Skip to main content

Pricing Your AI Engagements

If you price by the hour, you are training clients to see you as labor. You are inviting them to compare you to freelancers, micromanage your time, and focus on effort instead of outcomes. Nobody wins.

This guide is a framework for thinking through pricing on any AI consulting project. It is a living document, updated as we learn from real engagements.

The Core Principle​

Price by the transformation, not by the hour.

When you say "20 hours at $100/hour," the client focuses on the 20 hours. When you say "this system will save your team $240,000/year and the investment is $8,000," the client focuses on the ROI. The second conversation closes. The first one invites haggling.

Growth Center, Not Cost Center​

Position your work as a growth engine, not an expense.

Trying to sell AI as a way to cut costs puts you in a losing position. You are a new line item competing against existing ones. It is a losing battle to try to be paid for cutting costs, because you are a new cost.

But if you can grow someone's company, that is an easy sell. Frame your work around revenue unlocked, customers gained, capabilities built, and markets entered. Organizations that see applied AI engineering as a growth function scope bigger engagements, invest with conviction, and build lasting partnerships with the practitioners who deliver results.

This reframe changes everything: your pricing conversations, your proposals, your case studies, and the clients you attract.

Credit to Reid McCrabb and Jack Moffatt of Linkt for crystallizing this insight at Applied AI Live #2.

The Pricing Calculator​

For any new engagement, work through these factors in order.

Step 1: Quantify the Business Impact​

Every AI engagement maps to at least one of these:

Impact TypeHow to Quantify
Revenue unlockedNew revenue streams, increased conversion, faster sales cycles
Cost reducedHeadcount avoided, tool consolidation, error reduction
Time recoveredHours/week saved x loaded hourly cost of the people freed up
Risk mitigatedCost of compliance failures, data loss, competitive displacement

The formula: Estimate the annual impact in dollars. Your price should be 10-20% of Year 1 value. Position this as: "You keep 80-90% of the value. I take 10-20% for making it happen."

Step 2: Assess the Client Variables​

FactorLower Price SignalHigher Price Signal
Client revenuePre-revenue or small$10M+ established
Decision-maker accessTalking to a managerTalking to the CEO/CTO
Urgency"Exploring AI someday""We are losing to competitors NOW"
ComplexitySingle workflowMulti-system transformation
DurationOne-off build4-6+ month engagement
Strategic valueInternal convenienceCore business IP or competitive moat

Step 3: Check Your Own Variables​

FactorQuestion to Ask Yourself
Opportunity costWhat am I NOT doing while I work on this?
ExcitementDoes this energize me or drain me?
Case study valueWill this become a story I tell for years?
Relationship depthIs this someone I want in my orbit long-term?
Equity beliefDo I believe in this company's upside enough to take equity?
Cash needDo I need immediate cash, or can I weight toward equity/rev share?

Four Pricing Models​

Model A: Value-Based Project Fee​

Best for defined-scope builds with measurable ROI.

How to calculate:

  1. Quantify the annual business impact
  2. Charge 10-20% of Year 1 value
  3. Minimum floor: $5,000 for any meaningful project

Example: You build an AI voice agent for a home service company. They currently close 2 of 10 leads because the other 8 are not responded to fast enough. With the agent, they close 8 of 10. At $4,000 per deal, that is $24,000/month in recovered revenue, or $288,000/year. Your project fee: $8,000. That is less than a single month's value. The client does not blink.

Model B: Monthly Retainer​

Best for ongoing relationships where the system needs tuning and the client needs coaching.

How to calculate:

  • For revenue-generating systems: $1,000-3,000/month (fraction of monthly value created)
  • For time-saving systems: 15-20% of monthly time value saved
  • Always itemize what the retainer includes: monitoring, optimization, coaching, priority support, monthly impact reports

Why retainers work: A monthly impact report showing "47 leads contacted, 13 deals closed, $52,000 estimated revenue impact" turns a $1,000/month retainer into the most obvious investment the client makes.

Model C: Hybrid (Setup Fee + Retainer)​

Best for most engagements. Gives you upfront cash flow and recurring revenue.

Structure: Project fee for the build + monthly retainer for ongoing value.

Example:

  • $15,000-25,000 upfront for discovery + build (3-4 months)
  • $2,000-5,000/month retainer for ongoing coaching, optimization, evolution
  • Total Year 1 value to you: $39,000-85,000

The hybrid model works psychologically too. The upfront fee makes the client take the engagement seriously. The monthly retainer keeps you engaged and prevents the "build it and ghost it" dynamic that kills one-time projects.

Model D: Equity or Revenue Share​

Best for strategic partnerships where you deeply believe in the company's trajectory.

When to use this:

  • You genuinely believe the equity will be worth something
  • You like the people enough to be tied to them for years
  • The cash component still covers your costs
  • The work will produce reusable case studies or systems

Structure: Reduced cash fee + equity stake or percentage of net profit. Common ranges: 5-20% depending on depth of involvement.

How This Connects to the Client Journey​

The Business Owner playbooks walk your prospective clients through a three-stage readiness process before they ever talk pricing with you. Understanding where a client is in that journey changes how you price.

Client StageWhat They've DonePricing Implication
AI Quick CheckAssessed their own readiness with six yes/no questionsIf they scored "early," they need coaching before building. Price for discovery and education, not implementation.
Situation MapMapped their workflows, data, team, and gapsYou can skip weeks of discovery. Their gaps are documented. Price the build, not the diagnosis.
Pilot ScopeScoped a concrete pilot with problem, success metric, constraints, and budgetThis is a client who has done the work. They know what they want, what it costs them, and what they can spend. You can price with confidence because the scope is clear.

The ideal scenario: A client comes to you having completed all three stages. They hand you a Pilot Scope that includes the problem, the cost of the problem, their budget range, and their success metrics. Your pricing conversation is now about aligning your model to their documented needs, not convincing them the problem exists.

The common scenario: A client comes to you at the Quick Check stage or earlier. They know something is wrong but haven't mapped it. In this case, your first engagement should be a paid discovery (the Situation Map process), not a build. Price the discovery at $2,000-5,000 depending on complexity. The discovery itself becomes the foundation for the larger engagement.

Why retainers make sense: The Don't Accept Automation as the Goal playbook makes the case that one-time builds are the beginning, not the end. Real value comes from continuous improvement: measuring outcomes, optimizing the system, and iterating based on data. This is the philosophical foundation for why retainers are the right model. Automation is a process change. Improvement is a compounding system. Your retainer is what makes the compounding happen.

Matching engagement types to pricing: The Hiring Practitioners guide defines five engagement types (workflow automation, executive coaching, culture transformation, custom tool building, internal champion development). Each maps to a different pricing model:

Engagement TypeRecommended Pricing Model
Workflow automationValue-based project fee (Model A) or hybrid (Model C)
Executive coachingMonthly retainer (Model B)
Culture transformationHybrid (Model C) at enterprise price floors
Custom tool buildingHybrid (Model C) with heavier upfront fee
Internal champion developmentMonthly retainer (Model B) with defined graduation criteria

The Discovery Conversation​

You do not lead with the price. You lead with the diagnosis. If the client has already completed a Situation Map, you can move faster. If they haven't, the discovery conversation IS the situation map, and it should be a paid engagement.

Step 1: Diagnose the Pain​

Ask questions that reveal the cost of the problem:

  • "Walk me through what happens when [problem] occurs today."
  • "How many hours per week does your team spend on [manual process]?"
  • "What does that cost you in loaded salary?"
  • "What happens to deals when [bottleneck] slows things down?"
  • "How many opportunities have you lost because of this?"

Step 2: Quantify the Gap​

Reflect the cost back:

  • "So you're spending roughly $X/year on this problem today."
  • "And the opportunity cost of not solving it is another $Y."
  • "That means this problem is costing you about $Z per year."

At this point, the client is not thinking about your hourly rate. They are thinking about the $Z they are losing.

Step 3: Position the Investment​

Frame your price as a fraction of the value:

  • "I can solve this for [price], which is roughly [X]% of what it's costing you annually."
  • "You keep 80-90% of the value. I take 10-20% for making it happen."
  • "The system pays for itself within [timeframe]."

Step 4: Justify with Proof​

If you have them, reference comparable results:

  • "I built a similar system for [anonymized client] that saved them [X hours/week]."
  • "Organizations I've worked with typically see [X] ROI in the first quarter."

If you don't have case studies yet, start with a lower price to get one. Then double it on the next engagement. Keep doubling until you start hearing "no." If your close rate is 100%, your prices are too low.

Price Floors​

These exist because anything below them signals "cheap labor" and attracts the wrong clients.

Engagement TypeMinimum
Discovery/audit$2,000
1-month sprint project$5,000
3-6 month transformation$15,000
Monthly retainer (coaching + optimization)$2,000/month
Enterprise culture transformation$50,000+

Red Flags (Walk Away)​

  • They ask for your hourly rate before understanding the scope
  • They compare you to freelancers or Upwork contractors
  • "Can you just build me a chatbot?"
  • They want to pay per feature, not per transformation
  • No executive sponsor (a middle manager cannot authorize real spend)
  • They want to own everything with no ongoing relationship

Worked Examples​

Example 1: Lead Response System for a Services Company​

The scenario: Marcus is an applied AI practitioner. A regional plumbing company ($3M/year revenue) asks him to build an AI system that responds to inbound leads within 60 seconds, qualifies them, and routes hot ones to the dispatch team. Currently, 70% of leads go unanswered because the team is on job sites all day.

The math:

  • 40 leads/month, currently closing 12 (30% response rate)
  • With the AI system: 85% response rate, estimated 34 closeable leads
  • Average job value: $1,200
  • Delta: 22 additional closes/month = $26,400/month = $316,800/year

Marcus's price: $6,000 project fee + $800/month retainer. The project fee is under 2% of Year 1 value. The retainer is 3% of the monthly delta. The owner sees this as a no-brainer.

Example 2: Knowledge System for a Manufacturing Firm​

The scenario: Priya is building an internal knowledge base with three AI agents for a mid-size electronics manufacturer ($85M/year revenue). The agents handle supplier lookup, compliance document retrieval, and engineering spec search. Currently, engineers spend 8-12 hours/week hunting for information across shared drives, email, and tribal knowledge.

Discovery questions she asks:

  1. What decisions does this knowledge system need to support?
  2. Who are the users of the three agents? What do they do today without them?
  3. How many hours/week do people currently spend on what these agents would automate?
  4. What is the loaded cost of those people's time?
  5. Are there revenue implications? (Faster decisions, fewer errors, competitive advantage.)
  6. What is the cost of a bad decision that better knowledge access would prevent?

The math:

  • 15 engineers x 10 hours/week saved x $95/hr loaded cost = $741,000/year
  • Plus: faster quoting cycles, fewer compliance errors, reduced rework
  • Conservative Year 1 value: $500,000+

Priya's price: $50,000 project fee (4-month build) + $5,000/month ongoing retainer. That is 10% of conservative Year 1 value for the project fee. The client's VP of Engineering approves it in one meeting.

How she frames it: "This isn't a software build. It's a strategic intelligence layer for your engineering team. The knowledge system becomes institutional memory that compounds over time. The agents turn that memory into action. The question isn't what this costs. It's what it costs you to NOT have it for another year."

Example 3: Equity Play for an Early-Stage Company​

The scenario: Derek is approached by a friend's Series A startup ($2M ARR, growing fast) to build their entire AI operations layer. He loves the team, believes in the product, and wants long-term upside.

Derek's structure: $3,000/month retainer (covers his time costs) + 3% equity vesting over 2 years. If the company hits $20M ARR, that equity is worth real money. If it doesn't, he still got paid for his time. He only takes this deal because he genuinely believes in the company and wants to be tied to these people for years.

Let AI Interview You on Pricing​

Most practitioners undercharge because they never think through their pricing rigorously. They pick a number that feels safe, quote it on a call, and hope the client says yes.

The prompt below does something different. It turns your AI coding tool into a pricing strategist that interviews you about the engagement, pressure-tests your assumptions, and helps you arrive at a number you can defend with confidence. The interview format matters because it forces you to articulate things you would otherwise leave vague: the real cost of the client's problem, your own opportunity cost, what you actually need from this deal.

You will often discover that the engagement is worth significantly more (or less) than your gut said. You might realize the client needs a paid discovery phase before a build. You might catch a red flag you would have missed. The AI has no ego in the conversation, so it will ask the uncomfortable questions a friend might skip.

How to use it: Copy the block below. Paste it into Claude Code, Cursor, ChatGPT, or any AI tool that can fetch URLs. Don't fill in the bracketed sections yourself. Just paste it and let the AI interview you. The conversation will surface the details naturally.

# AI Consulting Pricing Advisor

You are going to help me think through pricing for an AI consulting
engagement. Before we start, read the full pricing framework and the
client readiness playbooks at these URLs:

- Pricing framework: https://docs.appliedaisociety.org/docs/playbooks/practitioner/pricing
- Quick Check: https://docs.appliedaisociety.org/docs/playbooks/business-owner/quick-check
- Situation Map: https://docs.appliedaisociety.org/docs/playbooks/business-owner/situation-map
- Pilot Scope: https://docs.appliedaisociety.org/docs/playbooks/business-owner/pilot-scope
- Beyond Automation: https://docs.appliedaisociety.org/docs/playbooks/business-owner/beyond-automation
- Hiring Practitioners: https://docs.appliedaisociety.org/docs/playbooks/business-owner/hiring-practitioners

After reading those, interview me. Do not ask me to fill out a template.
Ask me questions one at a time, in a natural conversation. Start by
asking about the engagement, then dig into the details you need.

Your goal is to understand:

**About me:**
- My background, expertise, and what I've charged before
- My constraints (time, other commitments, cash needs)
- What I value beyond money (case study potential, relationship,
  learning, equity upside)

**About the client:**
- Their business, industry, and approximate scale
- Who I'm talking to and whether they can authorize real spend
- Where they are in the readiness journey (Quick Check, Situation Map,
  or Pilot Scope stage)
- Their urgency and AI maturity

**About the engagement:**
- What I'm building or delivering
- Who the end users are and what they do today without this
- Timeline and complexity
- The cost of the problem this solves
- Whether there's an equity, rev share, or partnership angle
- Which engagement type this matches (workflow automation, executive
  coaching, culture transformation, custom tool building, or internal
  champion development)

Once you have enough context, walk me through:

1. The Pricing Calculator from the framework (business impact, client
   variables, my variables)
2. Your recommended pricing model with justification
3. A specific price range tied to ROI math
4. Where the client is in the readiness journey and what that means
   for scoping (should the first engagement be a paid discovery or a
   full build?)
5. Talking points for the pricing conversation using the Discovery
   Conversation framework
6. Any red flags from the Red Flags checklist
7. A draft proposal structure I can adapt

Push back if my instinct is to undercharge. Ask hard questions. Do not
be polite about it.

Building Your Proof​

You don't need 50 case studies to start pricing based on value. You need one solid proof point.

Structure your first case study like this:

  • Before: Client was spending [X hours/week] on [manual process], costing them [$Y/month]
  • After: Automated system reduced this to [Z hours/week], saving [$W/month]
  • What you did: Built [specific system] connecting [platforms/tools]. Implemented [key features]. Trained team on new process.
  • Result: Client recovered [X hours/week] and [$Y annually]. Paid [$Z] for implementation.

One documented before/after with real numbers is more powerful than any pitch deck.

Once you start retainers, send monthly impact reports. Show the numbers. You are constantly reinforcing your value and justifying the investment.

Further Reading​

For practitioners:

Understand what your clients see: