How to Price AI Automation Projects (Without Undercharging)

A step-by-step pricing model for AI freelancers and agency owners — with market benchmarks, contingency math, and how to hold your rate when a client pushes back.

The most common mistake AI automation freelancers make isn't technical — it's pricing. They quote based on what they think the client will accept, take the project, and end up working at $18/hour once scope creep and revisions are accounted for.

This post gives you a repeatable pricing model. It works for Make.com and n8n projects, for chatbot builds, for full intake automation systems — anything where you're delivering an AI-powered workflow to a client. It's the same model included in our AI Client Delivery OS.

Why AI Freelancers Underprice

There are three reasons this keeps happening:

  1. They price based on effort, not value. A 4-hour Make.com workflow that saves a client 15 hours/week is worth far more than $400. Effort-based pricing is a ceiling; value-based pricing is a floor.
  2. They forget what they can't see. The technical build is only part of the project. Discovery, documentation, client communication, revisions, and post-delivery support are real time — often 40–60% of total project hours.
  3. They're afraid to lose the deal. New freelancers especially price low to "get experience." The problem: clients who buy on price are the hardest clients to work with and the least likely to refer you to anyone.

The reframe: You're not selling hours. You're selling a working system with a measurable outcome. Price the outcome, not the build time.

The 5-Step Pricing Model

Step 1

Estimate Hours Honestly

List every task in the project and estimate hours for each. Do not estimate for the ideal case — estimate for the realistic case.

Most AI automation projects break down like this:

TaskTypical Hours
Discovery, scoping, proposal2–5 hrs
Platform setup / account configuration1–3 hrs
Workflow / automation build4–20 hrs (varies widely)
Integrations (CRM, API, webhook)2–8 hrs
Testing and QA2–6 hrs
Documentation1–3 hrs
Client communication throughout2–5 hrs
Revisions (budget 20–30% of build time)1–5 hrs
Delivery walkthrough call1–2 hrs

A "simple" automation project often runs 15–25 total hours once everything is accounted for. A complex one with multiple integrations can easily hit 40–60 hours.

Step 2

Calculate Your Minimum Hourly Rate

Your rate should be based on your target annual income — not what you think the market will bear.

Formula Target annual income ÷ billable hours per year = minimum hourly rate

Key assumption: you are not billing 40 hours/week. Between selling, admin, learning, and dead time between projects, most freelancers bill 15–25 hours/week, or roughly 800–1,100 hours/year.

Target IncomeBillable Hrs/YearMinimum Rate
$60,0001,000$60/hr
$80,0001,000$80/hr
$100,0001,000$100/hr
$120,000900$133/hr
$150,000800$188/hr

This is your floor. Going below this rate means you're working toward a number that won't support your business or life.

Step 3

Calculate the Project Floor

Formula Total estimated hours × hourly rate = project floor price

Example: 20 estimated hours × $100/hr = $2,000 floor. This is the minimum you can charge without losing money. Never go below this.

Step 4

Apply Complexity Adjustments

Certain factors justify pricing above your floor. Apply these as percentage increases to your base number:

FactorAdjustment
First time building this type of project+20–30%
Regulated industry (healthcare, legal, finance)+15–25%
Fast turnaround requested (<2 weeks)+20–40%
Complex integrations (multiple APIs, custom webhooks)+15–30%
Revenue-critical or customer-facing system+10–20%
You've built this exact project before (efficiency)-10% max
Referral or repeat client-10% max

Note: discounts are capped at 10% for a reason. Discounting more than that devalues the work and sets a precedent the client will expect to repeat.

Step 5

Add a Contingency Buffer

Add 15–20% for scope creep, unexpected complexity, and things you didn't know you didn't know. This is not optional. Scope always expands.

Formula Adjusted price × 1.15 = final project price (conservative) Adjusted price × 1.20 = final project price (recommended)

Market Benchmarks: 7 Common AI Automation Projects

These ranges reflect what experienced freelancers charge in 2025–2026. New freelancers often charge 30–40% less; that's a temporary positioning decision, not a permanent ceiling.

Project TypeTypical RangeNotes
Simple lead follow-up automation$800–$2,000Single workflow, 1–2 integrations
Appointment reminder + confirmation system$1,500–$4,000Multi-step, scheduling platform integration
CRM + email sequence setup$1,500–$3,500Varies by platform complexity
AI chatbot (single channel)$2,000–$5,000Trained on client content; custom flows
Full intake automation (dental/legal)$3,000–$8,000Multi-step, compliance considerations
E-commerce retention stack (cart + win-back)$2,500–$6,000Multiple flows, Klaviyo/Gorgias integration
Multi-channel voice + chat + SMS system$5,000–$15,000Complex builds; healthcare/legal command premium

Monthly retainer ranges: $500–$2,500/month depending on scope and ongoing build requirements.

On hourly vs. project pricing: Quoting hourly for a defined project scope is almost always a mistake. It caps your upside (if you work faster, you earn less), creates client anxiety (they watch the clock), and encourages scope negotiation. Quote project prices. If the scope is genuinely undefined, charge a discovery fee to define it first.

Holding Your Rate When Clients Push Back

Getting a proposal out is the easy part. Here's what happens when they push back.

"That seems like a lot"

The right question is: a lot compared to what? If you're building a system that saves 10 staff hours/week at $25/hour, that's $13,000/year in recovered labor. A $3,500 project pays back in 10 weeks. Say that out loud — most clients haven't done that math.

"Can you do it for less?"

You can offer two responses: (1) "I can scope it down — here's a Phase 1 that gets you [specific outcome] for $[lower number], and we build from there." Or (2) "The scope is what it is to deliver what you're asking for. I'm confident in the value, and the price reflects that." Never just discount without changing scope.

"We have someone who can do it cheaper"

"That's absolutely possible, and you should explore it. The questions I'd ask them: have they built this specific type of system before? Can they show you a working example? And what's included after delivery if something breaks? I'm happy to explain how my approach works — and if they're a better fit, no hard feelings."

The single rule: Never apologize for your price. Saying "I know it seems like a lot" signals ambivalence and invites negotiation. State the price. Then be quiet. The client fills the silence — and it's usually more productive than you preemptively defending yourself.

Payment Structure

For projects under $5,000: 50% at contract signing, 50% at delivery.

For projects over $5,000: consider milestone-based payment — 30% at signing, 30% at Phase 1 completion, 40% at final delivery.

Never start work without a deposit. It's not about trust — it's about alignment. A client who won't pay a deposit is a client who won't respect the scope.

The Retainer Opportunity

One-time projects are how you acquire clients. Monthly retainers are how you build a business.

Every project client is a retainer opportunity — but most freelancers never ask. The right time to raise it is at peak satisfaction: right after they've seen the system working, before the post-project glow fades.

The conversation isn't a pitch — it's an observation: "The clients who get the most out of this are the ones who stay in an ongoing relationship, where we keep improving the system rather than just hoping it runs forever on its own."

The Full System: Proposal → SOW → Delivery → Retainer

The pricing model above is one piece. The AI Client Delivery OS is the full infrastructure — proposal template, SOW, onboarding emails, kickoff agenda, handoff doc, pricing calculator, and a retainer upsell script. 13 documents, ready to use on your next engagement.

$97 — one-time purchase
Get the AI Client Delivery OS →