measure or manage it.
What gap in the industry were you trying to solve with the launch of NeuroRank and the concept of LLMO?
By mid-2024, we were already seeing ChatGPT misrepresent brands—hallucinating facts, recommending competitors and omitting companies that should have been cited.
Brands could monitor their Google rankings, but they had no way of understanding what AI models were saying about them.
There was no recognised discipline, no playbook and no platform built for this emerging challenge. Existing marketing tools were measuring the wrong surface altogether.
That’s where Large Language Model Optimisation (LLMO), also referred to as Generative Engine Optimisation (GEO), comes in. It is the discipline of governing how AI engines perceive, describe and recommend a brand.
NeuroRank enables brands and agencies to actively manage their AI search visibility. It goes far beyond monitoring. Monitoring only tells you whether you’re visible. NeuroRank completes the entire cycle—deconstructing how AI models perceive a brand, diagnosing inaccuracies, prescribing corrective actions, improving the information sources AI relies on, and measuring progress month after month.
We didn’t build NeuroRank because it was an interesting technology challenge. We built it because the gap was already costing brands business opportunities. AI search has grown rapidly, but expertise in this space hasn’t kept pace. NeuroRank helps brands close that gap before it becomes a competitive disadvantage.
What has been the core philosophy behind Pulp Strategy’s growth journey over the years?
Our philosophy has always been simple: identify major shifts early, build deep expertise and stay invested long enough to create meaningful value for brands.
Every major transformation in marketing has followed a similar pattern—search, social, mobile, martech and now AI. Each new platform creates an expertise gap that takes years for the industry to bridge.
Pulp Strategy has consistently moved early across these transitions. Over the past 15 years, we’ve earned 124 marketing awards through multiple waves of industry change. Those awards aren’t simply recognition—they validate the philosophy behind our approach.
The second part of our philosophy is equally important: we build, rather than simply advise.
Whether it’s Yukti, India’s first AI humanoid agent; Channel Command, our partner growth platform; or NeuroRank, our LLMO platform, we’ve found that the deepest understanding comes from building technology ourselves rather than simply operating within it.
NeuroRank is the latest expression of that philosophy. We recognised the AI opportunity in 2024, and today the platform is increasingly helping brands and agencies establish robust GEO practices.
How do you balance creativity with measurable business impact while building campaigns for brands today?
The question assumes creativity and measurement are opposing forces. They are not.
Creativity without measurement is decoration. Measurement without creativity is simply a spreadsheet that fails to move markets. Brands that prioritise one at the expense of the other will inevitably lose to those that embrace both.
Everything begins with a clearly defined business objective—one that even a CFO would confidently defend. Only then do you determine the creative approach most likely to achieve that outcome.
Awards should be the consequence of effective work, never the objective. If creative work is memorable enough to influence behaviour and measurable enough to demonstrate business impact, recognition naturally follows.
There’s also a new dimension to creativity today. The most measurable creative output is no longer just an advertisement—it is the content AI cites when consumers ask questions about a category. That content is still creative work, but it is now being evaluated by AI before it reaches the audience.
What are the biggest misconceptions marketers currently have about AI discoverability?
There are four major misconceptions, and each carries a significant cost.
The first is that strong SEO automatically translates into AI visibility. It doesn’t. Being number one on Google does not guarantee that AI models will recommend your brand. In fact, fewer than one in six links cited by AI engines overlap with Google’s top results.
The second misconception is that AI accurately describes every brand. It often doesn’t—particularly for non-branded searches. AI may omit your brand altogether, attribute your strengths to competitors or present outdated information as fact. Yet only around 14% of marketers actively track whether AI models cite their brand, meaning the vast majority are operating without visibility.
The third misconception is assuming that because ChatGPT or Gemini presents favourable results to you personally, your brand enjoys universal AI visibility. Personalised interactions are influenced by contextual learning and user history. That experience does not reflect how AI models respond across the broader user base.
The fourth—and perhaps the most expensive—is believing AI optimisation is a one-time project. It isn’t. AI models continuously retrain, information sources evolve and brand visibility changes over time. Brands that stop actively managing their AI presence gradually disappear from AI-generated answers.
NeuroRank was built specifically to address these challenges. Through its structured five-step monthly governance framework, it equips brands and agencies with the intelligence, diagnostics and ongoing optimisation required to manage AI discoverability effectively.
As consumer discovery shifts towards AI assistants and conversational platforms, how should brands rethink visibility?
Brands need to stop thinking about visibility as a ranking on a page and start thinking about it as a position within a conversation.
Traditional search was based on rankings—you competed for a slot on the first page. AI doesn’t work that way.
AI models typically recommend only a handful of brands. If you’re not included, you’re not simply ranked lower—you are completely absent from the consumer’s decision-making process.
That changes the discipline entirely.
Visibility today depends on becoming one of the most trusted sources AI can rely upon. AI models build responses from third-party reviews, industry publications, expert commentary and structured information far more than from a brand’s owned channels.
Optimising AI visibility therefore requires coordinated improvements across owned, earned and third-party ecosystems.
That’s exactly what NeuroRank enables. The platform identifies every source AI models rely upon—including competitor signals—and provides intelligence and best practices to strengthen visibility across all major AI engines over time.
Visibility used to be a media planning exercise. Today, it has become an enterprise-wide governance discipline. The brands that recognise this shift early will define the next decade.
What were some of the biggest learnings from stress-testing NeuroRank across 150+ brands and 65 industries?
Working closely with more than 150 marketing teams across multiple countries generated valuable insights that fundamentally improved the platform.
The first major learning was that AI models rely on different trust sources depending on the industry, geography and even the specific prompt. Brands often struggle because they don’t know which external sources are shaping AI responses.
For example, we encountered brands whose AI-generated descriptions repeatedly claimed they suffered from poor customer service despite internal customer satisfaction scores exceeding 90%. Improving their website alone couldn’t solve the issue because AI was relying on third-party sources. That insight led us to redesign our recommendation engine around source-level governance.
The second—and perhaps most significant—learning is that AI governance should not remain solely within marketing.
Think about how organisations allocate responsibility today. Investor information belongs to finance. Customer support belongs to service teams. ESG communications belong to sustainability leaders. Employer branding belongs to HR. Product information belongs to product teams.
But when ChatGPT, Gemini, Claude or Perplexity answer questions spanning all these domains, who owns the accuracy of those responses?
In most organisations, the answer is nobody.
The CMO often inherits responsibility simply because the conversation includes the word “search”. Yet AI representation extends far beyond marketing. It encompasses finance, customer service, sustainability, talent and product information.
This is fundamentally an enterprise governance challenge. Every C-suite function has a stake in how AI represents the organisation. Ultimately, governance should sit at the CEO and board level, with coordinated ownership across the CMO, CFO, CHRO, CSO and CPO.
Companies that establish this governance early won’t simply improve AI visibility—they will build a consistent, machine-readable representation of their organisation across every stakeholder group.
The third learning came from enterprise organisations, particularly in BFSI. They made it clear they would never deploy AI-driven content recommendations without structured approval workflows.
That insight led us to introduce NeuroRank’s Maker-Checker governance layer, ensuring every recommendation is drafted, reviewed and approved before implementation.
Overall, our biggest takeaway was that this category rewards governance rather than speed. Measuring the problem is relatively straightforward. Delivering the right treatment is where long-term competitive advantage lies.
Looking ahead, what does the next evolution after LLMO look like?
LLMO focuses on governing how large language models describe a brand when a person asks a question.
The next evolution is AI agents acting autonomously.
We’re moving from AI search to AI action. Instead of consumers asking which product or policy is best, AI agents will increasingly compare options, shortlist providers and complete transactions independently.
Brands will no longer be evaluated only by humans—they will also be evaluated by machines acting on behalf of humans.
The organisations investing today in trusted authority, structured data and continuous AI governance through LLMO will be the same organisations AI agents choose tomorrow.
Those that treat AI visibility as a one-time optimisation project risk becoming invisible twice—first to AI models and later to the AI agents making purchasing decisions.
NeuroRank was designed for the first stage of this evolution, but its architecture—governance, source optimisation and multi-engine intelligence—is already built to support what comes next. We are actively preparing for that future.
similar newsAvneet Singh on Medusa Beverages’ marketing playbook to challenge legacy beer brands
similar newsLubricants to future mobility: How Gulf Oil is reinventing marketing for the AI and EV era
similar newsDSP Mutual Fund’s new campaign challenges the frictionless digital investing illusion
similar newsPartha Pratim Sengupta on why Bandhan Bank is investing in India’s youngest future customers
similar newsI’d rather have a CEO call me with a problem than a campaign brief: Jitender Dabas
similar newsThe great social reset: Why brands can no longer treat social media as a channel
similar newsMarTech’s new brain: The AI takeover of social strategy
similar newsFrom campaigns to conversations: How social media is reshaping Indian advertising
similar newsThe new rules of financial PR: Ushang Sheth on AI discoverability, crisis, and power of storytelling
