How AI Consulting Is Priced, and Why We Tie Part of Our Fee to Results

AI consulting is sold on four common models: hourly or time and materials, fixed-price projects, monthly retainers, and outcome-based fees. Each moves risk to a different side of the table. Outcome-based pricing is growing fastest at the top of the market: McKinsey's global managing partner, Bob Sternfels, told Harvard Business Review in January 2026 that "about a third of our revenues total are underwriting outcomes." This post explains each model, who carries the risk under it, and how Cosmic prices its work.

Hourly and time and materials

You pay for the hours worked. It's flexible when the scope is unclear, and it puts the risk of slow or unfocused work on you: the longer the work takes, the more the firm earns. If you buy this way, cap the budget and review the work often.

Fixed-price projects

You pay a set price for a defined deliverable. The risk of overruns moves to the firm, which is why fixed-price proposals come with detailed scope and change orders. The weakness for AI work is that the scope is often the thing you're trying to learn. A fixed price for a large, fully specified build can lock in a plan that the first user test would have changed.

Retainers

You pay monthly for ongoing access to a team. Retainers suit steady work, such as running and improving systems after launch. They say little about results, so ask what gets reported each month and against which number.

Outcome-based fees

Part or all of the fee depends on a result, such as revenue, cost per unit of work or hours saved. The firm carries some of the risk that the work doesn't pay off. Outcome pricing only works when both sides agree on the number and the baseline before the work starts (the core of revenue-focused AI consulting), and when the number is something the work can plausibly move within the engagement.

Pure outcome pricing is rare for good reason. A firm that's paid only on results has to price in everything outside its control, from market swings to your own team's priorities. Most outcome-based deals combine a base fee with a variable part.

How Cosmic prices AI implementation

Each sprint has a fixed price, agreed before it starts, plus an outcome-linked fee tied to the number set in discovery. The fixed part covers the work. The outcome-linked part means some of what we earn depends on the number moving.

We don't bill by the hour or by story points. A sprint is a fixed piece of work with a fixed price, so you know what the next few weeks cost before they begin, and you decide after each one whether to continue.

The number comes from discovery. Every AI implementation engagement starts with a free 30-minute call, then a fixed-scope discovery session of two or four hours with your executive sponsor, where we agree on the number, measure its baseline and choose where AI goes first. That session is also where we set the terms of the outcome-linked fee.

Questions to ask about any AI consulting quote

• What exactly does the price cover, and what happens when the scope changes?
• Is any part of the fee tied to a business number, and how is that number measured?
• How often will I see results reported against the baseline?
• Can I stop after any sprint or phase without a penalty?

For how to weigh these answers against everything else, see how to choose an AI consultancy. Clients in the region can meet us in person: we're an AI consultancy near Washington, DC. You can also read what revenue-focused AI consulting means in practice.