How pricing works
One platform.
Built around your operation.
Enterprise software is usually priced on things that have nothing to do with whether it works: how many people log in, how many modules were bought, which tier the AI sits behind. None of those describe an operation.
Claritus.One is priced on the operation itself — how much of it there is, how much of it has to be reconciled, and how much of it you want the platform to carry. That is a conversation, not a price list, and it is a short one.
What we do not price on
- SeatsHowever many people need to see it.
- ModulesThere are none to assemble.
- Feature tiersOne version of the platform exists.
- Agent runsNo meter on the AI Workforce.
The model everyone else uses
Per-seat pricing asks you to decide, in advance, who is allowed to know what is happening.
The rest of the model has the same defect. Modules assume you already know which parts of the operation are connected. Upgrade fees assume improvement is an event. And pricing AI as a premium tier assumes the automation is optional. In a network of thirty-plus sites, none of those assumptions survive a quarter.
Traditional enterprise software
Six places cost enters
- 01
Priced per user
The people who most need to see the operation become the ones you ration access for.
- 02
Priced per module
Scheduling in one contract, analytics in another, and nothing that reads both at once.
- 03
Analytics sold as its own product
The system that holds the data and the system that explains it have different renewal dates.
- 04
New capability, new line item
The roadmap arrives as an invoice.
- 05
AI as a premium tier
The part that does the work is the part you have to justify twice.
- 06
Cost rises with headcount
Growth becomes the thing the contract penalizes.
Every line above is a negotiation, a renewal date, and a reason for someone to be left off the license.
Claritus.One
One line
One platform
There is one thing to buy.
The AI Workforce, included
Agents are how the platform does the work, not a tier above it.
Operational intelligence, included
Reasoning is not a reporting add-on.
Optimization is continuous
Improvement is a property of the platform, not an upgrade cycle.
Everyone accountable can see it
Access is not a lever we price on.
Investment tracks complexity
It follows the shape of the operation, not the size of the org chart.
No tiers to compare, no modules to reconcile, and no version of this platform that withholds the intelligence.
What actually shapes it
Three dimensions, and a seat count is not one of them.
Investment scales with how much operation there is to understand and how much of it you want the platform to carry. Those two things are knowable in a first conversation, which is why the conversation is short.
- Centers in operation, and the ones opening
- Markets and regions, and how independently they run
- Shared staffing pools that cross site boundaries
- Whether the network grew organically or by acquisition
Notice what is missing. Not how many people log in, not which features you turned on, and not how many agents you ran last month. None of those describe the operation, so none of them set the price.
What is included
There is one thing to buy, and this is all of it.
Nothing on this diagram is a tier, a seat, or a later phase. Naming them separately would suggest you could buy one, and the platform does not work that way — the intelligence is what makes the automation safe, and the automation is what makes the intelligence worth having.
- 01Operations
- 02AI Workforce
- 03Operational Intelligence
- 04Analytics
- 05Integrations
- 06Automation
- 07Governance
- 08Security
- 09Executive Briefings
- 10Insights
- 11Optimization
Every capability above is present from the first day of the engagement. What changes over time is how much of your operation runs through them.
Sizing the opportunity
Your numbers, our arithmetic, and every multiplier shown.
We will not publish an ROI figure as though it were a result somebody achieved. What we will do is show the calculation. Move the five inputs below and the page multiplies them by the coefficients printed beside each line. Nothing is hidden, and nothing here is a forecast.
Your operation
Every site the platform would see.
Across the network, not your busiest site.
30% of net revenue. Clinical and front-desk, network-wide.
If you do not track it, most networks land near 6%.
First-pass denials as a share of billed revenue.
Derived from the above
- Annual visits
- 792,000
- Net revenue, at $175 a visit
- $138.6M
Where the value would sit
Annual range
Recovered labor capacity
$302K – $706K
your overtime spend ($2.52M) × 12%–28% caught before the hours are worked
Overtime found in payroll is a report. Overtime found on Tuesday is a decision that can still be made.
Scheduling optimization
$336K – $924K
your annual labor spend ($42.0M) × 0.8%–2.2% exposed to schedule-to-forecast drift
Most networks are quietly over-scheduled and under-scheduled in the same week, at different sites.
Revenue cycle opportunity
$1.46M – $2.91M
your denied revenue ($9.70M) × 15%–30% preventable at the front end
A large share of denials are not clinical disputes. They are a member ID typed wrong at 8:40 in the morning.
Throughput retention
$554K – $1.11M
visits at risk when waits run long ($2.77M) × 20%–40% of wait-driven walkouts recoverable
Assumes 2% of arrivals leave before being seen once a site runs past its wait standard.
Potential annual operational impact
$0–$0
1.9% to 4.1% of the net revenue implied by your inputs
This is the sum of the four lines above and nothing else. It is arithmetic on numbers you entered, not a forecast, not a benchmark, and not a result any organization has reported to us. Treat it as a way to decide whether the conversation is worth an hour.
Executive time returned
2,760 hrs – 6,440 hrs
your site count across an operating year (1,840 site-weeks) × 1.5–3.5 hours per site per week · roughly 1.3–3.1 full-time equivalents
Hours, not dollars. What an operator does with time returned is worth more than any rate we could apply to it. So it sits outside the total rather than inside it.
Assumptions you did not enter, stated rather than buried
- 360 open days per site, per year
- $175 net revenue per visit
- 2% of arrivals leave when waits run long
- 46 operating weeks for the hours line
What you are actually buying
Nobody has ever improved an operation by installing something.
Software gets deployed. Operations get changed, and that takes people on both sides who understand what the numbers mean and are willing to argue about them. We are building this with a small number of operators who want that kind of relationship, which is why the engagement looks less like a purchase and more like a working arrangement.
Scoped against your operation
The first conversation is about how your network is actually structured, which systems disagree with each other, and what you would want automated first. It is not a feature walkthrough, and there is no template underneath it.
The first ninety days are specified
What gets connected, what the AI Workforce is allowed to do, and what you should expect to see change — written down before anyone signs. The part we can be precise about is the part we commit to in writing.
Widening authority is earned, not scheduled
Agents start where you are comfortable and take on more once you have watched them work. No engagement assumes a level of trust that has not been established yet, because an operator who does not trust the system will route around it.
We will say no
Some operations are not ready, and some problems are not ours. Hearing that on the first call costs you an hour. Hearing it eight months into an implementation costs considerably more.
None of that is a reason to pay more. It is the reason the investment is worth discussing against your operation rather than being posted on a page for anyone to compare on the wrong axis.
The questions that come first
Seven answers, none of which are a number we made up.
Where the honest answer is “it depends,” the answer below says what it depends on.
Because a published number would be either wrong or empty. A figure that has to cover a twelve-site operator in one metro and a two-hundred-site network assembled from six acquisitions is a figure with no information in it, and the range required to make it safe would be wide enough to be useless to both.
The three dimensions above are answerable in one conversation. We would rather have that conversation and give you a real number.
The conversation