From $1,500 to $1.2M a year — how the spend actually scales.

AI spending does not scale with company size. It scales with how much of your operation you have handed over — and the steps are more predictable than the market lets on.

The ladder, and what moves you up it

$1,500 — the first assistant. One workflow, your own content, two weeks. What it buys is evidence. You move up when it works and your team asks for more.

$3,500 — a full build. A complete chat, voice, web or mobile product with your rules and your branding. You move up when the assistant needs to reach into systems it currently cannot.

$7,500 — integrated. SSO, your data, your systems. You move up when one workflow is no longer the bottleneck — several are.

$24K–$95K — a production agent. It stops answering and starts doing, with audit trails and human override. You move up when the agent proves a measurable number and the next workflow is obvious.

$120K–$500K a year — a programme. Sequenced agents across a function, with operations and evaluation running continuously.

To $1.2M a year — platform. An AI workforce across the organisation, licensed and operated, with quarterly business reviews and outcome-linked terms.

Do not skip rungs

The single most expensive pattern we see is an organisation jumping from nothing to a programme because a board asked for an AI strategy. It fails for a reason that has nothing to do with technology: nobody has yet learned what their own content, processes and people do when an AI system meets them. That lesson costs $1,500 at the bottom of the ladder and six figures near the top.

Why we publish all of it

Because a buyer who can see the whole ladder can plan three years instead of one purchase, and because a published ceiling makes the floor credible. $1,500 is not what we are worth; it is the smallest honest unit of work we can deliver properly.