Here is the product in one paragraph. For $999, you get a written report on your own operations: a leak ledger that estimates what your recurring manual work costs each month with every assumption shown, a three-tier fix plan that includes the things you should leave alone, and a fixed price for each recommended build. The fee is credited toward your first build if you commission one. That is what an AI assessment for a small business should be, and this article walks through exactly what mine covers so you can judge it before spending anything.
I am a practicing CPA. The habits in this assessment come from financial work: trace the inputs, document the assumptions, show the math. To be clear about what it is not, this is a consulting engagement, not an audit, review, or any other assurance service, and it does not touch your financial statements.
Why the assessment is paid, and what the credit actually does
A free consultation is usually built to establish fit and scope, and there is nothing wrong with that. The paid assessment does a different job: the product is the report itself, a portable written analysis you keep whether or not I build anything. You can hand it to your own staff or any other builder, and every number in it shows its work.
The honest version of the incentives: I hope the report leads to a build, and the structure reflects that. The $999 is applied as a one-time credit against your first build invoice if you commission one. What the paid model changes is that the deliverable has to stand on its own math either way, in writing, where you and your own accountant can check it.
Leak math: estimating your monthly process cost
The core of the assessment is what I call leak math. A leak is a recurring process cost worth measuring: hours your team spends on work a system could carry, rework from errors, follow-up that slips. Measuring is the point, because the measurement decides whether the right answer is automation, a configuration change, a process change, or leaving it alone.
Each leak is estimated the same way: the hours it consumes each month, valued at the rate you tell me those hours are worth, plus measurable rework where your numbers support it, with overlaps removed rather than counted twice. The output is a leak ledger. Here is the sample from my site, with illustrative numbers, so you can see the shape:
| Leak | Illustrative monthly process cost |
|---|---|
| Duplicate data entry | $1,880 |
| Manual invoicing and follow-up | $1,540 |
| Scheduling by hand | $960 |
| Report assembly | $1,120 |
| Rework from errors | $740 |
| Estimated monthly process cost | $6,240 |
Two honest caveats belong next to any table like that. First, those are sample figures, not a prediction about your business. Second, estimated process cost is not the same as cash savings. Freeing ten hours a week creates capacity; it becomes money when that capacity reduces a real expense or gets used on billable work. The report says which of your leaks are cash and which are capacity, because pretending they are the same thing is how automation projects end up disappointing people.
If a cost cannot be supported by your own numbers, it does not go in the ledger. I would rather hand you a smaller, defensible total than an impressive one you would take apart in five minutes.
What an AI assessment for small business actually examines
The working session goes where small businesses tend to lose hours: data typed into more than one system, how invoices get created and chased, how scheduling really happens versus how the software assumes it happens, who assembles the monthly reports and how long that takes, and where errors start and what it costs to unwind them.
You do not need clean books or any AI experience to start. The intake form asks about your tools, your team, and where the time goes; the session fills in the rest. I am looking for the gap between how work is supposed to flow and how it actually flows.
The three-tier fix plan
Every leak lands in one of three tiers. Tier A is recommended first: the strongest candidates once estimated benefit, implementation cost, risk, and dependencies are weighed together, each with a fixed price for a written scope. Tier B is recommended next, so budget follows the strongest case rather than my convenience. Tier C is leave alone: workflows where the volume is too low, the judgment content is too high, or a tool you already own covers it once configured. Tier C is in the report in writing, and some of the most useful lines are the ones that say do not spend money here.
On pricing, fixed means fixed for the written scope. If a scoped fix takes me longer than I estimated, the price does not move. If you ask for something beyond the written scope, that gets priced and approved before any additional work starts. Anything I build is tested from a documented deployment process on a fresh setup against the acceptance criteria we agreed to, before handoff, and the assessment carries the satisfaction guarantee stated on my site.
One more boundary worth stating in an article about AI: tools assist parts of my process, but I review the evidence, the assumptions, the calculations, and the recommendations, and I am responsible for what the report says.
Who gets the most out of this, and who should skip it
The best-fit businesses have repetitive administrative work: the same data moving between systems, recurring invoicing and follow-up, hours that can be reasonably estimated, and an owner willing to share how things actually work.
Who should skip it: if your business runs almost entirely on judgment calls with little repetitive process, there is not much leak surface to measure, and I will say so at intake before any fee. If you have a capable operations person with time to configure the tools you already own, you may only need a direction. And if you want someone to bless an AI purchase you have already decided on, the report may frustrate you, because Tier C does not care what either of us hoped it would say.
FAQ
What exactly does the $999 include?
The intake review, the working session, and the written report: leak ledger with assumptions shown, three-tier fix plan, and fixed-scope pricing for recommended builds. The report is the product; there is no obligation to buy anything after it.
How does the build credit work?
If you commission your first build, the $999 is applied once as a credit against that build's invoice. It is not a discount on the assessment itself, and it applies to the first build rather than stacking across projects.
How long does it take?
Typically inside two weeks from completed intake to delivered leak ledger, assuming the working session gets scheduled promptly and I have the information I asked for.
How are my team's hours valued?
At the rate you state for them. The report shows the rate next to every calculation, so if you think a rate is wrong, you can change it and see exactly how the total moves.
Is the estimated cost the same as what I would save?
No. The ledger separates cash costs from freed-up capacity, and the report is explicit about which is which. No figure in it is a guaranteed saving.
What do you need from me?
The completed intake form, one working session, and honest answers about how the work actually flows. If key information is missing, I ask for it before the clock starts rather than guessing.
The short version
An AI assessment for a small business should be a measurement, not a pitch. Mine is paid so the report is the product, credited once against your first build, and delivered as a written leak ledger where every number shows its math, including the lines that say leave this alone.
If you want that report on your own operations, start the intake here. Completing the form does not charge you anything; I read every submission first, and if I do not see enough measurable opportunity to justify the fee, I will tell you before you spend a dollar. You can also read how I approach the workflows themselves or see the assessment on the main site.