73% of Companies in Southeast Asia Are Using AI — But Most Are Not Seeing Returns

A new report from McKinsey, EDB Singapore, and Tech in Asia paints a striking picture: 73% of companies in Southeast Asia are already piloting or scaling AI initiatives. That puts the region ahead of the global average, where only about 35% of companies have moved beyond experimentation.

But there is a second finding that gets less attention: the majority of firms that invested in AI are struggling to see financial returns. Here is a number that should make every business leader pause — over 60% of these companies say AI has contributed less than 5% to their earnings. Nearly one in five say it has had no noticeable impact at all.

These are not companies that avoided AI. These are companies that bought the tools, ran the pilots, and committed the budgets. And most are not seeing meaningful returns.

The adoption-returns gap

Nearly half of surveyed companies in Southeast Asia (46%) have moved beyond piloting to actually scaling AI. That is impressive. Companies across Asia Pacific expect AI to deliver strong financial returns in 2026. Spending is up. Budgets are growing. The average firm expects nearly three times return on their AI investment.

But expectation and reality are diverging. This is not an argument against AI. It is an argument for better implementation.

Three reasons AI investments fail

We have worked with enough businesses in Cambodia and the region to see the patterns clearly:

1. The pilot that never scales. A team runs a successful AI pilot. It works great in a controlled environment with motivated early adopters. Then it stalls. The pilot never becomes standard practice because nobody invested in changing the broader team's workflows. This is the most common failure mode.

2. Tool-first thinking. A company decides they need AI, evaluates vendors, picks a platform, and deploys it. Then they try to figure out what to use it for. This is backwards. The businesses seeing real returns start with a specific problem and find the right tool to solve it — not the other way around.

3. Underestimating the human side. AI adoption is a change management challenge disguised as a technology challenge. Buying the software is 10% of the work. Getting your team to actually change how they work is the other 90%. This is why workshops work better than training — the team needs to be involved in designing the change, not just informed about it.

What high-performing companies do differently

The McKinsey report identifies clear traits that set high-performing AI adopters apart:

• They align AI initiatives with clear business objectives
• They invest in people and processes, not just technology
• They start focused and scale what works
• They measure ruthlessly — before AI: this task takes 4 hours; after AI: 30 minutes

The companies seeing 3x returns did not start with a company-wide AI strategy. They started with one specific workflow that was clearly broken, automated that, proved the value, and expanded from there. A 12-person service company automated 60% of their admin work following exactly this approach.

What this means for Cambodia

Cambodia is part of this Southeast Asian AI wave, but most businesses here are still in the early stages. That is not a disadvantage. It means you can learn from the expensive mistakes others have already made.

The companies in Singapore and Indonesia that are struggling with AI returns largely made the same error: they bought tools before understanding their workflows. Cambodian businesses have the advantage of starting fresh — no legacy AI systems to untangle, no failed implementations to work around.

The practical playbook: - Pick one workflow that costs your team significant time every week - Map exactly how it works today, step by step - Identify the repetitive, predictable parts — those are the AI candidates - Involve your team in designing the solution, not just using it - Measure the impact with clear, specific metrics everyone can see - Expand to the next workflow once you have proven the value

This is not a six-month strategy project. For most businesses, the first automation can be running within weeks.

The bottom line

The question is no longer "should we adopt AI?" — 73% of your regional competitors already have. The real question is how to adopt it in a way that actually delivers results.

AI is not failing businesses. Businesses are failing at AI implementation. The technology works. The returns are real — for companies that do it properly. The difference between the companies seeing 3x returns and the ones seeing no impact is not budget or technology. It is how they approach the human side of adoption.

That is exactly the gap we help businesses in Cambodia bridge. Not selling AI tools, but helping teams implement them in ways that deliver measurable results. If you are considering an AI investment — or if you have already invested and are not seeing the returns you expected — let us talk about what is actually going wrong.

Book a free consultation to talk through where AI could make a real difference for your team.