What used to be a fifty thousand dollar project is closer to five

The number genuinely changed

Custom software used to be priced by how long it took a person to write. A quoting tool that read drawings, or an intake system that connected three platforms, was months of work and it was priced accordingly. Fifty thousand dollars was a normal starting point, which is why most businesses under a hundred people never seriously considered it.

The work that took those months is the part that got faster. Not the thinking about your business, and not the testing, but the writing and wiring, which was always the bulk of the hours. Projects I would have quoted at fifty now land nearer five.

That is the whole reason a twenty-person shop can have something built for them at all. If you priced this out a few years ago and put it aside, the answer you got then is no longer the answer.

Where it does not apply

Anything where the hard part is understanding your business rather than writing the code. A process nobody has ever written down still has to be worked out with the people who do it, and that takes the time it takes.

The same goes for work touching money, patient records or safety, where the checking costs more than the building. Be suspicious of anyone quoting those like a simple job.

Two separate numbers

Building it is one cost, paid once. Running it is another, ongoing, and mostly what the AI models charge per use plus hosting.

Running costs surprise people by being small. A booking assistant handling a few hundred calls a month usually costs less to run than a phone line. The build is the real number.

What actually drives the build cost

How many systems it has to touch. One system is straightforward. Four, one of which has no proper integration, is where the hours go.

How messy the inputs are. Clean structured data is quick. Handwritten forms and inconsistent PDFs need more work and more careful handling of the cases where it is unsure.

How expensive a mistake is. Something that drafts an internal summary needs less scaffolding than something touching customer money or patient records. That difference is real work.

How to tell if it pays back

Take the hours the task consumes each week, multiply by what that person costs you, multiply by fifty. That is the annual cost of doing nothing. Compare it to the build.

Most worthwhile automations pay back in months, not years. If a project cannot show that on the back of an envelope, it probably should not happen, and I will say so.

Start small on purpose

The way to avoid a bad spend is not a longer proposal, it is a smaller first project. One workflow, working in days, running against your real jobs. If it does not save time you have lost very little and learned something concrete.

Getting a real number for your situation

Everything above is how the pricing works in general. What it costs at your business depends on your systems, your inputs and what a mistake would cost you, which is a fifteen-minute conversation rather than a proposal process.

I will give you a figure on that first call. No discovery fee, no deck, and if the payback maths does not work I will say so rather than scoping something anyway.

Questions people actually ask

Is five thousand a real number or a headline?

It is where a well-scoped single workflow tends to land now: one process, connecting systems that have proper ways in, with a person still approving anything that reaches a customer. Bigger jobs cost more and I will say so. The point is that the starting number moved by roughly a factor of ten, not that everything is five thousand.

Why will you not put prices on your site?

Because a number without scope is either meaningless or misleading. What I will do is give you a real figure on the first call rather than after three meetings and a proposal.

Do you charge a retainer?

Not to talk, and not for discovery. Ongoing support after a build is usually a small monthly arrangement, and it is optional.

What is the smallest project worth doing?

Something that saves a few hours a week. Quote follow-up and after-hours booking are common starting points because they are quick to build and easy to measure.