The total cost of risk
The monthly line item is the part everyone compares. It is rarely the part that costs the most.
The question is not 'what does it cost' — it is 'what do I risk'
When a shop evaluates an AI front office, the instinct is to compare monthly figures. But the monthly figure is the one cost you can stop paying. What you cannot stop paying is everything you committed before you knew whether it worked.
We think the honest way to compare vendors in this category is by total cost of risk — five things, of which price is one.
The five components
The setup fee
Money spent before a single call is answered. It is the purest form of risk in this category, because it is spent whether or not the thing works.
The lock-in
How long you are committed before you can walk away if it is not working, and what leaving costs you.
The evaluation cost
The weeks of your own attention spent working out whether it is helping — which is real cost, and is usually unbudgeted.
The measurement gap
If nobody can tell you what changed against your own prior numbers, you are paying indefinitely on the strength of a feeling.
The opportunity cost
Every month spent on a system that is not recovering calls is another month of the calls still being missed.
What the category currently does — sourced and dated
We are only going to say things about other vendors that we can show you the receipt for. So here is the one structural fact we verified ourselves, with the method written down.
The best-funded vendor in this category does not publish a price on its website. To find out what it would cost you, you book a demo.
We are deliberately not quoting a competitor’s prices, setup fees, or contract terms — not because we have not heard figures, but because a vendor that publishes no prices gives us no authoritative source to cite, and repeating an unsourced number about someone else’s business is exactly the behaviour this page is arguing against.
Why an unpublished price is itself a risk
A price you have to book a call to learn is a price that can be different for you than for the shop down the road. That is a legitimate way to run a business. It is also a cost: it means you cannot compare without entering a sales process, and you cannot know whether you were quoted well.
It also tends to travel with the other four components — the implementation fee, the term, and the absence of a real trial — because all four exist to move risk from the vendor to the buyer.
Where we put the risk instead
We built our commercial model so that most of the risk sits on our side of the table. Concretely, and without a single figure on this page:
- No implementation fee. Nothing is spent before a call is answered.
- A free recovery audit first. Send your last ninety days of call logs and we will tell you what you are losing, using your own numbers. You keep the report whether or not you ever buy anything.
- A genuinely free first period before anything is billed.
- A published model. A low base plus a fraction of what we recover. If we do not recover, the fraction is small by construction.
- A measurement you can audit — against your own prior ninety days, not an industry average.
We describe our terms on our pricing page. You do not have to talk to anyone to read it.
What we will not claim
We will not tell you we are cheaper than a vendor whose price is not public. We do not know that, and neither does anyone else quoting it at you.
We will not tell you our system caused a revenue increase. We measure against your own prior ninety days and we describe that as a comparison, because that is what it is.
And we will not tell you we are the right fit for a large multi-brand operator with a procurement department and a certification checklist. We are built for the shop that is losing calls this week.