How pricing works
One set price. In writing. After mapping.
We do not publish fees, because a price without a map is a guess. After the mapping call you get one set price for the build, in writing, and it moves only if the scope does. What this page shows is how that number is built.
01How the set price is built
Five things move the number.
Every firm in this industry says pricing is custom, which tells you nothing. These are the five drivers we price from. Read them and you will know roughly where your build sits before you speak to anyone.
- 01ScopeOne process, one department, or the company layer. The map settles which, and the price follows the map rather than a tier chart.
- 02Number of workflowsEach workflow the system carries is designed, connected, run in review mode, and reported on. More workflows means more of each of those.
- 03Connections, and which way they runReading from a CRM is one thing. Writing back into it, the inbox, the phone line, and the payment tool, both ways, is another. Every connection is verified before anything depends on it.
- 04Approval complexityMessages, money, and sensitive decisions wait for a person. Each approval gate is designed, placed, and tested. Regulated work adds more of them, and they are usually the right call.
- 05Work after launchHosting, monitoring, the weekly report, and the next system. This is the separate monthly line, disclosed with the build price, and it scales with what we are running for you.
Why there are no numbers on this page
A range wide enough to cover everything above would be useless, and a narrow one would only be true for the company it was written for. So we do not print a figure we would have to walk back. On the mapping call you get a real number for your actual scope, with the reasoning attached, and if the scope is smaller than you feared, we say that too.
Model it yourself first
Five questions give you a tier and a timeline band. Never a price, and no email required.
02Scope
What is included, and what is not.
The exclusions matter more than the inclusions. Anything a vendor leaves unsaid at this stage is something you find out about in week eight.
Included in the set price
- The mapping: the operating map, the baseline, and the first system chosen by value
- The design in writing: what the system reads, what it does alone, where it stops, and which approval sits above each action
- The full build of every agreed system, into the tools you already run
- Every connection, verified before anything depends on it
- Review mode, approval gates, limits, escalation rules, and the audit trail
- Documentation and a runbook for every system we ship
- Training for the people who approve results and own the exceptions queue
- A direct line to the people on your build, not a ticket queue
The pod is sized from the map and priced into the set figure: two engineers for a focused system, a programme team staffed in phases for a larger build. Nothing is billed by the hour, and nothing is staffed all at once.
Not included
Third-party software and usage
Model usage, telephony minutes, and any software seats sit in your accounts and bill to you directly. We size them during mapping so there are no surprises, and you keep every vendor relationship.
Migrating your system of record
We build into the CRM you have. Nothing gets ripped out and nothing gets migrated. If a migration is what you want, it is its own project with its own scope and price in writing.
Work outside the written scope
Anything outside the agreed system stays outside the build. A new system mid-sprint gets its own scope and price in writing rather than being absorbed, which is what keeps a 30 to 90 day sprint from becoming twelve months.
03After launch
The separate monthly line.
Launch is the end of the build, not the end of the arrangement. Echelon stays on and runs it. The operation line is disclosed with the build price, runs month to month, and ends with thirty days' notice. No minimum term.
What the operation line covers
- 01HostingThe system runs on infrastructure we host and maintain. Your data stays in your accounts.
- 02MonitoringAlerting is wired to us. Most issues are fixed before you notice, and every action leaves a record.
- 03MaintenanceModels change, tools change, your pricing changes. The system is kept current as that happens.
- 04The weekly reportWhat ran, what it handled alone, what is waiting on you, and who owns it.
- 05The next systemBuilt on the layer already in place, scoped and priced in writing, without starting over.
The terms, in plain language
- The build
- One set price, in writing, after mapping. Fixed scope, the date the first build lands, and the launch date, in the same document.
- Changes
- The price changes only if the scope changes, and only in writing.
- After launch
- A separate monthly line for operation, disclosed with the build price. Month to month. No annual contract and no minimum term.
- Notice
- Thirty days, any time, no penalty.
- If you leave
- You keep every outcome the systems produced and all of your data, and we hand over the documentation and runbooks so your team knows exactly how the operation works.
- If something breaks
- Alerting is wired to us, so most issues are fixed before you notice. Every action leaves a record, so we can show you exactly what happened and why.
Our commitments, in writing
- One set price, in writing, after mapping. It moves only if the scope does.
- Review mode before any autonomy. Your team approves every result first.
- Narrow authority for proven routine actions only. The system never moves money on its own.
- Month to month after launch. Thirty days' notice, no minimum term.
- You own every result, the data, and the documentation.
- If at least one system we deploy does not measurably save time, reduce a cost, or support revenue, we keep working until it does.
04Ownership
You own every outcome. We own the upkeep.
Echelon builds and operates the AI systems we deploy for you. You license access to those systems for as long as we work together, and you retain full ownership of every outcome, result, and work product they generate: every lead answered, every quote sent, every dollar collected. We run the engine. Everything it produces is yours.
- Yours outright
- Every result the systems produce: leads, bookings, quotes, collections, transcripts, and the customer data behind them. Your data stays in your accounts.
- Yours to take
- The documentation, the runbooks, and the operating map we write for your build. If you leave, you leave knowing exactly how your operation works.
- Licensed to you
- The systems and the infrastructure they run on. We host, monitor, and improve them. That is what the monthly operation line buys, and why nothing rots after launch.
- Never ours to sell
- Your data does not train our models or anyone else's, and it is never pooled with another client's. Per-client isolation is a build standard, not an upgrade.
Said plainly because vagueness here is what buyers are right to distrust. We are not handing you a codebase to maintain, and we are not holding your operation hostage either. You own the results and your data. We own keeping the machine running.
05The guarantee
If at least one system we deploy does not measurably save time, reduce a cost, or support revenue, we keep working until it does.
Written into the agreement, next to the set price. It is the reason we map before we scope, and scope before we build.
06Getting to a number
Call, written scope, agreement.
How you get from this page to a figure. Nothing here is a sales sequence with a deadline attached.
- 01The mapping callThirty minutes. Bring one process your company still runs by hand. We map it with you on the call, and you leave knowing what we would build, what it would cost, and when it would go live.
- 02The scope and the price, in writingWithin days: the outcome, the scope, the boundaries, the acceptance criteria, one set price for the build, and the monthly line for operation, in one document. No tiers to compare and no deadline attached.
- 03Agreement and kickoffThe agreement matches the written scope exactly. Once it is signed, mapping starts and the direct line to the people on your build opens the same day.
07Commercial questions
What people ask before they send the agreement back.
Why don't you publish fees?
Because a price without a map is a guess. One process over a clean CRM and a company layer across several departments with a voice line on the main number are different builds. What we publish is what moves the number, above. What you get on the mapping call is a real figure for your actual scope, in writing.
How does the money work across the build?
Two lines. One set price for the build, agreed in writing after mapping. Then a separate monthly line for operation that starts when the system goes live and is disclosed with the build price. Third-party usage, such as model calls, telephony minutes, and software seats, sits in your own accounts and bills to you directly.
Do you bill by the hour?
No. Hourly billing charges you for the builder's uncertainty. A set price after mapping means you know the number before anything is built, and it moves only if the scope does, in writing.
What is the smallest sensible engagement?
One process, one clear outcome. Usually it is inbound: the leads that go unanswered. It is a smaller build, the first working build is in your hands within two weeks, it goes live in weeks, and it gives you something real to judge us on before the next system.
What changes the price after it is set?
Only the scope, and only in writing. If mapping finds the one process is really two, or you want a second system before the first is live, that is a scope change with its own price. Nothing gets absorbed silently and nothing gets billed by the hour.
Am I locked in after launch?
No. Operation is month to month with thirty days' notice and no minimum term. If you leave you keep every outcome the systems produced, all of your data, and the documentation and runbooks for how your operation works.
Who owns what you build?
Echelon builds and operates the AI systems we deploy for you. You license access to those systems for as long as we work together, and you retain full ownership of every outcome, result, and work product they generate: every lead answered, every quote sent, every dollar collected. We run the engine. Everything it produces is yours.
Do you have the capacity for something large?
Yes, in phases. Up to 20 to 30 engineers can work one programme, staffed against a written ramp plan with an architecture baseline agreed before the pods ramp. The first working build still lands in two weeks, and the first system is still live inside the sprint. Nothing is staffed all at once.
What if a system does not pay off?
If at least one system we deploy does not measurably save time, reduce a cost, or support revenue, we keep working until it does. Review mode means we see the numbers early, in the weekly report, while there is still time to change course. If a system is not returning what it was designed to, we say so and we fix it.
Your first department
Bring the process you still run by hand.
One mapping call. You leave with the system design, the price in writing, and the date it goes live. The first build lands in two weeks.
Size your project first
How big is the install?
Five questions, a tier, and a timeline band. Never a price, and no email required.
Size your project