By Dr. Mike Donaldson

Most small business owners ask one question when they look at an AI tool: what does it cost per month? That is a fair question. It is also the wrong one. The monthly fee is only the entry point, and if you stop there, you are going to be surprised. The real AI total cost of ownership for small business goes well beyond the pricing page, and the gap between the two numbers is where most owners get hurt.

Here is how it goes. A tool looks useful. The monthly price seems reasonable. You sign up. Then the real costs start showing up.

Setup time. Staff training. Workflow changes to accommodate how the tool actually works. The productivity dip while your team figures it out. The hours spent cleaning up AI output that made it out the door before anyone caught it. The cost of rebuilding your process when the tool changes its interface without warning. The cost of undoing everything if the tool raises its price or shuts down entirely.

None of that is on the pricing page.

For a small business, that gap is especially dangerous. You do not have the budget cushion that a large company has. You are conservative with spending because you have to be. Every dollar you put into a tool needs to do something real. You cannot absorb six months of disruption while a team figures out a new system.

Token costs add another layer. Some AI tools charge by usage, not by a flat monthly fee. The cost scales based on how much you use the tool and how you use it. Most small business owners have never heard the term token cost before they sign up. Then the bill comes, and it does not match what they expected. That surprise is not a billing error. It is a failure of expectation-setting before the tool went live.

More than twenty years in manufacturing operations before starting Thotos gave me one lens I keep coming back to: in Lean thinking, you always evaluate total cost, not unit cost.

Here is a simple version of the idea. Say you are running a production line and a supplier offers you a cheaper part. The unit cost goes down, so it looks like a win. But that part does not fit quite right. Your team has to rework it. Your reject rate goes up. Your throughput drops. The unit cost went down, but the total cost went up. The savings on the part were real. The cost of everything downstream was also real.

AI is the same math.

The monthly subscription is the unit cost. The AI total cost of ownership for small business includes everything else: setup, training, workflow redesign, productivity loss during the adjustment period, the ongoing cost of reviewing and correcting AI output, and the cost of dependency — what happens when you build your operation around a tool and then have to rebuild when that tool changes.

A concrete example: take an AI writing tool, maybe ten or fifteen dollars a month. Your team starts using it to draft emails. The problem is the output does not sound like you. So your team spends five to ten minutes rewriting every email to match your voice. Every email. Every time. Now ask yourself: how long would it take to just write that email from scratch? If the honest answer is close to five minutes, you are not saving time. You are spending the same time, plus the subscription fee, plus the mental overhead of reviewing output you do not fully trust. Most owners never stop to run that math.

Pick one AI tool you are currently using. One tool. Then answer these three questions.

First: do you know what it actually costs per month? Not what you think it costs. What it actually costs, including any usage-based charges.

Second: do you know how many hours went into setting it up? Your time, your team's time, any outside help. Add it up.

Third: do you know what you would do if that tool stopped working tomorrow? Not eventually. Tomorrow.

If you cannot answer all three, you do not actually know what the tool costs you. That is a starting point, not a criticism.

Before adopting any new tool, ask for a full cost breakdown: monthly fee, setup time estimate, training time for your team size, and ongoing maintenance expectations. If the vendor cannot answer those questions, that is useful information on its own.

If the tool charges by usage, get a realistic estimate before you commit. One way to do this is the PERT formula: take your worst-case cost estimate, add four times your most likely estimate, add your best-case estimate, then divide by six. That calculation gives you a realistic projection that accounts for variation rather than assuming everything goes smoothly.

The step most owners skip: build a recovery plan. What happens if the tool changes, breaks, or goes away? If your answer is I have no idea, you have a dependency you have not accounted for. That dependency is a hidden cost sitting in your operation right now.

When owners evaluate an AI tool, they tend to compare it to not having it. That is a broken comparison because it leaves out everything that shows up after day one.

The right comparison is this: what does it cost to buy, set up, train your team on, maintain, and recover from when something goes wrong, versus what does it actually return? That is the total cost of ownership question. That is what tells you whether the tool is worth it.

Before looking at AI agents for my own work, I ran through this same process. I went through my own framework, looked at my processes, my data, my workflows. What I found shaped the direction I went and kept me from spending time and money on tools my foundation was not ready to absorb.

You have to know where you are before you know what you actually need.

The price of an AI tool and the cost of an AI tool are two different numbers. Knowing both is how you make a real business decision.

By Dr. Mike Donaldson

Every small business owner who has watched an AI agent demo wants the same thing. An agent that reads the inbox, drafts the follow up, files the paperwork, and clears a piece of the backlog before the day even starts. Before you look seriously at AI agents for small business, there is one question worth asking first. What happens when the agent is handed the same undocumented mess a new hire would inherit? That is exactly what happens in most first attempts, and it is why an agent gets blamed for a problem it did not create.

An agent handed a task does not clean up the environment around it. It works with what it finds. If the process is unclear and the files are scattered, the agent produces an unclear and scattered result, only faster and at a larger scale than one person could manage alone.

This mirrors a lesson from Lean manufacturing. A business does not put a robot at the end of a broken production line and expect the robot to fix the line. The line gets fixed first, then the robot runs the corrected process. AI agents work exactly the same way. They execute what is in front of them. A clear, documented process gets run well. An unclear process gets run at speed, and the mess that was already there does not shrink. It multiplies.

This is not a knock against the technology. The demos are not lying. The capability is real, and so is the burnout small business owners are trying to solve with it. Repetitive tasks, follow ups that fall through, decisions that should not require an owner personally, all of that is genuinely a good fit for an agent, provided the work handed to it is already defined.

Before I built my first two agents, I sat down and looked honestly at my own file structure. What I found was a system built for a person, specifically for me, not for a machine. Folders were nested six and seven levels deep. File names made complete sense to me and would have meant nothing to an agent trying to locate the right document. I knew exactly where everything lived, and that was the problem. My knowledge was the system.

I made three changes before I built a single agent. First, I capped folder depth at four levels. If a file could not be found within four clicks from the top, the structure itself was wrong, not the person searching. Second, I created a firm naming convention for every file type, a rule rather than a guideline, with no exceptions. Third, I wrote a plain language file of agent rules that any agent in my environment has to follow: where files belong, what to do when something is unclear, and one rule that never bends. The agent can archive a file. It cannot delete one.

This connects to what I call the Digital Twin pillar at Thotos, a documented, navigable version of the business that exists outside of any single person's head. When that documentation is solid, an agent can operate inside it. When it is not, the agent is guessing, and it will not tell you that it is guessing.

If you are ready to put an agent to work, follow this sequence in order.

Pick one process. Not the whole operation, one task that repeats, takes real time, and has a clear output you could describe completely to someone with no background in the work.

Document that process fully. Write every step, every file involved, and every decision point. Write it for a person on their first day who knows nothing about how the work actually gets done.

Validate the process before you touch a tool. Read each step aloud and do only what it says, nothing more. Note every gap: assumed knowledge, a missing instruction, a file referenced but never named. This step gets skipped more than any other, and it is the one that determines whether the agent works.

Audit the environment the agent will operate inside. Check folder depth, naming consistency, and whether the data involved is clean and accessible. Write down the rules the agent has to follow before it runs its first task.

Documenting a process and validating one are different tasks, and the difference is where most first attempts at AI agents for small business fail. Documentation is writing down what you believe happens. Validation is walking through the process exactly as written and finding out what is actually true.

A person doing a task fills in gaps automatically, because they have done it before and understand the context around it. An agent does not have that context. It either follows exactly what is written, or it stops, or it guesses, and a guess can travel through several steps before anyone notices the output is wrong. The validation walk-through feels slow and a little tedious. It is also the fastest path to an agent that actually works, because everything built after it stands on a foundation that has already been tested.

An AI agent will always do exactly what the process says, not what the owner meant, and the businesses getting real value from agents right now are the ones who did the unglamorous work of writing that process down and testing it before they ever turned an agent loose on it.