Back to All Writing
    e272026-05-20

    The founders who scale in SEA are rebuilding customer experience from the operating layer up

    Originally published in e27 on 2026-05-20.

    Read Original Article on e27

    Core Argument

    AI-powered CX is not a technology investment. It is a scaling decision. Most founders in Southeast Asia get customer experience wrong at exactly product-market fit, when the product works, customers are coming in, and the temptation is to hire more people to serve more people. That is not scale. That is growth that will eventually collapse under its own weight. The challenge that kills the next phase is almost never the product. It is the operating layer, the systems and touchpoints between a product and its customers, built for the previous phase, not the next one. Southeast Asia is not one market. Customers in Jakarta, Colombo and Ho Chi Minh City do not respond to the same tone, channel or cadence. AI with natural language processing across Bahasa, Tamil, Vietnamese, Sinhalese and Thai is deployable infrastructure now, letting businesses deliver personalised service at a market depth no human team could reach cost-effectively. ASEAN's AI adoption grew 38 percent year over year, with 29 percent of businesses using it in some form. The problem is sequencing, not adoption. A founder achieves product-market fit, then implements AI in pockets, a chatbot here, an automated email there, and the result is fragmentation, not scale. A chatbot placed on top of a broken support workflow does not fix the workflow. It automates the frustration. An AI-native customer journey, where routing, personalisation, escalation and feedback loops are designed together, creates compounding returns: response time reductions of 70 percent and first-contact resolution improvements of up to 40 percent reported by businesses running integrated platforms. Western markets carry decades of legacy CRM and helpdesk infrastructure that is expensive to unwind. Asian businesses do not carry that weight. A business in Colombo, Manila or Bandung can build an AI-native customer operating layer before the next layer of legacy tech calcifies. That window is 18 to 36 months.

    What I'd Revise Now

    The macro numbers this column leans on have since been confirmed rather than superseded. Google, Temasek and Bain's e-Conomy SEA 2025 report, released four months after this column, puts the region's digital economy at $300 billion in GMV for 2025, exactly the trajectory this piece described, with AI investment reaching roughly $2.3 billion across more than 680 regional AI startups. The 18-to-36-month leapfrog window this column named is now roughly four months shorter. Nothing in the subsequent data suggests the window has widened; if anything, the same Google/Temasek/Bain report frames AI-native customer experience, what it calls "experience redefinition", as one of the region's three defining transformation categories for the coming decade, which suggests more businesses are now racing for the same window rather than fewer. The specific 70 percent response-time and 40 percent resolution figures in this column come from vendor-reported platform data rather than an independent regional survey, and I have not found a comparable regional benchmark to confirm or revise them. Treat them as directionally right rather than independently re-verified.

    Key Takeaways

    • Customer experience gets rebuilt wrong at exactly product-market fit, when hiring more people looks like scale but isn't
    • The sequence matters: redesign the operating layer first, then deploy AI into a clean architecture, not the reverse
    • Integrated AI-native CX reports 70 percent response-time reductions and up to 40 percent first-contact resolution improvement
    • Southeast Asia's lack of legacy CRM infrastructure is a genuine leapfrog opportunity, open for 18 to 36 months
    • Southeast Asia is not one customer; language, trust behaviour and channel preference genuinely differ by market

    Related Insights