Almost every founder I speak to in Singapore is already using AI somewhere in their marketing, usually ChatGPT for a first draft. Very few can say it has moved the numbers. That matches the research: the tools are now cheap and everywhere, and the gap is in how well a team puts them to work.
What the research says
The State of AI in Marketing 2026 report from MMA and Decision Lab surveyed 143 marketers across Southeast Asia. A few findings stood out:
Skills, not tools, are the problem. Around three in four respondents said their biggest challenge is skills and training. Among the most advanced teams, the talent gap was also the number one blocker to using AI agents.
Most teams use AI for content, not results. Content and creative are where AI is used most. Measurement, attribution and media allocation are the least scaled uses, even among the leaders.
Search is shifting. Almost a third expect the move from traditional search to AI search to shape their marketing this year.
Few teams manage the risks. Half worry about AI making up answers, but well under half have a formal plan for AI risk.
The survey covers Southeast Asia as a whole, and most respondents work at agencies or media companies, so treat it as a regional read rather than Singapore data. But it matches what I see with small businesses here.
Where AI pays off for a small business
1. Getting found in AI search. Your customers now ask ChatGPT and Google's AI answers who to use. Making your website easy for those tools to read and quote is one of the best-value pieces of marketing work available right now. For SupplementHub, orders from AI assistants rose 49% year on year after we did this work. I explain how in this case study.
2. Reporting that tells you what to do. Most small teams have the data and no time to read it. An AI agent can pull figures from your ad accounts, analytics and email platform every week and write a short plain-English summary with three actions. I run this for a subscription app client every Monday. It replaces hours of spreadsheet work and catches problems days earlier.
3. Taking repetitive work off your plate. Drafting follow-up emails, researching prospects, checking ad accounts for overspend and preparing meeting notes are all good jobs for AI agents, as long as a person approves anything that goes out. I run my own business this way: agents draft replies across five inboxes each morning and check client ad accounts daily. That is how I can serve twice as many clients without cutting corners.
Where it does not pay off (yet)
Publishing AI content unedited. It reads like everyone else's, it can state things that are not true, and it does nothing for your brand. Use AI for the first draft and research, then edit hard.
Buying tools before fixing the basics. If your tracking is broken or your offer is unclear, AI will just help you do the wrong things faster.
Letting AI decide budgets on its own. Ad platforms already automate bidding. You still need a person setting goals, checking the numbers and deciding where the money goes.
Anything that sends messages without approval. One wrong email to a customer costs more than the time it saved.
How to start without wasting money
Pick one painful, repetitive job. A weekly report or a follow-up email is ideal. Do not start with a full AI strategy.
Use the tools you already pay for. Most of what a small team needs runs inside Gmail, Google Workspace, Notion and your ad accounts.
Keep a person in the loop. Agents draft, people approve. That is the simplest and most effective risk policy there is.
Measure the time saved and the results. If a workflow does not save hours or improve a number within a month, drop it.
If you would like help, I set up AI workflows and AI reporting for small marketing teams, and you can see everything I offer on my AI marketing page. If you want to start with search, ask for a free AI visibility check.
