AI speeds up marketing. It doesn't replace human judgment.
At the Pitch CMO Summit 2026 in Mumbai, senior marketers made one thing clear: AI is an engine, not the driver. It accelerates analysis and execution, but the brand's "why" still lives with people.
In a fireside chat, Sundar Kondur (Chief Revenue Officer, The Hindu Group) joined Diptakirti Chaudhuri (CMO, Casagrand) and Naarayan T V (CMO, Akasa Air) to unpack how to use AI without losing the human signal.
AI answers the what and how. Humans own the why.
Opening the session, Kondur pointed to a question raised by AI pioneer Geoffrey Hinton: can artificial intelligence become "artificial wisdom"? His take: AI handles speed and specificity, but meaning and intent are human work. Data explains patterns. People decide purpose.
AI is here to stay-put it to work where it compounds
Naarayan framed it simply: a decade ago, data was the new oil. Now AI is the tool that refines it into output. It's already useful in customer segmentation, churn prediction, and lifetime value models. Brand building, however, still needs human insight and taste.
If you're still testing the edges, start where AI has clear traction: prediction, clustering, creative variations, and decision support. Keep brand calls with people.
The funnel isn't a funnel-it's a loop
Chaudhuri challenged the classic AIDA model. From the top view, it's closer to a circle than a funnel. In high-consideration categories like real estate, timelines zigzag: some prospects buy in weeks; others pause for months and re-enter with new needs.
AI may speed research, scoring, or outreach, but it won't magically compress the purchase cycle. It was always messy-we now have better data proving it.
Personalisation pays-until it feels creepy or chaotic
Chaudhuri flagged a common miss: people still see ads for what they already bought. That's not an AI issue. That's a data discipline issue. AI only reflects the data-and the logic-you feed it.
He also warned against "vibe marketing": pumping out hundreds of AI-made creatives without clear cohorts. If you don't know your audience, AI will amplify the wrong message, faster.
Want a primer on tested playbooks? See AI for Marketing for practical uses across personalisation, campaign optimisation, and analytics.
Performance now vs. brand tomorrow
Naarayan highlighted the pressure: immediate ROI gets the budget, but trust compounds slowly. Both matter. Ignore short-term performance and you miss targets. Ignore long-term brand and you shrink pricing power and retention.
The smart move: set parallel goals, split budgets intentionally, and keep measurement honest about time horizons.
Marketing is bigger than the marketing team
Chaudhuri reminded the room: every function shapes the brand-sales, support, frontline ops. As AI tools spread across teams, marketers must anchor the work in customer insight. Tools will change. The consumer stays the North Star.
What to implement this quarter
- Define the "human why": write a one-page brand belief, problem solved, and non-negotiables. Use it to check every AI output.
- Pick 3 AI use cases with measurable upside: segmentation refresh, churn-risk alerts, and creative iteration for top campaigns.
- Fix data basics: event tracking, ID resolution, consent status, and suppression rules (especially post-purchase). Garbage in, garbage out.
- Tighten personalisation guardrails: frequency caps, recency windows, and exclusion logic by lifecycle stage.
- Protect the brand: build a tone guide, banned claims list, and visual do's/don'ts that AI tools must follow.
- Balance your mix: ringfence a fixed percent for brand (e.g., 40%) and the rest for performance. Review quarterly, not daily.
- Prove value with time-based metrics: short-term (CAC, CTR, conversion rate) and long-term (LTV, aided awareness, search share).
- Enable the front line: arm sales and support with AI-assisted FAQs and talk tracks that match current campaigns.
- Audit "vibe marketing": kill creative volume that isn't tied to a cohort, a message, and a metric.
- Run a monthly red team: humans review AI outputs for bias, brand drift, and broken logic.
For CMOs building an AI-capable org
If you're formalising the roadmap-org design, processes, and governance-see the AI Learning Path for CMOs for structure, maturity models, and roll-out checklists.
Bottom line
As Kondur put it, technology is a powerful engine, but the driver still matters. Use AI to move faster and see clearer. Keep humans in charge of meaning, trust, and the moments that make a brand worth choosing.
Your membership also unlocks:
