This is the question almost everyone asks and almost nobody answers directly. You get "it depends" or a vague range so wide it's useless, or a proposal you have to sign an NDA to see.
Here's the honest version.
Yes, it depends. But it depends on specific things — and I can tell you what those things are, what the ranges actually look like, and how to think about whether the investment makes sense before you spend anything.
Why the Range Is So Wide
"How much does automation cost" is a lot like asking "how much does construction cost." The answer for a bathroom refresh and the answer for a custom home are both technically correct. Completely different numbers.
The same is true here. A simple automation that fires when a form is submitted and sends a notification email lives in a completely different cost category than a full CRM build with document generation, e-signature collection, and AI-assisted follow-up workflows.
What actually drives the cost:
**Complexity of the process.** The more steps, conditions, and edge cases involved, the more time it takes to build, test, and document.
**Number of platforms involved.** Connecting two tools that already have a pre-built integration is fast. Connecting five tools where two of them require custom API work is not.
**No-code vs custom.** Configuring an existing tool is faster and cheaper than writing something from scratch. Custom code gives you more control and eliminates platform fees, but it takes more to build.
**What happens when something goes wrong.** Error handling, logging, and recovery paths add time. Skipping them makes the build cheaper and makes your future more expensive.
**Ongoing maintenance.** A build isn't finished when it launches. Platforms update. Processes change. Someone has to maintain it. That's either factored into the project or it becomes a surprise later.
Real Ranges (Not Fake Ones)
These are honest ranges based on real work. Your situation will land somewhere in here based on the variables above.
Simple automation (one or two platforms, straightforward trigger-action logic):
$300 to $1,500 in build time. Think: form submitted, record created in CRM, notification sent. Fast to scope, fast to build, low maintenance.
Mid-complexity automation (multiple platforms, conditional logic, some custom configuration):
$1,500 to $5,000. Think: lead comes in, gets routed based on criteria, triggers a sequence, creates tasks for the team, sends documents for signature. More moving parts, more testing required.
Complex build (full workflow systems, document pipelines, CRM configuration, multi-step processes):
$5,000 to $20,000+. Think: a client's entire intake and onboarding process automated end to end, or a full CRM built and configured around how a specific business actually operates. This is a significant project with significant return when done right.
Custom app (replacing a SaaS subscription with something you own):
$3,000 to $30,000+ depending on what it does. A simple internal tool on the lower end. A full web application that replaces a $500/month platform on the higher end. You pay once. The monthly bill disappears.
AI integration (chatbot, document processing, AI-assisted workflows):
$2,000 to $15,000+ for implementation, plus ongoing model costs depending on usage. Highly variable based on what the AI is doing and how it's connected to your systems.
MCP and agent builds (AI that takes autonomous action in your systems):
This is the frontier right now. Costs vary widely based on scope, monitoring requirements, and complexity. Expect this category to be in the $10,000+ range for anything production-ready.
Hourly vs Project Pricing — What the Difference Means for You
Some vendors price by the hour. Some price by the project. Both have tradeoffs.
**Hourly** is transparent. You see exactly what you're paying for. The risk is that a poorly scoped project runs longer than expected and the bill grows with it. Ask for estimates and check in regularly.
**Project pricing** gives you a fixed number to budget against. The risk is that scope creep (changes mid-project that weren't in the original scope) leads to change orders that add up. Read the contract carefully and make sure scope is defined clearly before you sign.
We price hourly. Not because it protects us — it means we have to be efficient and accurate with our time estimates. But because it keeps the relationship transparent. You know what the work actually takes.
The Cheap Build Problem
There's a version of this where someone quotes you $500 for something that sounds like it should cost $3,000.
Sometimes that's because it's genuinely simple and the high quote was padded. Sometimes it's because the builder is cutting corners you won't discover until it breaks. Sometimes it's because they're scoping something simpler than what you actually need, and the scope expands after you've committed.
The $500 build that requires $2,000 in fixes six months later wasn't cheap. The $3,000 build that runs cleanly for two years probably was.
Ask what's included in error handling. Ask what the documentation looks like. Ask what support looks like after launch. The answers tell you a lot about what you're actually buying.
How to Think About ROI Before You Commit
The cost question gets a lot clearer when you put it next to the return.
Figure out what you're actually spending on the problem right now. If the process takes four hours a week across your team at an average cost of $25 per hour, that's $5,200 a year in labor for that one task. A $3,000 build that eliminates it pays back in seven months and then runs for free.
That math works even more clearly when the cost is client experience, not just labor. Delays in follow-up. Errors in documents. Inconsistent onboarding. Those have costs that don't show up on a timesheet but show up in client retention.
The question isn't "is this expensive." The question is "what is the problem costing me right now, and does fixing it justify the build cost?"
Sometimes the answer is yes, obviously. Sometimes it's not yet — the volume isn't there or the process is still changing. That's a legitimate outcome of actually running the numbers.
Why We Start With an Audit
This is why the Automation Audit exists before any build conversation.
We want to understand what you're working with before we give you a number. Not because we're being evasive. Giving you a number before we understand the scope would be irresponsible. The right build for your situation might be simpler and cheaper than you assumed. Or more involved. We don't know until we look.
The Audit gives you a clear picture of what's there, what the opportunities actually are, and what it would realistically cost to address them. You get that picture whether or not you hire us to build anything.
That's the lowest-risk way to start this conversation.
Michelle Onizuka is co-founder and Systems Architect at Onizuka Studio. She has been pricing and building automation projects for small and mid-size businesses since 2013.
[Start with an Automation Audit](/automation-audit/) and know what you're working with before you commit to anything.