Priced on the spend we govern, not the savings we grade.
A flat monthly fee on the AI spend you put under management. Not a cut of what we save you, not a charge per head. You know what Margin costs before you know what it saved, so the grader’s paycheck never rides on the grade.
It is also what lets us say what a savings-paid vendor cannot: keep the pricier route when what it lands is worth more than it costs.
Why no figure is printed here →Drawn, not plotted: no fee figure is published. The measured event sits under the chart.
- Our fee
- Flat, whether Margin saves you a little or a lot.
- A share of savings
- The model we refuse: it pays the grader more for a bigger number.
- Why it matters
- A flat fee has no reason to inflate the savings it reports.
The refusal is what the flat fee buys you: a route at 30.6% of the incumbent’s bill, refused on quality, our fee unchanged.
is_simulated=falseSource: provenance/autoroute_defended_savings_frac.json, committed, read at build time.
Never a share of the savings.
Do we make more money if you keep the wasteful agent? No. We grade parity on your own tasks, so we are never paid on our own grade.
Never per seat.
Hiring a teammate does not change your bill. The fee is set against what your agents spend, not the people you employ.
We have not earned the right to post a number yet.
With no customers yet, a posted figure would be a guess dressed as a fact. We set it against your actual spend, in a conversation, and move it as we learn.
The meter starts when your first metered outcome lands. Discovery and the scoping call are on us.
A pilot runs 45 days against one test we agree on before it starts: your provider’s invoice reconciles to the spend we measured. It passes or it fails, and you see which.
See what you would be paying for.
Watch the loop run live on open-source agents, every number tagged real or simulated, before we talk about a fee.
or email subh@trymargin.io directly; it reaches the founder, not a queue.