The short answer
- A single automated workflow typically costs $1,500 to $6,000 to build, plus $20 to $150 a month in software.
- A connected set of workflows across quoting, scheduling and follow-up usually lands between $8,000 and $25,000.
- Cost is driven far more by how many systems have to talk to each other than by the complexity of any one step.
- Payback is hours saved per month multiplied by loaded hourly cost, plus revenue recovered from faster response. Most useful builds pay back in four to nine months.
- If a workflow runs fewer than a handful of times a week and takes under five minutes, it is usually cheaper to leave it manual.
The ranges, stated plainly
Automation pricing feels opaque because most firms quote a project, not a unit. The useful unit is a workflow: one triggered sequence with a defined start, a defined finish, and a place the result lands. Priced that way, the numbers become comparable.
A single workflow — a form submission that creates a CRM record, notifies the right person, and sends a confirmation — is generally a $1,500 to $3,500 build. Add conditional routing, a document generated from a template, or a two-way sync with an existing system, and the same workflow moves to $3,500 to $6,000 because it now has failure modes that must be handled rather than hoped away.
A department-level build — quoting, scheduling, invoicing handoff and follow-up all connected — is usually $8,000 to $25,000. That range is wide because it is dominated by integration surface, not by step count. Two workflows touching one CRM cost less than one workflow touching a CRM, an accounting package, a field app and a legacy spreadsheet.
Ongoing cost is smaller and easier to forecast: $20 to $150 a month for the automation platform and connectors on a typical small business volume, plus whatever the underlying tools already cost. Budget a maintenance allowance too. APIs change, vendors deprecate endpoints, and a workflow nobody maintains eventually fails quietly, which is the expensive kind of failure.
The five things that actually move the price
When two quotes for what sounds like the same project differ by a factor of three, one of these is the reason. Ask about each before comparing numbers.
- Integration count. Every additional system is an authentication method, a rate limit, a data model and a support relationship. The second system roughly doubles effort; the fourth does not quadruple it, but it is not free either.
- Data quality. Clean, consistently formatted records automate cheaply. Records with three spellings of the same customer name need a reconciliation step, and that step is real engineering.
- Exception handling. The happy path is maybe 40% of the work. What happens when the payment fails, the address does not validate, or the technician marks a job complete twice determines whether the build is trustworthy.
- Compliance. Anything touching health information, payment data or signed agreements needs auditability, retention rules and access controls. This is not optional and it is not fast.
- Change tolerance. If the process itself is still being argued about internally, you are paying to automate a moving target. Settling the process first is the cheapest thing you can do.
How to calculate payback before you spend
Do the arithmetic yourself; it takes ten minutes and it will disqualify half your candidate projects, which is the point.
Start with hours. Count how many times the task runs per month and how long it takes end to end, including the context switch back to real work. Multiply by loaded hourly cost — salary plus taxes and benefits, typically 1.25 to 1.4 times the base rate. That gives you the labor recovered per month.
Then add the revenue side, which is usually larger and almost always ignored. If automated routing takes your average response to a new inquiry from four hours to four minutes, more of those inquiries convert. Take your monthly inquiry volume, your current close rate, and a conservative improvement — two or three percentage points, not the vendor's number — and price it at your average job value.
Divide the build cost by monthly labor plus recovered revenue. Under nine months, build it. Nine to eighteen months, build it only if it also removes an error class that costs you customers. Over eighteen months, leave it manual and revisit when volume grows. Our automation ROI calculator runs this same model if you would rather not do it on paper.
Where the cheap wins hide
The lowest-cost, highest-return automations are almost never the ones people ask for first. They cluster around handoffs — the moments where information moves between two people or two systems and someone retypes it.
Instant inquiry acknowledgment and routing is the standard example. It is a small build, it runs constantly, and it changes a metric that directly drives revenue. Quote and proposal generation from a template is the second: high value per run, low complexity, immediate time saving for whoever currently rebuilds the same document each week.
Reporting is the third and the most underrated. If someone spends two hours every Monday assembling numbers from three places, that is 100 hours a year of skilled time recovered by a build that rarely exceeds $2,500.
By contrast, automating a nuanced judgment call — pricing an unusual job, deciding which client gets escalated — tends to cost the most and deliver the least, because the rules are not actually rules. Automate the plumbing around the judgment and leave the judgment to a human.
Costs people forget to budget for
Three line items appear in almost every honest automation budget and almost no informal estimate.
First, process documentation. Someone has to write down what the current process really is, not what the handbook says. Half a day to two days, and skipping it is how you end up automating a process nobody follows.
Second, staff training and the adoption dip. For the first two or three weeks, output goes down while people learn the new path. Plan for it and it is a footnote; ignore it and it becomes the reason the build gets abandoned.
Third, monitoring. A workflow needs to tell you when it fails. An alert into a channel someone actually reads costs almost nothing to add during the build and is close to impossible to retrofit attention for later.
Add roughly 15% to any build estimate for these and your budget will survive contact with reality.
Questions people ask next
One workflow typically costs $1,500 to $6,000 to build and $20 to $150 a month to run. A connected set of workflows across several departments generally lands between $8,000 and $25,000, driven mostly by how many systems have to exchange data.
For repetitive, rule-based work, usually yes. A build that recovers 20 hours a month is roughly equivalent to a quarter of an administrative hire but costs once rather than annually. For work requiring judgment or relationship handling, hiring is still the better answer.
Platform and connector subscriptions of $20 to $150 a month for typical small business volume, plus a maintenance allowance for API changes and process adjustments. Budget a few hours a quarter rather than a fixed retainer unless the workflows are business critical.
A single workflow is usually two to three weeks from scoping to live. A department-level build is six to twelve weeks, with most of that spent on process definition and testing rather than on building the connections themselves.
Four to nine months for well-chosen workflows. If your estimate exceeds eighteen months, the task is either too infrequent or too complex to be a good first candidate.
Yes, and you should. Building one workflow first proves the data is clean enough and gives your team a low-risk introduction. Ask that the first build use a platform and data model that will extend, so the second workflow is cheaper than the first rather than a restart.
Want this applied to your business rather than explained? Book a strategy call or read the rest of the guides.