The New OS for Small Business: How AI Will Replace the Agency Model

I’ve been doing this for a while now, in fact, my whole life. The key is, what is “this”. Well, if I’m being honest, and I am, it has been being of service to those who paid me. Basically, I would look for a need and offer to fill it. This has taken me down a lot of rabbit holes and certainly has taught me to be careful what I agree to. Let’s just say, I have paid a lot of money for my education. So from shoveling snow out of peoples driveways, to collecting bugs off their backyard trees, to building complete business operations software to help them run their day-to-day operations, I have learned plenty about what it means to be of service. I already gave you the first, which is completely define the scope of work about to be done. This establishes expectations and avoids costly overruns. Clients have expectations, I vision in their heads. Even if this vision is undefined, it’s there. Your job is to get it out of their head and down on paper, otherwise, you find them saying “but I wanted that”, or “I expected that to be part of this.” I’m not saying that a customer is wrong, just that you are the pro and it is your job to understand the client, to be of service to the clients.

I’ve heard too many stories about a designer agreeing to the price, then when the job extended past what they understood it to be, they charged more. A “Afterall, the client pays me for my time and it’s not my fault if they change their mind.” I’m here to tell you yes, it is. You are either an extension of your client, like a tool, or you are the driver of the client’s vision, which needs thorough investigation.

Why am I telling you this? Well, because with the above attitude, I have learned quite a bit about businesses, how they make their money, the different approaches a small business uses going to market vs an enterprise, and most importantly, I’ve learned about people. What they want, how they use their businesses to achieve this, and then, I’ve learned that there are a lot of business owners out there who deliver half on their promises.

You see, this goes both ways.

I have run a digital agency for about 15 years and in that time I have worked extensively to hone my skills and those of my employees. I haven’t done this perfectly, but I have achieved a level of success in this area that makes me and what my company does very effective and even better than what is out there.

As an agency owner born from the digital marketing revolution, I have been exposed to all kinds of technology. Some of it worked and most of it didn’t. I saw thousands, literally thousands of companies pop up offering applications or do-it-yourself solutions that were supposed to change the companies who used them. Most of these have fallen by the wayside and some of the ones who stuck and stayed, SalesForce, for example, have grown to be multi-billion dollar companies.

The second thing I learned early on in the emergence and rocket-like trajectory of digital marketing is people love shiny new things. They gravitate towards them and then realize they don’t have the time to man them and support them. Even the simplest thing, like a chat bot (circa 2018) required a lot of setup. So a small business owner gets excited to get it on their website, they create an account, log in, and then begin setting it up. Before long they are frustrated or bored. The issue is they just wanted the solution. I’ve seen this especially in AI. Here now all of a sudden, a business owner is empowered with an intelligent tool that can compile reports, compile marketing plans, create images, tell you what’s wrong with your website, your people, heck, even you company. But the problem is, it’s fun and empowering until it’s not. And this is the dropoff.

So yes, these frontier models and tools being built off of them will definitely empower people to do more with less, but the point being missed here, and this is important, is the vision is being missed. It takes a certain level of expertise to uncover the end result one is shooting for and AI certainly won’t get you there, unless of course, you are using my company’s AI GIA. People think they want to do it all themselves and even believe they can, but eventually they will fall back on an expert to help them get there. So the second thing I learned is people want do-it-for-you solutions.

AI carries the promise of this. In fact, AI carries the promise of being able to do ALL of it for you. Which means, you are no longer necessary. However, we are a little away from this being reality, but this is the course gotcha! is on at our company. So let’s assume that this is a true statement: AI will one day run a business from sales to inventory or service execution, to customer support. If this is the end game, then we need to work backwards to where we are now, which is people subscribing left and right to dozens of tools that they have to prompt to get results, which, although they are proud of them, probably will underperform.

So we are at DIY moving to DIFY. At the DIY phase, clients are having a lot of fun generating reports and spawning Sora 2 videos and posting them on social media. Soon however, they will tire of this and look for someone to do it for them. This is where agencies come in. Agencies will transition into specialized prompters who orchestrate actions together to deliver outcomes for their clients. Clients will be relieved and agencies will be busy. But then enter the laws of scalability. Smaller agencies are going to struggle with the ability to generate enough outcomes to grow beyond small. They will hit a ceiling. Clients are demanding (rightfully so) and they have expectations. This takes time and consideration if it will be done correctly. When a small designer or agency hits a ceiling there are only so many choices; hire more experts, raise prices, or deliver poorer quality. Not poorer AI quality but less time planning and strategizing and more time executing. gotcha! It will be at this phase that the tables will turn. Companies like gotcha! Will then begin gobbling up these companies as clients because we will have built our system on research, strategy, planning first (as well as lifetimes of experience) and all our products will execute with such precision small agencies, even large ones won’t be able to keep up. 

This will be great for a while as our system grows bigger, the clients need for people in the loop will grow smaller. Eventually, very few or even none at all will be needed.

This is the evolution and whether it takes a few years or a few decades I am not waiting around.

Besides this, all these “solutions” available in the marketplace are questionable in my opinion. Most are GPT wrappers and the ones who do a little more work than that are just a step above, yet, almost none are considering the business and the business owners.