Two Different Games

The Old Deal

When I was running a social media and web agency in university, the deal looked like this.

Charge a client $3,000 AUD to build a website. Hire a developer for RM 3,000. Wait one week for the prototype. Wait another three weeks for the full build. Deliver. Collect the remaining 60%. Move on to the next one.

That was the business. Every project was a small operation. Every client required a chain of people. Every delivery was a coordination exercise.

It worked. I made money. But I was never more than one missing person away from a problem.

What Running an Agency Actually Teaches You

Nobody tells you this when you start an agency: the business model is less important than the operational discipline it forces on you.

I learned cash flow the hard way. Early on I took a client deposit, paid my video editor the full amount upfront, and the editor didn't deliver. I had to put in my own money to hire someone else to save the project. After that, I restructured every freelancer arrangement. Twenty percent upfront. Eighty percent on delivery. The same protection my client had over me, I built for myself over them.

I learned that a client paying late is not the same as a client not paying. You still have freelancers to pay. You still have timelines to hit. Managing the float between what comes in and what you owe is a skill nobody teaches you in school.

I learned that personal brand is not vanity. In a services business, you are the product before the product exists. People hire who they trust. Trust comes from visibility. Visibility comes from showing up consistently before anyone needs you.

These weren't lessons I read anywhere. They were lessons I paid for with my own money and my own mistakes.

The Problem With Scaling an Agency

At some point you hit a wall that every agency owner hits.

To make more money, you need more clients. To serve more clients, you need more people. More people means more management, more payroll, more coordination, more things that can go wrong. The business doesn't scale. It just gets heavier.

If an agency contract earns you RM 100,000, you might spend RM 40,000 to 50,000 on the people who actually do the work. Another RM 100,000 means another RM 40,000 to 50,000. The revenue grows but so does the cost, in lockstep, every time.

An agency can make you comfortable. It can give you a decent life. But the ceiling is lower than it looks, because every ringgit of new revenue drags a significant cost behind it.

SaaS is a different game entirely. The first months are hard and expensive. Building the product, acquiring the first customers, fighting churn. But once the product works and customers stay, each new customer costs almost nothing to serve. The margin improves as you grow, not in spite of it.

That asymmetry is the whole difference.

The Shift I Watched Happen

Something changed in the last two years that I don't think enough people have stopped to fully process.

Building used to require a team. A developer to write the code, a designer to make it look right, a project manager to hold it together. The barrier wasn't just money. It was coordination. You needed multiple skilled people moving in the same direction at the same time, which meant hiring, managing, waiting, and hoping nothing fell apart.

Then the tools changed.

I can now build a working website prototype in fifteen minutes. Deliver a full project in a week. The total cost in software subscriptions is under RM 1,000 a month. What used to take a team a month now takes one person a week, at a fraction of the cost.

This isn't a small efficiency gain. The economics of building collapsed. The barrier to creating real software dropped dramatically and quickly. Things that work, things clients can actually use.

The Moment I Felt It Most Clearly

I went to get a company stamp made for my new business.

While waiting, I looked at the stamp maker's website on my phone. It was outdated and needed rebuilding. I sat down at the cafe next door and built them a demo website. Fifteen minutes later I walked back in, collected my stamp, and asked to speak to someone in management.

That demo led to a proposal. The proposal didn't convert. But that's a different story.

What stayed with me was the fifteen minutes. That a working, presentable website could exist in the time it takes to drink a coffee. Three years ago that same website would have cost me RM 3,000 and a month of someone else's time.

Something had fundamentally shifted. I was lucky enough to see it clearly because I knew what it used to cost.

Why Doing It the Hard Way First Mattered

Here is what I think gets missed in the "just use AI tools" conversation.

The tools amplify what you already know. If you don't know how to manage a client, the tools don't teach you. If you don't understand cash flow, the tools don't save you when a client pays late. If you can't sell, the tools don't close deals for you.

Running an agency first gave me the business fundamentals that the tools now run on top of. I know how to price a project because I've bled from pricing it wrong. I know how to structure a freelancer payment because I've lost money getting it backwards. I know how to show up consistently because I've seen what happens when you disappear.

The tools made building faster and cheaper. The agency made me someone who could use them properly.

SMMA taught me how to run a business. AI tools gave me the ability to build one.

The Honest Comparison

If you're starting today and you want money fast, an agency works. You'll make decent money. You'll learn real things. You'll also hit the ceiling sooner than you expect, and the weight of managing people and projects will become the main job faster than you'd like.

If you're building for the long term, software is the better model. Harder to start, slower to revenue, but the economics get better as you grow rather than heavier. And right now, the cost of building has dropped far enough that the gap between "I have an idea" and "I have a product" is smaller than it has ever been.

I'm not saying everyone should stop running agencies. I'm saying I knew, once I felt what the tools could do, that the game I wanted to play was different from the one I'd been playing.

The fifteen minutes at the cafe made it obvious.