Most agencies won’t tell you. So here are real numbers, and the reasoning behind them.
Short version: a single automated workflow, built properly and handed over, costs between $6,000 and $25,000 in the US market. Below $6,000 you are usually buying a script. Above $25,000 you are usually buying a project that should have been split into two.
What you actually pay, by route
A freelancer on an hourly rate — $50 to $150/hour. Cheapest per hour and genuinely fine for a contained job: one integration, one script, one report. The risk is not skill, it is continuity. When the person moves on, so does the only copy of how the thing works.
A specialist studio — $6,000 to $25,000 per workflow, fixed price. You are paying for scoping, documentation and someone accountable when it breaks at 2am. Worth it when the workflow touches more than two systems or when being wrong has a cost.
A large consultancy — $50,000 and up. Appropriate for enterprise compliance and procurement requirements. For a 40-person company automating lead intake, it is the wrong shape of engagement.
In-house — $8,000 to $15,000 of loaded salary cost for a mid-level engineer spending three to four weeks on it. This is the option people forget to price. It is often the right answer, and the real question is whether that engineer has four weeks that are not already spoken for.
The number nobody quotes: what it costs to keep
Every automated workflow carries roughly 10 to 20% of its build cost per year in maintenance. APIs change, credentials expire, a vendor deprecates a webhook.
This is the line that turns a good automation into a bad one. A $12,000 system that saves $30,000 a year is excellent. The same system, unmaintained, silently failing for two months before anyone notices, is worse than the manual process it replaced — because at least the manual process failed loudly.
Budget for it, or do not build it.
How to tell if a workflow is worth automating
Rough arithmetic, and it is usually enough:
- Hours per month × loaded hourly cost × 12 = annual cost of doing it by hand.
- If that number is under $5,000, automation probably will not pay back. Write a checklist instead.
- If it is over $20,000, the question is not whether to automate but which part to automate first.
- Add the error cost. A missed lead or a wrong invoice usually costs more than the time saved, and it is the part people leave out of the math.
A workflow taking 8 hours a month at a $60 loaded rate costs about $5,760 a year. A $10,000 build pays back in under two years — acceptable, not exciting. The same workflow at 20 hours a month pays back in under eight months, and that is a clear yes.
Why fixed price beats hourly for this work
Hourly rates put the risk of a bad estimate on you. If the integration turns out to be harder than it looked, your invoice grows.
Fixed price puts that risk on whoever wrote the estimate — which is the correct place for it, because they are the one who understood the problem well enough to quote it. The trade is that scoping has to happen first and properly, which is why serious studios charge for the scoping phase rather than giving it away.
What should be in the price
Before you compare quotes, check they include the same things:
- Monitoring and alerting. Otherwise you find out it broke from a customer.
- Retries and error handling. Every integration fails sometimes; the question is whether it fails safely.
- Documentation. A diagram and a runbook, so a different engineer can pick it up.
- Ownership. Code, credentials and infrastructure in your name — not in an agency account you rent access to.
A quote missing these is not cheaper. It has moved the cost somewhere you cannot see it yet.
What we charge
We publish ours: $2,500 for a two-week Automation Map that scopes the work, credited against the build. $8,000 to $18,000 for a first system, fixed scope. From $6,000 a month for ongoing partnerships.
If the arithmetic above says your workflow is not worth automating yet, we would rather tell you that than sell you a build.