What to Charge for AI Automation Services: Exact Pricing by Project Type

One of the most common questions AI agency owners ask is: "How much should I charge?" It's also one of the questions with the most bad advice floating around. You'll see forums recommending everything from $200 for a simple chatbot to $50,000 for an enterprise AI deployment. Neither extreme is useful for the agency owner trying to build a sustainable business serving local and mid-market businesses.

This guide gives you the exact pricing benchmarks we've seen work across hundreds of AI automation engagements in 2026, organized by project type. We'll also cover the pricing frameworks that the most successful agencies use to move clients from sticker shock to "where do I sign."

The Three Pricing Models for AI Automation Agencies

Before we get into specific numbers, you need to understand the three models and when each one is appropriate:

Model 1: Project-Based Pricing

A one-time fee for building a specific automation. The client pays once, owns the deliverable, and you move on. This model is fastest to close but creates revenue instability. Best for: simple automations, first-time clients who are hesitant to commit to ongoing, or when the automation truly requires no maintenance.

The hidden problem with project-only pricing: you're on a hamster wheel. Every month you need new clients just to maintain revenue. Agencies that rely solely on project fees typically plateau around $8,000-$12,000/month because there are only so many hours in a day. Project pricing also attracts one-and-done buyers rather than the long-term relationships where the real money is made.

Model 2: Retainer Pricing

A recurring monthly fee for ongoing management, maintenance, optimization, and support. This is the gold standard for agency revenue stability. Best for: all automations that connect to external APIs, run continuously, or need regular updates based on performance data.

What goes into a retainer? More than clients realize. You're covering: platform hosting and subscriptions, API costs, monitoring the automation for failures, updating prompts and logic as the client's business changes, handling edge cases that break the system, responding to client questions, and running monthly performance reviews. When you itemize this for clients, $997/month starts looking like a bargain.

Model 3: Hybrid (Setup + Retainer)

An upfront setup fee to build the automation plus a monthly retainer for ongoing management. This is the most common and arguably best model for most AI agency engagements. It aligns your interests with the client's long-term success and ensures you're compensated fairly for the initial build work.

The setup fee also serves as a commitment device. Clients who pay $2,500 upfront are far less likely to cancel after month one than clients who paid nothing. You're not just covering build costs — you're buying 6+ months of client retention before you even launch.

For a full breakdown of how to structure your agency's pricing tiers, see our AI agency pricing guide which covers everything from starter packages to enterprise retainers.

Exact Pricing by Automation Type

1. Missed Call Text-Back Automation

This is the entry-level automation that every local business needs and understands immediately. A missed call triggers an automated SMS within 60 seconds. Complexity: low. Build time: 2-4 hours.

• Setup fee: $500 - $1,500
• Monthly retainer: $297 - $597/month
• Typical total first-year value: $4,064 - $8,664

Many agencies use this as a "foot in the door" product — charge a modest $797 setup and $297/month, then upsell to more comprehensive packages after 60-90 days when the client is seeing results and trusts you.

Real-world example: A plumbing company averaging 25 missed calls per week closes 20% of them into jobs worth $450 average. That's 5 jobs per week or $2,250/week recovered — $9,000/month. Your $497/month retainer is a 18x ROI on paper. When you walk a plumber through that math on a discovery call, the conversation shifts from "is this worth it" to "when can we start."

2. AI Lead Follow-Up Sequence

A multi-step SMS and email follow-up sequence triggered by web form submissions, ad leads, or inbound inquiries. Includes AI-generated personalized responses, qualification questions, and appointment booking. Build time: 4-8 hours.

• Setup fee: $1,500 - $3,000
• Monthly retainer: $497 - $997/month
• Typical total first-year value: $7,464 - $14,964

What drives pricing within this range: number of lead sources integrated (1 source = lower end, 3+ sources = upper end), complexity of qualification logic, number of sequence steps, and whether you're building the CRM integration from scratch. A real estate agent running Facebook ads and a Google Local Services profile with two different lead funnels that both need 7-touch sequences is a $2,500 setup, not $1,500.

Speed-to-lead is the key ROI hook here. Studies consistently show that leads contacted within 5 minutes are 9x more likely to convert than leads contacted after 30 minutes. That research point belongs in every proposal for this type of automation.

3. AI Chatbot (Website Lead Capture)

A conversational AI chatbot embedded on the client's website that qualifies leads, answers FAQ, and books appointments. Built in Voiceflow or Botpress and connected to the client's CRM or calendar. Build time: 8-15 hours depending on complexity.

• Setup fee: $2,000 - $4,500
• Monthly retainer: $597 - $1,200/month
• Typical total first-year value: $9,164 - $18,900

What justifies the higher end of this range: a knowledge base with 50+ FAQ entries, multi-language support, integration with a proprietary CRM (not just HubSpot or GoHighLevel), e-commerce product recommendation logic, or a bot that handles complex triage (like a medical spa that needs to route different treatment inquiries to different staff members). Simpler bots that qualify and book — five intents, one calendar integration — are $2,000-$2,500 setups. Complex, deeply trained bots are $4,000-$6,000.

4. AI Appointment Booking and Reminder System

End-to-end appointment management: AI books appointments via conversation, sends confirmations, sends reminders at 48 hours, 24 hours, and 2 hours before, and triggers re-booking workflows for no-shows. Build time: 6-12 hours.

• Setup fee: $1,500 - $3,500
• Monthly retainer: $497 - $1,200/month
• Typical total first-year value: $7,464 - $17,900

The no-show reduction angle is your best ROI argument here. A dental practice with 80 appointments per month and a 15% no-show rate is losing 12 appointments. At $200 average revenue per appointment, that's $2,400/month in lost revenue. Even if your reminder system cuts no-shows from 15% to 8%, you've recovered $1,400/month. Your $600/month retainer pays for itself twice over — and that's before counting the new patients who would have gotten those slots.

5. CRM Integration and Data Enrichment Automation

Automating lead import, data enrichment (company size, LinkedIn profile, phone number validation), lead scoring, and CRM update workflows. Typically built in n8n connecting multiple APIs. Build time: 10-20 hours.

• Setup fee: $2,500 - $6,000
• Monthly retainer: $797 - $1,997/month
• Typical total first-year value: $12,064 - $29,964

This automation type is best sold to B2B companies with sales teams. The prospect needs to be spending significant time on manual CRM work — updating records, pulling contact data from LinkedIn, manually scoring leads — before this makes financial sense. A 5-person sales team each spending 2 hours per week on CRM maintenance is 10 hours/week or ~43 hours/month. At $50/hour fully loaded labor cost, that's $2,150/month in wasted time. Your $797/month automation more than covers it.

6. Cold Email Outreach Automation System

A full cold email infrastructure including domain setup, warm-up, AI-personalized email sequences, reply detection, and calendar booking. This is a higher-complexity build that many agencies specialize in exclusively. Build time: 15-30 hours.

• Setup fee: $3,000 - $8,000
• Monthly retainer: $1,500 - $4,000/month
• Typical total first-year value: $21,000 - $56,000

Where the range moves: lower end is a single sender domain, one campaign, basic personalization (first name + company name). Upper end is 5+ sending domains, AI-personalized first lines using scraped prospect data (recent funding, new hire, LinkedIn posts), multi-step sequences with conditional logic, and integration with a full sales stack. Agencies specializing in cold email often move to performance-based components — a $2,000/month base plus $200-$500 per booked meeting — which aligns incentives and can push effective revenue well above these ranges.

For clients who need cold outreach, be sure to understand the technical requirements. See our cold email deliverability checklist before pricing these projects.

7. Full AI Employee (Voice + SMS + Email Agent)

A comprehensive AI agent that handles inbound calls, texts, and emails simultaneously — qualifying leads, booking appointments, answering questions, and escalating to a human only when necessary. This is the premium product that the most sophisticated agencies are selling. Build time: 20-50 hours.

• Setup fee: $5,000 - $15,000
• Monthly retainer: $2,500 - $6,000/month
• Typical total first-year value: $35,000 - $87,000

The comparison that makes this an easy sell: a full-time receptionist or inside sales rep costs $45,000-$60,000/year in salary alone. Add benefits, payroll tax, training, and management overhead and you're at $65,000-$80,000/year. Your AI employee at $3,000/month ($36,000/year) is cheaper, available 24/7, never calls in sick, handles unlimited simultaneous conversations, and can be updated instantly. Frame it this way and your pricing becomes a staffing cost reduction story, not a technology cost story.

How to Price Your Services: The 4 Frameworks

Framework 1: Value-Based Pricing

This is the most powerful pricing approach and the one used by the highest-earning agencies. Instead of pricing based on your time or cost, you price based on the value you create for the client.

Formula: Your price = 10-20% of the annual value you create.

Example: An HVAC company gets 40 missed calls per month. Each missed call represents a potential $2,500 job. If your automation recovers 15% of those (6 calls), that's $15,000/month or $180,000/year in recovered revenue. 10% of $180,000 = $18,000/year or $1,500/month. This is your justified retainer price, and the client is still getting a 9x return.

To use this framework effectively, you need to ask good discovery questions that surface the actual numbers: average job or transaction value, volume of leads, current close rate, number of staff hours spent on manual tasks, and the cost of their current solutions. The more specific you can get, the more compelling your ROI calculation becomes. A prospect who tells you their average job is "around $800-$1,200" gives you less leverage than one who says "our average service call runs $1,100 and we do about 60 a month when we're fully booked."

Framework 2: Tiered Package Pricing

Offer three packages — Starter, Growth, and Scale — with clear feature differentiation. This anchoring technique makes the middle tier look like the obvious choice, and the top tier makes the middle seem reasonable.

• Starter ($997/month): Missed call text-back + basic follow-up sequence
• Growth ($1,997/month): Full AI follow-up + chatbot + appointment booking
• Scale ($3,497/month): Everything in Growth + voice AI + custom integrations + monthly strategy call

Packaging psychology matters here. The Starter option should feel slightly incomplete — good enough to work, but missing 1-2 features the client will clearly wish they had. The Scale option should feel like it's designed for businesses larger than theirs. Growth should feel like it was built exactly for them. Most of your closes will land in Growth. The Starter gives hesitant prospects a lower entry point (and a natural upsell path). The Scale gives confident prospects permission to invest more.

Add-ons amplify this model. After presenting your packages, mention optional add-ons: custom reporting dashboard (+$200/month), additional location integration (+$300/month per location), priority support SLA (+$150/month). These feel small against the base package price and convert well.

Framework 3: Cost-Plus Pricing

Calculate your actual costs (tool subscriptions, API costs, your time at a target hourly rate) and add a margin. This is the floor — never price below this. For most AI automations, your actual cost is $50-$300/month per client including tool costs and time. A 5-10x markup is standard.

Break down a typical small business automation client's cost structure: n8n or Make.com hosting ($15-$30/month), Twilio or similar SMS costs ($20-$60/month depending on volume), OpenAI API costs ($10-$40/month), your time for maintenance and support (1-2 hours/month at your target rate). Total hard costs: $45-$130/month. At a $597/month retainer, your gross margin is roughly 78-92%. At $297/month (the minimum you should ever charge), margin is still 56-84% — acceptable, but leaves little room for surprises. This is why $297/month is a hard floor, not a starting point.

Framework 4: Competitive Anchoring

Research what full-time employees doing similar work would cost. A human receptionist in the US costs $35,000-$50,000/year in salary plus benefits — roughly $3,300-$4,500/month all-in. Your AI automation at $1,500-$2,500/month is a bargain, and framing it this way makes the price feel like a steal.

You can also anchor against what clients are currently spending on inferior solutions. A local business paying $300/month for a basic web chat widget that doesn't qualify or book is already in the habit of spending money on this problem. Your $797/month fully automated system is a 2.5x price increase for something 10x more capable. That's not an objection — that's an easy trade-up conversation.

Industry-Specific Pricing Adjustments

Not all industries pay the same for the same automation. Your pricing should reflect the economics of the client's business. Here are the key variables:

• High average transaction value + low volume (law firms, med spas, financial advisors): Price at the upper end of every range. A law firm where each new client is worth $5,000-$15,000 in fees can justify $2,500-$4,000/month for automations that improve lead conversion by 15%. The math is easy.
• High volume + moderate transaction value (restaurants, gyms, barbershops): Price at the lower to mid range but stack multiple automations. A gym might pay $997/month for missed call text-back + lead follow-up + membership trial automation. The per-automation price is low, but the package is compelling.
• Multi-location businesses: Add 30-50% per location. A dental group with 3 locations should pay $1,500/month where a single-location practice pays $997/month. You're managing more volume, more integrations, and more edge cases.
• Businesses running paid ads: Premium is justified. A roofing company spending $5,000/month on Google ads needs their leads handled perfectly — slow follow-up means that $5,000 is partially wasted. They're already in the mindset of spending money to grow, and protecting ad spend is an ROI argument they understand immediately.

How to Handle the "That's Too Expensive" Objection

This objection almost always means one of three things: they don't understand the value, they can't afford it, or they don't trust that it will work. Each requires a different response.

If they don't understand the value: Return to the ROI calculation. "I want to make sure I explained this clearly — you mentioned you lose about 20 leads per month to slow response time. At $1,800 average job, that's $36,000 in potential revenue. We're asking for $1,500/month to recover a meaningful portion of that. Does the math make sense to you?"

If they genuinely can't afford it: Offer the Starter package or a single automation to start. "If the full Growth package feels like too much right now, we can start with just the missed call text-back and follow-up sequence at $797/month. Once you see results, we can talk about expanding. Most clients who start here move to the full package within 90 days."

If they don't trust it will work: Offer a performance guarantee or a 90-day trial with clear success metrics. Alternatively, offer to show them a live demo using their own business data. Nothing builds trust faster than showing a prospect their own missed calls being automatically followed up in real time.

What you should never do: drop your price without getting something in return. If a client asks for a discount, tie it to something — a longer commitment ("I can do $1,797/month if you commit to 6 months instead of month-to-month"), a referral agreement ("I can offer 15% off if you agree to make an introduction to two other business owners you know"), or a reduced scope ("I can reduce the setup fee by $500 if we skip the CRM integration for now and add it later").

Common Pricing Mistakes to Avoid

• Pricing per hour: Never quote hourly rates. This puts the focus on your time rather than results, creates endless scope creep discussions, and limits your income to your hours worked.
• Underpricing to get clients: A $300/month retainer attracts clients who are the most demanding and hardest to retain. Higher-priced clients respect your work more and churn less. Never go below $497/month for any ongoing engagement.
• No setup fee: The setup fee serves two purposes: it compensates you for the build work, and it creates skin in the game for the client. Clients who pay nothing upfront cancel more frequently.
• Same price for every client: Your pricing should reflect the client's industry, business size, and the complexity of their situation. A 3-location med spa should pay more than a solo HVAC company.
• Forgetting API cost escalation: If you include API costs in your retainer (rather than passing them through), build in a buffer. A client who suddenly starts running 10,000 leads through your system will eat into your margins fast. Either include a fair-use cap in your contract ("up to 500 contacts/month, additional volume billed at $X/1,000") or pass through API costs at cost with a clear itemized breakdown.
• Discounting without a reason: If you discount without conditions, you train the client to expect discounts, signal that your original price wasn't real, and devalue your work. Every discount should come with a tradeoff.
• Month-to-month for complex automations: Any automation that took more than 10 hours to build should have a minimum 3-month commitment. You need time to prove ROI, and the client needs time to see results. Month-to-month contracts for complex builds mean you're one bad month away from losing clients you spent significant time onboarding.

When to Raise Your Prices

You should raise prices when:

• You have 3+ clients with strong results and testimonials
• Prospects are saying yes too easily (a sign you're underpriced)
• You're turning down work due to capacity
• You've added significant new capabilities to your stack
• Your close rate exceeds 40% on discovery calls

Most successful agencies raise prices 20-30% every 6 months in their first two years. Grandfathering existing clients at their original rate while charging new clients more is standard practice and usually appreciated.

A practical trigger: when you book your 5th client at a given price point, raise prices for the next prospect. You now have enough volume to test the new rate without risking your whole pipeline. If you close 2 of the next 5 at the higher price, the increase is validated. If you close zero, re-examine your pitch — the price may not be the issue.

For the full picture on building a scalable agency, read our guide on how to start an AI automation agency and our post on how to productize your services into packages.

What to Include in Your Proposal

A proposal that closes is not a list of deliverables. It's a document that mirrors the client's problem back to them, quantifies it, and presents your solution as the obvious fix. Structure every proposal like this:

• The Problem (their words, not yours): Use exact language from the discovery call. "You mentioned you're losing an estimated 15-20 leads per month because your team can't respond fast enough..."
• The Cost of Inaction: Quantify the current pain. Missed leads × average job value × 12 months. Make the number real.
• Your Solution: Describe what you're building and why each component exists. Not features — functions. Not "n8n workflow with Twilio integration" — "automated text response within 60 seconds of any missed call, 24/7."
• Expected Outcomes: Be specific. Based on similar clients, here is what typically happens in 30/60/90 days. If you have a case study, include it. If you don't, use conservative industry benchmarks and say so.
• Investment: Setup fee + monthly retainer + total first-year investment. Then show the ROI calculation: "If this system recovers just 4 additional jobs per month at $1,800 average, that's $86,400 in recovered revenue over the year against a $26,964 investment — a 3.2x return."
• Next Steps: A clear call to action with a deadline. "To get started this month, we'd need your signed agreement and setup fee by [date]. That puts your system live by [date]."

Sample Pricing Conversation Script

After you've demonstrated value on a discovery call, here's how to present pricing without hesitation:

"Based on everything you've shared — you're losing roughly 15-20 leads per month to slow response time, and each of those leads is worth $2,500 to $3,500 — the ROI here is significant. Our Growth Package at $1,997/month includes the full AI follow-up system, the chatbot on your website, and automated appointment booking. There's a one-time setup investment of $2,500. At minimum, this system needs to recover 2 additional jobs per month to pay for itself — and in our experience with similar businesses, we typically see 6-10 additional conversions in the first 90 days. Does the timing work to get started this month?"

Notice: you presented the ROI first, the price second. You gave a specific number of expected conversions based on experience. And you closed with a forward-moving question, not a "what do you think?"

One more thing: deliver the price with confidence and then stop talking. The silence after stating a price is uncomfortable. Most people fill it by backpedaling, offering discounts before the client has objected, or over-explaining. State the number. Stop. Let them respond. Their response tells you exactly what objection you're handling, which is far more useful than preemptively defending a price they might have accepted without hesitation.

Ready to build your AI agency? Join 215+ AI agency owners in our free community. Join the free AI Automation Sprint on Skool.