Rethinking the role of video infrastructure in Asian retail

Entrance counts with conversion reveal something POS systems cannot: How many people came in and chose not to buy.

Walk into almost any mall, supermarket, or flagship store across Southeast Asia today, and you're being watched, not in a sinister sense, but in the most literal one. Cameras line the entrances, the aisles, the checkout counters, and more. Retailers across the region have spent millions building out this surveillance backbone, largely for loss prevention and safety.

And yet, ask most retail leaders a simple question — how many customers walked through your doors last month — and the answer is often an estimate pulled from POS transaction counts that miss every browser who didn't buy. Ask a harder question, like what the average dwell time was in your highest-margin product zone last week, or at what point your queue length started costing you conversions, and there is often no response.

This is the paradox at the heart of retail in Southeast Asia: An industry sitting on enormous volumes of visual data, generated continuously, store by store, day after day, and largely failing to use any of it.

A region positioned to lead, but not yet ready
The opportunity in front of retailers here is not small. Asia Pacific is set to drive around two-thirds of the world's new retail sales over the next five years, powered by more than 4.3 billion shoppers and 18 megacities. Nearly three-quarters of consumers in the region are already using artificial intelligence (AI) to discover, compare, and learn about products, and retail leaders broadly expect that shift to accelerate.

But there's a gap between the scale of this opportunity and the operational maturity needed to capture it. Despite rising investment, only around 30% of Asia Pacific consumer businesses report that at least 40% or more of their AI initiatives reach production, with implementation challenges cited amongst the top barriers.

Research from Deloitte Southeast Asia tells a similar story closer to the ground: Across Indonesia, Thailand, and Vietnam, consumer, and retail companies are still at a comparatively early stage of adopting data analytics technologies, which is why most are choosing to follow proven approaches rather than lead with new ones.

That caution is understandable, but it also means a lot of the data retailers already generate — including the video data captured by infrastructure they've already paid for — is sitting unused.

From footage to footfall intelligence
The retailers starting to close this gap aren't necessarily spending more. They're spending differently by adding on the layer of intelligence that sits between the camera and the decision-maker.

Consider dwell time.

A camera pointed at a high-margin shelf captures hours of footage every day, but footage only tells you what happened at a point in time. Paired with the right analytics, that same feed can tell a category manager precisely how long shoppers linger at a display before deciding to buy, abandon, or move on. and whether last month's planogram change actually worked, instead of waiting for a quarterly sales report to find out indirectly.

Consider queueing.

Long lines at checkout are one of the most well-documented drivers of abandoned baskets in physical retail, yet most stores still manage queue length reactively, based on a manager's gut feel or a customer complaint. Video-derived queue analytics can flag the exact threshold measured in people, or in minutes, at which conversion starts to drop, giving operations teams a trigger point to open another counter before revenue is lost.

Consider footfall itself.

Entrance counts paired with conversion data immediately reveal something POS systems cannot: How many people came in and chose not to buy. That single metric reframes a ‘slow sales day’ from a demand problem into a possible merchandising, staffing, or store-layout problem, each of which has a very different fix.

None of this requires retailers to rip out and replace what they already have. They just need to connect the infrastructure they already have to the questions they need answered.

The shift retailers need to make
The agentic, AI-driven shift in retail across Asia Pacific is not just about chatbots and recommendation engines on the customer-facing side. It is equally about what happens inside the four walls of the store, where most purchase decisions are still made. Retailers who treat their video infrastructure purely as a security cost center will keep missing this.

Retailers who start treating it as a continuous, real-time source of operational truth will be the ones who can finally answer the basic questions their boards are already asking: Who is coming in, what are they doing, and where exactly is the business losing them.

The cameras are already there. The data is already there. What is missing is the ability to turn that data into operational intelligence that store managers, merchandisers, and executives can act on in real time.

In an industry where margins are tight and customer expectations continue to rise, turning video into actionable insight may prove to be one of the most practical, and most immediate sources of competitive advantage.