Build-Along: start an AI consulting business

Time: about 40 min reading + 30 days of practice

The gist

A product (SaaS, software people pay for by subscription) is a box you package once and sell a thousand times. It usually takes a long time to build and even longer to pay for itself.

Consulting is you as the product. The client isn't buying a box. They're buying your expertise, your time and your attention. It doesn't scale the way SaaS does, but:

• You can start faster than with a product: there's no software to build and no infrastructure to run
• Your costs stay small (no servers, no thousands of users to support)
• One client pays far more than one SaaS subscriber, so you need many fewer customers to get going (but finding that client and winning them over is your job)
• Every client gives you the kind of feedback a SaaS founder spends months chasing in customer interviews

🎨 Picture this: a consultant is a doctor in private practice. Not a pharmacy (a product), not a hospital (an agency), not an insurance company (SaaS). One doctor, one patient, one specific problem, paid directly for the visit. Expensive by the hour, but the patient pays because they trust the doctor, not because the price is low.

This lesson is a 30-day plan for starting an AI consulting practice from scratch and getting to the point where you offer your first client a retainer (a monthly fee for ongoing help). The amounts in this lesson are examples for doing the math, not an income forecast and not a price recommendation: your results depend on your niche, your market and your work.

This is a compressed scenario. It assumes you already have a niche, service packages and a first case study or two from the earlier lessons. If you're starting from zero, take the same steps at a calmer pace: the general 90-day plan comes two lessons from now, in Graduation: your 30-60-90 day AI business plan.

🎯 What the plan aims for by Day 30

If you work the plan, by the end of 30 days you should have:

• ✅ A first retainer proposal: sent to a client or already signed (you and the client agree on the terms; whether and when they sign is up to them)
• ✅ A reusable engagement playbook: a step-by-step process you can repeat with clients #2, #3 and #4
• ✅ Personal brand assets: a LinkedIn profile, a landing page and one case study
• ✅ A pipeline of 3-5 prospects for month 2: people who might become clients and already know you (warm, not cold)
• ✅ Your first case study: your first client as social proof for the next ones
• ✅ Experience: you've been through the whole process once and know which steps work for you and which to change

This is a target for 30 days, not a guarantee: the timing and the outcome depend on your niche, your market and your work. Getting to a first client can take longer than a month. If it does, the order of the steps stays the same and the weeks stretch.

💰 Pricing structure: four levels

Consulting isn't sold "by the hour." You sell a result plus access to an expert. Here's how the pricing is structured:

You set the actual amounts yourself, based on your costs (time, tools), the value of the result to the client, and the going rates in your niche.

How the funnel works:

Audit  →  Implementation  →  Retainer
(entry)   (proof of value)   (recurring)

🎨 Picture this: a car dealership. The test drive (the audit) → selling the car (implementation) → regular service visits (the retainer). Nobody buys a car without a test drive, and nobody books a service visit for a car they don't own.

Never open with a big retainer. The client doesn't know you yet, so that's too big a step. A small audit is the "first date," where both sides find out whether there's chemistry.

WEEK 1: Foundation and pipeline

Goal for the week: 5-10 discovery calls with potential clients + 2 proposals sent.

Days 1-2: Niche and brand assets

Step 1: Pick your niche (industry × function)

The most common beginner mistake is "an AI consultant for everyone," which really means "an AI consultant for no one." Without a niche:

• You don't know what to say to people who've never heard of you
• You can't reuse your work from one client to the next
• Every client is a new field you have to learn from scratch
• You're competing with the entire market

The formula: Industry × Function.

Examples of niches that work:

• "AI for law firms with 5-20 people: automating document review"
• "AI for real estate agents in Miami: listing descriptions generated in English, Spanish and Portuguese"
• "AI for direct-to-consumer (DTC) brands with $1M-$10M in revenue: automating customer support"
• "AI for financial advisors: automating research and reporting"
• "AI for online course creators: content production and student support"

The narrowing rule: if your niche takes more than 15 words to describe, you haven't narrowed it enough, or you've narrowed it the wrong way. "AI for legal" is too broad. "AI for immigration lawyers in the US who work with clients from Latin America" is precise.

Step 2: Personal brand assets (1 day)

The minimum you need before your first call:

1. An updated LinkedIn profile

   • Headline: "I help [niche] put AI to work so they can [specific result]"
   • About: 3-4 paragraphs: who you are, what you do, who you do it for, a case study, how to reach you
   • Featured: a link to your landing page + 1 post
   • Photo: a professional headshot, not a selfie

2. A landing page (built with Carrd or Notion)

   • Hero section: what you do and who it's for (one sentence)
   • Services: 3 packages with prices (Starter, Growth, Scale)
   • Case study: 1 case (even a made-up one, clearly labeled "illustrative example")
   • About: your background and why you do this work
   • Contact: a Calendly link to book a discovery call

3. An email signature + Calendly setup

   • Calendly with 30-minute "Discovery Call" slots
   • Email signature: your name + headline + Calendly link

Step 3: Your first case study

If you have real experience bringing AI into a business, write it up as a case study. If you don't, write a made-up case study with a clear disclaimer:

🔍 EXAMPLE: "Anonymous Law Group" (a made-up firm)

PROBLEM: The team spent 15 hours a week on manual document review.

SOLUTION: Set up the Claude API to pre-sort documents.
The attorneys get a summary plus the flagged sections
and make the legal calls themselves.

RESULT: Document review time dropped from 15 to 4 hours a week.
That's attorney time saved (in a real case study, use the numbers you actually measured).

* This is an illustrative example of a typical project, not a real client.
  When you write up real work, change names and figures as your NDA requires.

It does the job. A made-up example with an honest label beats an empty page.

Days 3-4: Outreach list and cold outreach

Step 1: Build an outreach list (50 ICP companies)

Use a mix of:

• Apollo.io: a database of B2B contacts for building prospect lists
• LinkedIn Sales Navigator: if your budget allows
• Hunter.io: for looking up email addresses

These services change their prices, so check the plans on their websites.

Your list = 50 companies × 1-2 decision-makers each (the CEO at a small business, the CMO or CTO at a mid-market company).

ICP (Ideal Customer Profile) criteria:

• Size: 10-100 employees (a rough guide; adjust it to your niche)
• Revenue: enough to have a budget, but no in-house AI team
• Industry: the niche you picked
• Geography: where you're comfortable (language, time zone)
• Signal: a pain that AI can solve (manual processes, slow operations, trouble keeping up as they grow)

Step 2: Cold outreach templates

A cold email template (adapt it to your niche):

Subject: Quick question, [First Name]

Hi [First Name],

I noticed that [Company] is [specific thing; find it on their LinkedIn or blog].
It reminds me of work we did with [similar company/industry]:
they cut [specific metric] by [X%] by bringing AI into
[specific function].

I'm offering a free 30-minute review call for [niche] companies. In one call
we usually find a few quick wins [specifics from your real experience].

Worth a conversation? My calendar is here: [Calendly link]

Best,
[Your name]
[LinkedIn]

Keep the line about work with a similar company only if that work really happened. If you don't have a case yet, cut it and say what you noticed about their processes instead.

A LinkedIn message template (shorter):

Hi [First Name], I saw that [Company] recently [specific thing].

I help [niche] companies put AI to work: [a specific result from your real case].
I'm doing free 30-minute review calls. Interested?

Calendar: [Calendly link]

Step 3: Send 20 messages a day

20 a day × 2 days = 40 messages. Write to the rest of your list next week. Track your numbers:

• Open rate (you only see it if you send through an outreach tool with tracking; if it's low, the subject line is the problem)
• Reply rate (if it's low, the message isn't relevant to them)
• The share that turns into booked calls (if it's low, you aren't hooking into a real pain)

There are no universal benchmarks, so compare your first batches with each other. Don't promise savings you haven't measured.

Never send a generic mass blast. Spend 1-2 minutes personalizing each message (mention a specific post, project or piece of company news). Cold email is regulated by law (for example, the CAN-SPAM Act in the US, CASL in Canada and the GDPR in the EU): check the rules where you are and where the people you're writing to are. The legal section of the lesson Cold outreach that gets replies covers the basics.

Days 5-7: Discovery calls

The plan's target for the end of week 1 is 5-10 discovery calls booked. If you have fewer, that's common with cold messages: keep writing, and add warm contacts, people who already know you (see the lesson How to find your first clients).

Discovery call template (30 minutes):

MINUTES 0-5: Build rapport
- Say hello and thank them for their time
- "Tell me briefly about [Company] and your role"

MINUTES 5-15: Diagnose the pain (the main part)
- "Which 2 or 3 processes take up the most time at your company?"
- "Where do you see the most repetitive manual work?"
- "What have you tried to automate? What worked and what didn't?"
- "Roughly what share of your team's time goes to routine work versus strategic work?"

MINUTES 15-25: Show value (not selling)
- Describe 2-3 AI use cases for their specific situation
- Share a relevant case study (anonymized)
- "In your situation, I'd start with [specific opportunity]"

MINUTES 25-30: Next step (call to action)
- "If you're interested, I can do a formal one-week audit for [price].
   You'd get a detailed roadmap with 5-10 opportunities ranked by ROI"
- "If it's not a fit, that's totally fine. It was great to meet you"
- Schedule the next call or send a proposal

After every discovery call:

Log it in your CRM (the tool or spreadsheet where you track clients and deals; the free version of HubSpot is often enough to start):

• Pain points (what hurts)
• Budget signal (did money come up?)
• Authority (can this person make the decision, or do you need someone else?)
• Timeline (when they want to start)
• A Hot/Warm/Cold rating

Goal for week 1: 2-3 hot prospects → send a proposal by the end of day 7.

WEEK 2: Proposal and contract

Goal for the week: send a proposal and agree on terms (when they sign and pay a deposit depends on the client).

Days 8-10: Proposal and negotiation

Proposal structure (use PandaDoc, Better Proposals or a regular document):

1. EXECUTIVE SUMMARY (1 paragraph)
   "This proposal lays out how [Company] will bring AI into [function]
   and reach [specific outcome] within [timeframe]."

2. CURRENT STATE (1 page)
   - What you do today
   - Where you lose time or money
   - Constraints we need to account for

3. PROPOSED SOLUTION (1-2 pages)
   - 3 key opportunities, each with a description
   - The tools and tech stack you'll need
   - A week-by-week roadmap

4. EXPECTED OUTCOMES (1 page)
   - Specific metrics: $X saved / Y hours freed up / Z% faster
   - Timeline to results: your honest estimate (for example, first quick wins in 2 weeks, the full effect in a few months)
   - Risk factors (be honest)

5. INVESTMENT (1 page)
   - Discovery Audit: [price] (1 week; deliverable: a detailed report)
   - Implementation: [price] (3 weeks; deliverable: a working AI solution)
   - Ongoing Retainer: [price]/month (continuous optimization)
   - Total Year 1: [amount]
   - ROI calculation: only from numbers you've measured or agreed on with the client

6. NEXT STEPS
   - Deposit to start (you set the amount in the contract, for example 50%)
   - Kickoff call within 7 days of signing
   - First deliverable in 7 days

Negotiation patterns:

• "It's too expensive" → "What in the proposal do you find most valuable? What could we take out to fit your budget?" (don't discount right away)
• "We need to think about it" → "What specifically gives you pause? What can I clarify?" (bring the real objections to the surface)
• "We're going to look at other proposals" → "Makes sense. What will be the deciding factor for you?" (learn their criteria)
• "Can you do it for less?" → "I can. If we take X out of scope, it comes to $Y. Does that scope work for you?" (never drop the price without changing the scope)

🎨 Picture this: changing the price without changing the scope is like a store clerk crossing out the number on a price tag and writing a lower one while you watch. The client sees that your prices are random, and trust drops. Changing the scope when the price changes is like ordering at a restaurant: skip the appetizer and the check gets smaller. That's normal.

Days 11-14: Sign and kick off

Step 1: Get the contract signed (Bonsai or DocuSign)

The minimum a contract should cover:

• Scope: what's included and what's NOT included (the second part matters more)
• Timeline: milestones and deadlines
• Payment terms: 50% upfront, 50% on delivery (or by milestone)
• Cancellation clause: what happens if either side walks away from the project
• IP ownership: who owns the deliverables (usually the client, once they've paid in full)
• Confidentiality: a baseline NDA
• Limitation of liability: standard protection for you

Don't write a contract from scratch. Bonsai has ready-made templates for consulting agreements that you can adapt to your project. Have a lawyer review the final contract: contract rules differ by state and by country, and this lesson isn't legal advice.

Step 2: Collect the 50% deposit

Use Stripe (it isn't available in every country; the list is at stripe.com/global) or another payment service that works where you are. Don't start work until the deposit has arrived: a client who stalls on the deposit is a red flag for payment problems later.

Step 3: Kickoff call (60-90 minutes)

Kickoff call agenda:

1. Confirm the project goals (10 min)
   - Make sure everyone is on the same page about the outcomes

2. Stakeholder map (15 min)
   - Who the decision-maker is
   - Who will work with me directly
   - Who needs to be interviewed (usually 5-10 people)

3. Communication cadence (10 min)
   - A weekly sync call (Tuesdays at 10 a.m., for example)
   - A Slack channel or email as the main channel
   - Agreed response times

4. Data access (15 min)
   - Which systems I need access to (read-only wherever possible)
   - An NDA, if one isn't signed yet
   - Security and compliance requirements (including what company data may go into AI tools)

5. Discovery interview schedule (15 min)
   - A list of 5-10 people to interview
   - Book time on their calendars over the next 2 weeks

6. Spot the quick wins (10 min)
   - What clearly hurts and can be tackled first

Step 4: Discovery interviews (5-10 stakeholders, week 2)

45 minutes each, covering:

• What they do (the current process)
• How much time goes to routine work
• What they'd change if they could
• Where they see opportunities for AI
• What they DON'T want AI to do (this matters for trust)

Document the current state and the pain points: that's the raw material for the audit deliverable.

WEEK 3: The discovery audit deliverable

Goal for the week: a polished audit report + a presentation + a commitment to implementation.

Days 15-17: Analysis

After the discovery interviews you have 10-30 pages of raw notes. Now you turn them into a structured audit.

Audit report structure:

1. EXECUTIVE SUMMARY (1 page)
   - Top 3 findings
   - Top 3 recommendations
   - Estimated ROI

2. CURRENT STATE ANALYSIS (3-5 pages)
   - Process map (how things work today)
   - Pain points ranked by impact
   - Time and cost breakdown by process

3. AI OPPORTUNITIES RANKED (5-10 opportunities)

   For each opportunity:
   - Description (what exactly to do)
   - Impact: $X/year saved or Y hours/week freed up
   - Effort: hours or weeks to implement
   - Risk: low / medium / high
   - Dependencies: what has to happen first
   - Recommended priority: P0, P1 or P2 (P0 = do first)

4. RECOMMENDED ROADMAP (1-2 pages)
   - Phase 1 (weeks 1-4): Quick wins, no infrastructure changes
   - Phase 2 (months 2-3): Larger implementations
   - Phase 3 (months 4-6): Strategic AI integration

5. INVESTMENT REQUIRED
   - Tools/subscriptions: $X/month
   - Consulting (implementation): $Y one-time
   - Ongoing (retainer): $Z/month
   - Total Year 1: $W
   - Expected ROI: $Q (ratio Q/W)

6. NEXT STEPS
   - Begin with [specific opportunity]
   - Timeline: start in 1 week
   - Budget: $X to commit

A way to rank opportunities (a simplified version of the RICE method):

Give each opportunity four scores from 1 to 10:

Total score = Reach × Impact × Confidence / Effort. Rank the opportunities by total score: the higher the score, the sooner it's worth doing.

Days 18-21: Polish and present

Deliverable formats (hand over both):

1. A slide deck (15-20 slides): for the presentation
2. A written report (a 10-15 page PDF): to share around the company and keep on file

Presentation meeting (1.5 hours):

0-10 min: Context recap
- What we've been doing for the past 3 weeks
- Who we interviewed

10-30 min: Findings
- Current state map
- Top pain points

30-60 min: Recommendations
- A deep dive into the top 5 opportunities
- The roadmap, visualized

60-75 min: Investment + ROI
- The numbers conversation
- Q&A

75-90 min: Commit to the next step
- "Are you ready to start on opportunity #1 next week?"
- "Whose budget approval do we need?"
- Sign the implementation contract or set a decision date

The goal of the meeting isn't "get some feedback." It's a signed commitment to the implementation phase, or at least a set decision date.

WEEK 4: Implementation and the retainer pitch

Goal for the week: show your first quick win and offer a monthly retainer (whether and when the client signs is their call).

Days 22-25: Your first quick win

From the audit, pick the opportunity with the highest impact and the lowest effort. Examples:

• AI-generated first drafts of email replies (set up in 3-5 days)
• Automatic meeting summaries (Otter or Fireflies + a Claude prompt setup)
• Marketing copy automation (templates + the Claude API)
• A document sorting pipeline (Claude + spreadsheets)
• A customer support FAQ assistant (Claude + a knowledge base)

Implementation pattern:

1. Day 22: Set up the tools, configure the prompts, test on 5-10 examples
2. Day 23: Improve the quality, train the client's team (1-2 key people)
3. Day 24: Run it on real cases
4. Day 25: Measure the results: time saved, quality held, errors

Document the ROI carefully (an example with made-up numbers; plug in the ones you measure):

BEFORE:
- Task: Write 30 customer support responses/day
- Time per response: 5 minutes average
- Total: 30 × 5 = 150 minutes/day = 2.5 hours/day
- Cost: at $50K/year (about $25/hour), 2.5 hours/day × 250 days
  = 625 hours/year = $15,625/year on this task

AFTER:
- Task: Review AI-drafted responses + send
- Time per response: 1 minute (review only)
- Total: 30 × 1 = 30 minutes/day = 0.5 hours/day
- Savings: 2 hours/day × 250 days = 500 hours/year
- Dollar value: $25/hour × 500 = $12,500/year saved

These numbers are your evidence when you pitch the retainer.

Days 26-28: Show results and pitch the retainer

Results meeting (60 minutes):

0-15 min: Demo
- Show the quick win working live
- A before/after comparison

15-30 min: Metrics
- Concrete numbers: hours saved, dollar value
- Quality held steady or improved

30-45 min: What's next
- "This was one opportunity from the audit. There are 4-5 more."
- "To roll all of them out, we need an ongoing engagement."

45-60 min: Retainer pitch
- Tiers (Starter, Growth, Scale, priced from your packages)
- "For your situation, I recommend [tier]"
- Sign or set a decision date

Retainer benefits for the client (your pitch):

• Continuous improvement (not a one-shot rollout)
• Priority access (fast answers to their questions)
• Accumulated knowledge (you already know their business, so every new project goes faster)
• A predictable budget (instead of hourly billing with surprise invoices)
• AI tools change fast, and someone needs to keep up with what's new

Days 29-30: Pipeline and playbook

Day 29: Pipeline review

From a list of 50 companies and 10 discovery calls (a made-up example), you might end up with:

• 1 client (if everything went well)
• 2-3 hot prospects who haven't signed
• 5-7 warm prospects who need a follow-up
• 30-40 cold prospects you can stay in touch with over time

Action items:

• Hot prospects → send a follow-up: "here's what I did for client X" (referencing your case study)
• Warm prospects → schedule check-in calls in weeks 5-6
• Cold prospects → invite them to your monthly newsletter (if you write one)

Day 30: Write down your playbook

Record what worked and what didn't, so you can repeat it:

• Which outreach template got the best reply rate
• Which discovery call questions led to hot leads
• Which proposal sections held up in negotiation
• Which quick wins impressed the client most
• Where you lost time, and what to delegate next time

This playbook is an asset that grows with every client. Client #2 should go faster, and client #5 faster still.

💼 Service packages: 3 tiers

Standardize your pricing: it saves time on every deal. The amount of work in each package below is an example structure; you set the prices yourself. You worked through how to build packages in the lesson Packaging your AI services into three clear offers.

Starter

For: small businesses with 5-15 people working with an AI consultant for the first time

• 4 hours of consulting/month
• 1 monthly strategy call (60 min)
• Slack/email access (replies within 24 hours)
• Quarterly business review
• Setup of the recommended tools

Growth

For: small and mid-size businesses with 15-50 people that already use AI in some everyday processes

• 8 hours/month
• 2 strategy calls (60 min each)
• 1 implementation project/month (one quick win)
• Monthly business review
• Slack replies within 4 hours
• Advice on which tool subscriptions to keep

Scale

For: mid-market companies with 50-150 people going through a serious AI transformation

• Unlimited reasonable access (up to 20 hours/month)
• Weekly strategy calls
• Hands-on AI implementation help (where it applies)
• On call for critical issues (4-hour response)
• Quarterly executive briefing
• Custom training for the team

🎯 Niche ideas for 2026

The table below is the author's rough guide, not market statistics: check the demand in your own niche.

Niches where budgets tend to be bigger:

In law, healthcare and finance, you automate the admin work and paperwork around licensed professionals. The AI doesn't give legal, medical or financial advice, and the professional stays responsible for the work.

Niches with mid-size budgets:

• Marketing agencies (they sell AI as a service to their own clients: a bit recursive, but real)
• SaaS companies that want AI features (helping them build AI in-house)
• Online course creators (content production + student support)
• Coaches and consultants (also a recursive niche, but there's a market)

Which one should you pick first? The one where you have a personal connection or past experience. Breaking into an unfamiliar industry cold takes much longer than working in one you already know.

⚠️ Common pitfalls

🛠 Tools you need

Add up the plans for the services you choose yourself. Start with the minimum (free tiers plus one paid tool for finding clients) and add the rest as you need it. Current prices: What's current.

What you don't need: a corporate website, a custom CRM, a fancy office, business cards, or a registered company on day one. When the time comes, the common options for a solo consultant in the US are a sole proprietorship or an LLC; choose the setup with your accountant or a lawyer. The rules for working for yourself and for taxes differ by state and by country, so check them before you send your first invoice.

📊 The real economics

How to figure your effective hourly rate:

• Divide the client's monthly fee by the hours the work actually took (calls, prep, emails)
• Count in the hours you spend finding clients and doing admin
• Compare the result with your minimum rate: if it's lower, change the amount of work, the price or what's in the package

Why a retainer beats hourly billing:

Hourly, you're selling time. On a retainer, you're selling access plus results. An hourly client will try to keep your hours down. A retainer client pays for peace of mind: you're in their corner.

Ways to scale:

There are no income figures here: they depend on your niche, your market and your work.

✅ Your checklist for the end of Day 30

Check yourself against this list:

• ✅ A first proposal sent to a client (or already signed, if the client was ready)
• ✅ A repeatable engagement playbook (notes on what worked)
• ✅ Personal brand assets (LinkedIn + landing page + 1 case study)
• ✅ A pipeline of 3-5 prospects (warm, not cold)
• ✅ A tool stack that works
• ✅ Tested outreach templates (you know what works in your niche)
• ✅ A refined discovery call script (from 10 calls)
• ✅ A proposal template (tested on at least one client)
• ✅ Experience: you've been through the whole process once and know what to change

This is a starting point you can act on right away. If some items are missing, keep working the same steps next month: the outcome depends on your niche, your market and your work.

🚀 A plan for year 1

This is a rough outline of stages, not an income forecast.

Months 1-3: Foundation

• Month 1: the 30-day plan from this lesson
• Months 2-3: widen your funnel and look for the next clients

Months 4-6: Growth

• Grow through a repeatable process, not one-time luck
• Spend 1-2 hours a week on content (LinkedIn posts, case studies)
• Inbound leads may start coming in (referrals + content)

Months 7-9: Optimize

• Review your clients and packages: which ones are worth your time and which aren't
• Bring in a helper for admin and scheduling if you're short on time

Months 10-12: Scale or specialize

• Path A (solo premium): fewer clients, deeper work
• Path B (hybrid): retainers plus separate implementation projects
• Path C (productize): launch a course or templates as a separate product

The actual income and workload depend on your niche, your market and your work.

Related lessons

• Your first clients: where to find them and how to close
• Finding potential clients: an optional lesson in the deep-dive library on collecting lists of companies into a spreadsheet
• Cold outreach with Claude: subject lines, follow-ups and sequencing
• The factory model: templates and repeatability: how to set up solo consulting as a business

Tools and resources

• Apollo.io: a B2B prospect database for outbound
• HubSpot Free CRM: a free CRM to start with
• PandaDoc: proposals with e-signature
• Bonsai: contracts + invoicing for freelance consultants
• Calendly: booking discovery calls
• Hunter.io: email lookup for cold outreach
• Better Proposals: an alternative to PandaDoc
• DocuSign: e-signatures if you don't need Bonsai (check its current plans)
• Stripe: payment processing
• Y Combinator: Startup School: frameworks for validating ideas (they apply to services too)
• McKinsey Insights: a reference for proposal language ("McKinsey-level briefs")

Key takeaways

Consulting is you as the product. It doesn't scale like SaaS, but it starts faster, because there's no software to build and no infrastructure to run. It suits people who want to put their AI know-how to work without a big upfront investment; the outcome depends on your niche, your market and your work.

The funnel: Audit → Implementation → Retainer. Never open with a retainer: it's too big a step for a client who doesn't know you yet. The audit is the first date; the retainer is the marriage.

Niche = Industry × Function. "AI consultant" is for no one. "AI for immigration lawyers" makes it clear who you're selling to. Without a niche, marketing, growth and repeat clients are all out of reach.

30 days of following the plan give you a direction: a first proposal to a client + a reusable playbook + a pipeline. The result isn't guaranteed: it depends on your niche, your market and the quality of your outreach. Track your funnel on your own data: how many contacts → replies → calls → proposals → agreements.

A one-person consulting practice with no employees and no outside funding is also a standalone, detachable project in your portfolio (see Detachable projects, an optional lesson in the deep-dive library). Income figures depend on your niche, your market and your work, which is why this lesson doesn't give any.

Next lesson

→ The factory model: templates for repeatable client work: how to get faster with every similar project