What Malaysian SME Owners Can Learn from Pocket FM’s AI Model

You may not run an audio storytelling platform, but you probably face a similar business problem: your customers expect more content, faster replies and more personalised service, while your small team is already stretched. Marketing posts, product descriptions, sales follow-ups, training materials and customer updates all compete for the same limited hours.

The useful lesson from Pocket FM is not that every business should replace people with AI. It is that a company can keep human judgement at the centre while using AI to handle repetitive production work. That distinction matters when you want to grow without making your operations chaotic or lowering quality.

TL;DR: Pocket FM reportedly uses AI for 93% of its catalogue and 99% of new content, while people still develop the ideas and stories. For Malaysian SMEs, the practical takeaway is to use AI as a supervised production assistant: standardise repeatable work, keep approval with your team and measure whether customers respond better.

The company says AI made content production approximately 80 times cheaper and reduced the time needed to create 100 hours of content from about a year to one day. Source: TechCrunch You should treat those figures as Pocket FM’s own claims, not as a guarantee for your business. The more important point is the operating method behind them.

What This Means in Plain Language

Pocket FM began as a platform for serialised audio stories. According to the report, human creators remain responsible for ideas and storytelling, while AI helps turn those concepts into finished audio content at scale. Its AI tools support creative writing and text-to-speech, using production data and listener engagement signals. Source: TechCrunch

For a small business, this is similar to creating a workflow where you provide the knowledge, rules and approval, while software prepares the first draft. You might give an AI tool your product specifications and brand tone, then ask it to prepare several social media captions. Your staff checks the facts, adjusts the language and approves the final version.

That is different from asking AI to run your marketing or customer service without supervision. The system can produce text quickly, but it does not automatically understand your promises to customers, local sensitivities, regulatory responsibilities or the context of a difficult complaint.

The best small-business AI workflow is not “people versus AI”. It is people deciding what matters, with AI helping them produce the repeatable parts faster.

Pocket FM also shows why content volume alone is not enough. The company said its library grew to more than 770,000 audio series and that it had more than 250 million listeners across over 20 countries. Source: TechCrunch For your business, producing more posts, messages or documents only helps if the work remains accurate and useful to the right audience.

How This Applies to Malaysian SMEs

1. Retailers can build a repeatable product-content system. If you sell through a shop, website, marketplace and social media, the same product information may need to be rewritten several times. You can maintain one approved product sheet containing the name, dimensions, materials, usage instructions, warranty terms and frequently asked questions. An AI workflow can then prepare a Shopee description, Instagram caption, WhatsApp reply and website summary from that source. Your team only needs to check that each version is accurate and suitable for its channel.

This is especially useful when you carry many products or add new stock regularly. You can also prepare versions in English, Bahasa Malaysia and, where appropriate, Chinese or Tamil. Human review remains important because direct translation may not match how your Malaysian customers actually speak. Your staff should confirm product claims, technical terms and local wording before publishing.

2. Service businesses can respond faster without losing a personal touch. A renovation contractor, tuition centre, clinic, repair company or professional firm receives similar questions every week. What are your service areas? What documents are needed? How long does a typical process take? What happens after a customer submits an enquiry? You can turn approved answers into a searchable knowledge base and use automation to draft replies.

For example, when a customer sends a WhatsApp enquiry, an automated workflow can identify the topic, suggest the relevant answer and alert a staff member if the question involves a complaint, unusual request or sensitive personal information. This reduces repetitive typing while keeping a person responsible for the final response. Do not allow an automated reply to make promises about delivery dates, refunds or outcomes unless those rules are clearly defined.

3. Manufacturers and wholesalers can make internal knowledge easier to use. Many Malaysian SMEs depend on a few experienced employees who know how to handle suppliers, machines, quality checks or common customer problems. If that knowledge stays in one person’s head, onboarding becomes slow and daily operations become fragile when that person is absent.

You can record short explanations, collect existing documents and ask an AI tool to organise them into draft standard operating procedures, checklists and training quizzes. A supervisor should verify every procedure, especially those involving machinery, safety, quality control or regulated products. The result is not a replacement for experience; it is a clearer way to share experience with new employees.

4. Food and hospitality businesses can plan content around real customer questions. Restaurants, caterers and small hotels often need regular updates about menus, operating hours, events, promotions and booking instructions. Instead of creating each post from scratch, you can prepare a monthly content calendar based on your actual services. AI can help turn one approved announcement into a Facebook post, Google Business Profile update, email message and staff briefing.

However, your team must confirm availability, allergens, operating hours and booking conditions before anything goes live. A wrong menu item or inaccurate opening time can create more customer frustration than an unwritten post.

A Simple Operating Model for Your Business

Start with work that is frequent, structured and easy for a knowledgeable employee to check. Avoid beginning with decisions that require judgement, confidential data or direct financial consequences. A small pilot can use one workflow, one owner and one approval checklist.

Workflow area	AI can help prepare	Human must verify
Marketing	Captions, content ideas and draft calendars	Claims, brand tone, promotions and images
Customer service	Suggested replies and question classification	Complaints, promises, refunds and exceptions
Operations	Draft checklists and document summaries	Safety, compliance and actual procedures
Sales	Follow-up drafts and enquiry summaries	Product fit, commitments and customer details

Keep a record of the original information used to create each output. If an AI-generated answer contains an error, you need to know whether the problem came from outdated product data, unclear instructions or an unsuitable prompt. Assign one person to maintain the source documents and review the workflow regularly.

Practical Takeaways

• Choose one repetitive workflow rather than trying to automate everything at once.
• Write down your approved facts, tone of voice, service rules and escalation conditions.
• Use AI for drafts, summaries, classification and formatting before giving it authority to act.
• Require human approval for customer complaints, sensitive data, safety matters and business commitments.
• Keep your product and service information in one maintained source of truth.
• Measure response time, correction frequency, customer satisfaction and staff hours saved.
• Review whether faster content actually produces more enquiries, repeat purchases or completed bookings.
• Train employees to identify confident-sounding errors instead of accepting every generated answer.

The Bigger Picture

Pocket FM’s reported growth shows how AI can become part of a company’s production system rather than a tool used occasionally by one employee. The company said its annualised revenue run rate rose from about $250 million a year earlier to $500 million, and it attributed the expansion partly to increased AI use, international growth and new user-generated content. Source: TechCrunch

That does not mean AI caused all of the growth, or that the same results will apply to a Malaysian SME. It does suggest that operational design matters. Businesses that document their knowledge, standardise routine work and keep quality controls in place will be better prepared to use new tools responsibly.

The long-term advantage may belong to companies with the clearest internal information, not simply the newest software. Your customer questions, approved processes, product details and service history are valuable operating knowledge. When that information is organised, your team can use automation more consistently across sales, support, marketing and administration.

Start small. Pick one process your team repeats every week, document what a good result looks like, add an approval step and review the outcome after a fixed period. If the workflow saves time without creating extra corrections or customer confusion, expand it carefully. That is how AI becomes a practical part of your business: not through blind automation, but through disciplined workflows that help your people do better work.

Ready to Streamline Your Operations?

Your business should run itself. AutoRunBiz deploys AI agents to automate your daily operations — WhatsApp orders, invoicing, customer follow-ups, and accounting. Book a free 15-min ops audit to see where automation fits your business →