What AI and automation really cost a small business

A fourth cost never appears on an invoice: your team's time during the changeover. Two weeks of slightly slower work while people learn a new habit is normal. Plan for it rather than being surprised.

Not the monthly fee. The value of the hours and the missed work. Multiply the hours a week the task consumes by your loaded hourly cost — wage plus taxes plus overhead, not just the wage — and multiply by fifty. Ten hours a week at $35 loaded is roughly $18,000 a year of capacity.

Then add the revenue side, which is usually bigger and almost always ignored. If you lose a handful of jobs a month by replying second, the value of replying first dwarfs any subscription. Work out what one average job is worth to you and how many you think you are missing. That single calculation decides more projects than any feature comparison.

• Aim for the first fix to pay for itself inside a few months, not a few years. If the payback math needs three years and an optimistic assumption, pick a different first project.
• Start with one process, not a transformation. A single finished automation earning its keep beats four half-built ones.
• Prefer a fixed price for defined work. Open-ended hourly arrangements on a vague scope are where budgets disappear.
• Insist the scope names what you'll be able to do differently when it's done — in your words, not technical ones.
• Ask what happens if it doesn't work. A serious partner will tell you how you'd know and what they'd do about it.

Run the cheapest audit there is before you spend anything: list every software subscription, the monthly cost, and who actually logged in last month. Overlapping tools, seats for people who have left, and subscriptions nobody can explain turn up nearly every time. In a lot of businesses that cleanup covers a meaningful share of the first project outright.

The second source is the capacity you get back. Hours returned to a team that was about to hire, or to an owner working Sundays, has a real dollar value even though it never shows on an invoice.

• Nobody can tell you what the change is worth in hours or dollars. If it can't be measured, don't buy it.
• The proposal is a technology list rather than a description of your process.
• The price is per hour with no ceiling and no defined finish line.
• You're being sold a platform when you described one problem.
• The plan starts with replacing systems that currently work fine.

We quote real numbers in conversation, not on a web page, for one reason: the same automation is worth very different amounts to a two-person shop and a twenty-person practice, and a public price list either overcharges the small business or undersells the work. What you get instead is a complimentary consultation where we look at your actual week, name the two or three places your money is going, and tell you what fixing each one is worth before anyone talks about fees.

If you want to see how the engagements are structured — free conversation, paid assessment, fixed-price implementation, ongoing optimization — that's laid out on how we work.