What automation actually costs (and what it pays back)
Honest pricing for small-business automation: what drives cost, typical project ranges, how to figure payback in months, and when we'll tell you not to build it.
7 min readBy Nick Hayghe
This isn't a sales pitch dressed up as a guide. It's the math.
If you've been burned by a software pitch before, you have every right to be skeptical. Most automation consultants talk about "efficiency gains" and never show you the invoice. So let's start there: what it costs, what drives the price, and how to tell whether it pays back. If the math doesn't work, we don't build it. That's the whole pitch.
> lead: captured → qualified → booked ✓ 11:48pm
That's the kind of line we're going after — real outcomes, timestamped. Not vibes.
Why most automation pricing is vague
Ask ten consultants what an automation project costs and you'll get ten answers, all different, none useful. There's a reason: automation isn't one thing. A two-app connection between your booking tool and your invoicing tool is a fundamentally different job than an AI system that reads incoming quote requests, scores them, and routes the good ones to your best tech.
Anyone who gives you a flat number before seeing your systems is guessing. Anyone who won't give you a range after seeing them is hiding.
Here's what we'd rather do: show you the levers. Then you can price anything, whether you hire us or not.
What actually drives the cost
Forget the AI for a minute. The cost of an automation project comes down to five things: 1. Number of systems involved. Two tools talking to each other is simple. Five tools with conflicting data formats is not. Every system you add multiplies the testing surface. A project connecting QuickBooks and your scheduling app costs less than one tying together QuickBooks, your scheduler, your CRM, and your dispatch board — because each pair has to be tested against the others. 2. Edge cases. The happy path is easy. It's the exceptions that cost money: the invoice with two line items from different jobs, the booking that changes three times, the form where the customer writes "call me sometime Tuesday-ish" in the address field. If your business has a lot of "well, except when…", expect that to show up in the price. 3. Data cleanup. Automation is only as good as the data it reads. If your customer records have five spellings of "Main Street" and three duplicate entries for your biggest account, somebody has to clean that up before anything reliable can run. Sometimes that's a day of work. Sometimes it's a week. This is the line item people forget — and it's often the one that matters most. 4. How much AI is really involved. Plain workflow automation — moving data between tools on rules you define — is the cheapest tier. AI enters when the work requires judgment: reading a messy email and pulling out the job details, drafting a quote in your voice, deciding whether a lead is worth a fast response. AI adds cost because it has to be trained, tested, and monitored for the times it gets it wrong. More judgment, more money. 5. Who owns the result. A contractor who hands you a system you can run and change yourself prices differently than an agency that keeps you on a retainer forever. Ask about this directly. If the quote is cheap but you're locked in for life, that's not cheap.
Typical project ranges
These are illustrative ranges based on what we've seen in the market for small businesses — not quotes, and not promises. Your project lands where it lands based on the five levers above.
| Project type | What it looks like | Typical range |
|---|---|---|
| Simple integration | Two tools connected, e.g. form submissions → your CRM | $3,500–$6,000 |
| Multi-system workflow | Booking → dispatch → invoice → follow-up, across 3–4 tools | $7,500–$15,000 |
| AI-assisted workflow | AI reads incoming requests, drafts quotes or routes work, human approves | $12,000–$25,000 |
| Custom AI build | After-hours lead response, self-writing reports, real judgment work | $20,000+ |
Notice the gap between "simple integration" and "AI-assisted workflow." That's not padding — it's the testing. A rule-based integration either works or it doesn't. An AI system works 95% of the time, and the other 5% is where the engineering goes.
And notice what's missing: hourly billing. We don't do it. You get a fixed price before we start, and if we underestimated, that's our problem, not yours. The reason is simple: hourly billing rewards slowness, and we'd rather be judged on the outcome.
If the math doesn't work, we don't build it. That's the whole pitch.
The payback question
Cost is only half the math. The other half: what does it give back?
The formula is simple enough to do on a napkin:
Hours saved per month × fully loaded hourly cost = monthly value
Then: project cost ÷ monthly value = payback period in months. Say a tech costs you $35/hr fully loaded (wages, taxes, insurance, the truck — everything). If automation saves her 10 hours a month of admin work, that's $350/month of value. A $7,000 project pays back in 20 months.
Is 20 months good? Depends. Software that saves 10 hours a month but breaks twice a month isn't worth it. Software that saves 40 hours a month pays back in 5. The math doesn't lie — but it only works if the hours are real.
That's the trap to watch for: consultants who claim your team "saves 30 hours a week" by counting time nobody actually tracked. Ask for the arithmetic. Make them show the hours, the rate, and the payback months. If they can't, the savings aren't real.
Want the full worked example? It's in Automation payback math: a worked example.
The fixed-price pitch, honestly
Here's our pitch, and the reason we can make it: fixed price, delivered in 2–4 weeks, you own everything.
Fixed price means you know the number before we start. 2–4 weeks means we scope small enough to actually ship. You own everything means the code, the accounts, the docs — if you fire us, the system keeps running and your next person can pick it up.
That's not generosity. It's how we'd want to buy software. And it keeps us honest: if a project won't pay back, saying so costs us nothing, because we only get paid for work that makes sense.
"We'll tell you not to build it"
This is the part most consultants won't say. Sometimes the answer is no.
If you're doing 12 invoices a month, automating invoices is a $6,000 answer to a $600 problem. If your process is broken — if nobody agrees on how a job gets dispatched — automating it just makes the mess run faster. If there's no owner who can say "yes, that's right" when the system does something unexpected, it will fail in ways nobody catches.
We have a whole post on this: When not to automate: we'll tell you ourselves. It's the shortest sales pitch in our business: sometimes the best advice we can give is "don't."
Where this goes from here
This guide is the hub. Four posts hang off it, each going deeper on one piece of the math:
- What business automation actually costs in 2026 — price bands, what moves a project up or down
- Build vs. buy: when off-the-shelf software wins (and when it doesn't) — the honest decision framework, including the subscription treadmill
- Automation payback math: a worked example — step-by-step arithmetic with clearly labeled illustrative numbers
- When not to automate: we'll tell you ourselves — the anti-hype closer
Read them in any order. They're all written the same way this one is: no hype, real numbers, and the answer even when the answer is no.
Get a free tool-stack audit
Before you spend a dollar on automation, find out what you're already paying for and what it's actually doing. Our free tool-stack audit maps the tools you use, where the busywork hides, and whether automation would pay back — including telling you "don't" if the math says so. Book a free tool-stack audit.
