Interview conducted by Brij Pahwa for MarTechAI.com’s Leadership Series. Conversation participants included Rajesh Jain, Brij Pahwa, Anupama Mitra and Mekhala Maitra. This text has been edited for clarity while preserving the speakers’ meaning and intent.
Martech is at an inflection point, says Rajesh Jain, Founder & Managing Director of Netcore Cloud. The problem is not a shortage of tools, but a mindset that overpays for reacquiring customers brands already have. In this conversation with MarTechAi.com’s Brij Pahwa, Jain argues that the next wave will be agentic AI that delivers true one-to-one engagement through “brand twins,” a data and decisioning layer that represents each customer to the brand in real time. The payoff he wants marketers to chase: profitable growth, not vanity metrics.
Jain’s thesis is blunt. Most budgets still chase acquisition on ad platforms while neglecting retention, referrals and lifetime value. He calls for CMOs to become “chief AI and profits officers,” build zero and first-party data pipelines, and unify identity across touchpoints. He believes multi-agent systems will make the “impossible” possible, letting a team of one orchestrate thousands of hyper-relevant campaigns every day. He also sets a new profitability rule he abbreviates as UNAM: acquire a customer only once, never pay again to reacquire.
Brij Pahwa (BP): You have watched martech evolve from a tool-scarce world to a universe of thousands of products. How do you frame the real problem martech is supposed to solve, and where are we now?
Rajesh Jain (RJ): If you back up a little, marketing has two elements. There is acquisition, largely through ad tech platforms like Google and Meta that deliver audiences to websites and apps. Then it is martech’s job to take over and build engagement, retention and monetization. Profitability is a function of this interplay. The challenge today is that about 90 percent of marketing budgets end up on acquisition, and a big chunk of that is spent on reacquiring customers you already had. Because martech was not used well to build relationships, you pay again and again to monetize the same customer. That is what hurts EBITDA.
I like to say brands pay a kind of tax of 20 to 30 percent of revenue to three places. First, ad tech platforms for reacquiring. Second, marketplaces that bring revenue but give you little first-party data. Third, discounts to win back lost customers. In the future, you will pay AI platforms and chat interfaces to bring customers back. To become profitable, you need a mindset change. Acquire once, then use martech to build lifetime engagement and maximize lifetime value, including referrals. AI will make a huge difference here.
(BP): Third-party data is fading. For categories like FMCG and CPG that find it hard to capture identity, how can brands build first-party data at scale?
(RJ): Every customer has two digital identities today: a mobile number and an email address. My advice is to collect both with consent. If you only have a mobile number, pushing messages is expensive and deliverability is getting tougher, especially on WhatsApp which wants to protect a P2P inbox.
For FMCG and CPG, think value exchange. Byron Sharp’s work talks about physical availability and mental salience. You need top-of-mind recall. Build a brand daily: a short 15 to 60 second touchpoint that adds value. If you deliver something the customer wants, they will share identity with an opt-in. Also collect zero-party data, the preferences customers tell you. Many brands are one-and-done with 60 to 65 percent buying once and never again. My framework is Best, Rest, Test, Next. You need identity and permission to build engagement hotlines, then send useful content tied to real problems your product solves. Without that, you rely on intermediaries and discounts.
(BP): Marketers talk about ROI more than ever, yet many still struggle to measure it. How do you want CMOs to reframe ROI in the age of AI?
(RJ): The problem is ROI became ROAS, return on ad spend, because 90 percent of budgets are skewed to acquisition and reacquisition. Martech done right can deliver ROI that is profitability oriented. Marketers tend to think only of growth. The single biggest change needed is for the CMO to think like the chief AI and profits officer. CFOs look at costs and CEOs look at many things, but who owns growth and profitability together? With AI, marketers can make that transformation.
(BP): Agentic AI is the buzz, but many are still unclear on what it unlocks beyond content. What is the practical impact you see?
(RJ): Think of AI in two ways. First is the efficiency play: faster, better, cheaper for content and campaign planning. The second is the impossibility play. A brand may have 10 million customers, but a marketing team can realistically maintain only 8 to 10 segments because creating content and campaigns daily is humanly impossible. With agents, a department of one can run 10,000 hyper-relevant campaigns every day. That gets you to N equals 1 personalization. Men 25 to 34 is not the answer. A Brij is different from a Rajesh even in the same cohort.
Two ideas matter. One is multi-agent systems that make operations efficient, where you give instructions and agents take over. The second is for a segment of one: the brand twin. Every customer should have a twin controlled by the brand that the customer can talk to in natural language. It is built on that person’s data plus category and real-world context. We are different people through the day. Agents can represent that. These twins are created by a twin factory. Each twin is like an agent.
(BP): How far are we from a world where a million customers means a million brand twins running in production?
(RJ): Given the pace of AI, I think we will start seeing twin factories within the next 12 months. Telecom needed 30 years to go from 1G to 5G. We are not even three years into ChatGPT and we have seen five versions. Costs are falling, inference is improving. In the physical world, a salesperson reads you quickly. A twin can hold all the data, see what people like you are doing and understand intent through conversation. This leads to a twin ledger. Profitability is the sum of profitability at the customer level. There should be a ledger of debits and credits for each customer: what it cost to acquire, what you spend to serve, what you spend to win back, what margins and referrals you earned. The bigger question is why five hundred billion dollars is wasted on reacquisition. Brands are not building one-to-one relationships. Agentic AI is the path to sustainable, systematic, exponential profitable growth. Outcomes should matter more than inputs.
(BP): Data fragmentation fuels irrelevance. Many companies still run on silos and conflicting signals. What needs to change in the data layer to make your vision work?
(RJ): You need a unified layer where all data is pooled. Call it a CDP, but the next CDP cannot be passive. It must be active because the twins sit on it. It is not just martech data. Customers have lives outside your site or app. They search, talk on social, react to events. Weather, concerts, category patterns, all of it matters. Bring it together into what I call a large customer model for the brand. Once you have that, you can build brand twins.
For the agentic future, move to a panel-less world. Eliminate dashboards. Let a marketer talk to the data in natural language. Ask what happened yesterday and what to change today. Customers should be able to do the same. We cannot do that with layers of static dashboards. The foundation is identity and unity. Identity means two pieces of information for every person so you can reach them without waiting for them to return. Unity means consolidate the data. Then segment based on forward-looking lifetime value using Best, Rest, Test, Next. With that in place, strategies for loyal, lapsed and lost customers become very different, and twins get better every day.
(BP): Privacy laws like India’s DPDP and Europe’s GDPR are raising the bar. How do you capture, store and activate consent while still personalizing?
(RJ): The best data is zero-party data. It is what we voluntarily share with brands because we get something back. Make it a value exchange with micro-incentives, what I call atomic rewards. If brands focus on zero and first-party data, there is no data problem. Third-party data is where issues come from. Every touchpoint can collect zero-party signals. In every email, show two images and ask hot or not. Hill stations or beaches. Very few are doing this. Every transactional message is a chance to capture preferences and reduce third-party dependency. People want personalized experiences. Agentic AI makes that possible. Marketers of tomorrow will be white-coding. They will tell the system what they want, it will write the code and give answers, and marketing becomes 24 by 7 because customers are always on.
(BP): We also see debate on model sizes and where this all goes next. What is your end-state vision for marketing if you look out a few years?
(RJ): In business the job is to make money, and money comes from customers. What gets lost is that we look at P&L in aggregate. Economics has a principle called methodological individualism. The unit of analysis must be the single customer, not the collective. Recommendations and P&L should go to N equals 1. It may not happen instantly, but that is the goal. On acquisition, think UNAM: only once, never again. Acquire a customer only once and never pay again to reacquire. Otherwise, in many brands, advertising and marketing spend grows faster than revenue. That is not sustainable.
(BP): If you had to leave CMOs with one operating rule for 2025, what would it be?
(RJ): Chase intelligence, speed and measurable business impact. Build the identity and data foundation, collect consented zero and first-party data, unify it, talk to it in natural language and put agents to work so you can deliver N equals 1 personalization. Move the organization to profitability thinking. Make outcomes matter more than inputs.