Every small business runs on a set of repeatable processes — writing proposals, pricing jobs, onboarding new clients, handling common decisions, documenting how things get done. Most of the time, these processes live inside someone's head, get recreated from scratch each time, or rely on whoever's available.
That's not a people problem. It's a systems problem. And for the first time, you can solve it without hiring an operations consultant, a developer, or a business analyst. You just need AI and a few hours.
This final playbook closes the loop on everything you've built in Playbooks 1–5. Your website is live, your content is running, your CRM is automated, your support is handled, your data is analysed. This one makes the underlying business engine run consistently — regardless of who's doing the work.
✅ What you'll have when done: A complete toolkit of AI-powered internal tools — a custom proposal generator, a pricing calculator prompt, a full SOP library for your core processes, a client onboarding toolkit, a hire/no-hire decision framework, and a standard decision prompt you can apply to any recurring business choice.
What Is an AI Internal Tool?
An AI internal tool is any prompt or AI workflow that automates a task your business does repeatedly. You build it once, save it, and anyone on the team can run it consistently — without needing the original person who knew how to do it from memory.
The Tools Worth Building First
Not every business needs every tool. These are the 6 highest-ROI internal tools for most small businesses — ranked by how much time they save and how directly they impact revenue:
Proposal Generator
Turns client intake info into a complete professional proposal
Pricing Calculator
Codifies your pricing logic so anyone can quote accurately
SOP Library
Documents every repeatable process so anyone can run it
Client Onboarding
A complete new-client experience, automated and consistent
Hiring Rubric
Structured AI-powered candidate evaluation framework
Decision Framework
Turns recurring hard decisions into structured AI analysis
Before and After — What These Tools Actually Change
| Task | Before AI Tools | After This Playbook |
|---|---|---|
| Writing a proposal | 2–4 hours of writing from scratch, every time | 15–20 minutes: fill in inputs, review AI output, send |
| Quoting a job | Mental math + gut feel, inconsistent between team members | Structured prompt produces detailed line-item quote in 5 minutes |
| Onboarding a new client | Ad hoc — whoever does it first decides what to send | Documented 5-step onboarding sequence, same every time |
| Training a new hire | "Shadow me for a week" — knowledge stays in one person's head | Written SOP for every process; new hire can self-train |
| Making a hard business decision | Deliberate for days, then go with gut, second-guess yourself | Run decision framework prompt — structured analysis in 10 minutes |
⚠️ Affiliate disclosure: Some links on this page may be affiliate links. We may earn a commission at no cost to you. We only recommend tools we actively use and test.
The 6-Step Playbook
Map Your Repetitive Tasks
20 min · Claude · Identify exactly which tasks to automate first
Before building any tool, you need to know which processes are costing you the most time and producing the most inconsistent outcomes. This prompt audits your business operations and outputs a ranked list of tool-building candidates:
You are a business operations consultant helping me identify which internal processes to systematise first using AI tools. My business: [name and what I sell] Team size: [e.g. just me / me + 2 contractors / team of 5] Revenue stage: [e.g. under $100k/yr / $100k–$500k/yr / over $500k/yr] Biggest time drains right now: [e.g. writing proposals, answering the same client questions, quoting jobs] Walk through each of these operating areas with me and help me identify which tasks should be turned into AI internal tools: OPERATIONS AREAS TO AUDIT: 1. Sales process: [how do you find, qualify, and close clients? where does it slow down?] 2. Proposal / quoting: [how long does it take? how consistent is it? does it vary by person?] 3. Client onboarding: [what happens the first week after a client signs? is it documented?] 4. Service delivery: [how is the actual work done? are there repeatable steps? SOPs?] 5. Communication: [how do you handle routine client updates, check-ins, follow-ups?] 6. Hiring / team management: [how do you evaluate candidates? train new people? manage performance?] 7. Finance / admin: [how are invoices, expenses, monthly reviews handled?] 8. Decision-making: [what recurring decisions take the most mental energy?] For each area I describe to you, evaluate: - Time currently spent: how many hours per month does this take across the whole business? - Inconsistency risk: how likely is the output to vary based on who does it or how they're feeling that day? - AI suitability: can a structured prompt reliably handle 80%+ of this task? - Priority score: rank 1–10 for tool-building priority OUTPUT FORMAT: 1. TOP 5 TOOL OPPORTUNITIES — ranked by time savings × consistency impact, with a one-sentence description of what each tool would do 2. QUICK WIN — one tool I could build and start using within 2 hours that would have immediate impact 3. OPERATIONS GAP ANALYSIS — areas where I described a process but it's clearly undocumented or inconsistent — these need SOPs before they need AI tools 4. SUGGESTED BUILD ORDER — the sequence to build these tools in, with reasoning
Build a Proposal and Quote Generator
45 min to build · 15 min per proposal · Claude · Highest-ROI tool for service businesses
For most service businesses, writing proposals is the single most time-consuming sales activity. This prompt turns your existing proposal style, pricing, and client intake information into a custom proposal generator that produces professional, on-brand proposals in minutes.
You'll run this prompt once to create your generator — then save the output as your reusable proposal tool:
You are a business writing expert helping me build a custom AI proposal generator for my business. My business: [name and what I sell] My ideal client: [describe who you typically work with — industry, size, situation] My services / packages: [List each service or package with a brief description and typical price range] [e.g. "Brand Strategy Package — 4-week engagement, $3,500–$6,000 depending on scope"] My current proposal structure (describe what your proposals typically include, or say "I don't have a consistent structure"): [e.g. Cover page, executive summary, proposed approach, deliverables, timeline, investment, next steps] My tone and style: [e.g. professional but warm / formal / conversational and direct / premium/luxury feel] A recent proposal I was happy with (describe it briefly or paste sample sections): [Paste or describe a proposal section you liked, so I can match the voice] PART 1 — BUILD MY PROPOSAL GENERATOR: Create a custom AI prompt I can use as my proposal generator. The prompt should: - Ask me to fill in 8–12 client-specific inputs (the variables that change per proposal) - Use my services, pricing, tone, and structure automatically - Output a complete ready-to-send proposal with all sections - Include a pricing presentation section that presents my fee in the most favourable way - Include a next steps section with a clear call to action - Be designed so that I (or a team member) can run it in under 20 minutes per proposal PART 2 — TEST IT NOW: Using the generator you just built, create a sample proposal for this fictional client: [Describe a typical ideal client scenario — e.g. "A 12-person law firm in Chicago that needs a new website. Budget is $8,000–$12,000. They've had bad experiences with agencies before and care about communication. Timeline is 8 weeks."] The output should be ready to send with only minor personalisation needed.
💡 Save both parts. Save the generator prompt (Part 1 output) to your Notion or Google Docs tool library. Save the sample proposal as a style reference. Each time you need a proposal, fill in the 8–12 client inputs and run it. The generator improves over time — after a few uses, add notes about what clients responded well to.
Build a Pricing Calculator Prompt
30 min · Claude · End inconsistent quoting forever
Inconsistent pricing is one of the most common small business problems — different team members quote different prices for similar jobs, discounts get given without logic, and complex jobs get underquoted because someone forgot a line item. This prompt codifies your pricing logic into a structured calculator that anyone can use:
You are a pricing strategist helping me build a consistent AI-powered pricing calculator for my business. My business: [name and what I sell] Business model: [project-based / hourly / retainer / productised service / product sales / mix] My current pricing structure (describe as much detail as you know): [e.g. "We charge $150/hr for strategy work, $95/hr for execution, $75/hr for admin. Projects are typically 20–80 hours. We sometimes offer a 10% discount for long-term commitments. Rush work is 1.5x rate."] [If you sell products, describe cost of goods, typical margins, volume pricing, etc.] Factors that affect my pricing (what makes a job more or less expensive?): [e.g. timeline urgency, project complexity, client size, revision rounds, deliverable count, travel required] Common mistakes or inconsistencies in our current quoting: [e.g. forgetting to add rush fees, underestimating revision time, not charging for discovery] Discount policy: [e.g. no discounts / 10% for 3+ month commitments / case by case] PART 1 — BUILD MY PRICING CALCULATOR: Create a custom AI pricing prompt that: - Walks through all relevant pricing factors systematically - Applies my pricing logic consistently with no missed line items - Provides a detailed line-item breakdown the client can see - Shows a recommended investment range (low/mid/high) based on scope - Flags any scope items that typically cause budget overruns - Includes a recommended way to present the price (anchoring, packaging, etc.) PART 2 — BUILD A PRICING FAQ TOOL: Create a second prompt I can use to answer common client pricing questions: - "Why is this more expensive than I expected?" - "Can you do it cheaper?" - "What's included vs. extra?" - "How does your pricing compare to others?" These responses should be professional, confident, and value-focused — never defensive. PART 3 — TEST BOTH: Run a test calculation for this scenario: [describe a typical job or order with its relevant details]
⚠️ Price anchoring matters. How you present a price affects whether it gets accepted. When the AI outputs your pricing, it should present the mid-tier option first (not the lowest), frame the investment in terms of the outcome value, and always include at least two options so the client is deciding between your options — not between you and a competitor.
Create Your SOP Library
60 min · Claude · Document every process so the business runs without you
An SOP (Standard Operating Procedure) is a documented, step-by-step description of how a task gets done in your business. Most small businesses don't have them — which means every task relies on the person who's "always done it that way." This creates fragility: one sick day, one departure, one new hire and the process falls apart.
AI can write SOPs dramatically faster than doing it manually. You describe the process, AI structures it. Here's the anatomy of a good SOP, plus the prompt to generate them:
Purpose & Scope
One sentence: what this process does and who runs it. Removes ambiguity about when to use it.
Trigger
What event or condition starts this process? (e.g. "When a client signs the contract…")
Inputs Required
What information, access, or materials are needed before starting?
Step-by-Step Instructions
Numbered steps, specific enough for a new person to follow without asking questions. Screenshots where relevant.
Decision Points
If X, do Y. If Z, do W. Every common fork in the process is handled explicitly.
Expected Output
What does "done" look like? What has been produced or confirmed when this SOP is complete?
Common Mistakes
The 3 things most often done wrong. Prevents repeat errors without repeating the conversation.
You are a business operations specialist helping me document a business process as a clear, usable SOP. My business: [name and what I sell] Process to document: [e.g. "New client onboarding", "Monthly invoicing", "Responding to a support ticket", "Closing a sales call"] Who runs this process: [e.g. me / my VA / anyone on the team] Experience level of the person who might do this: [e.g. experienced / new hire / contractor unfamiliar with our systems] Here is how I currently do this process (describe it in your own words — don't worry about structure, just brain-dump it): [Describe the process as you currently do it, including every step you can remember, the tools you use, common variations, and anything that often goes wrong] Any tools, logins, or systems involved: [e.g. HubSpot, Stripe, Google Drive, Notion, Slack, our project management tool] Common mistakes or things that often get forgotten: [List anything that has gone wrong before or tends to get skipped] Create a complete, usable SOP following this structure: 1. PROCESS NAME: [short, clear title] 2. PURPOSE: One sentence — what this process accomplishes and why it matters 3. TRIGGER: What event or condition starts this process 4. BEFORE YOU START: What access, information, or prep is needed 5. STEP-BY-STEP INSTRUCTIONS: Numbered, specific, clear enough for a first-timer 6. DECISION TREE: Common "if this, then that" situations in this process 7. EXPECTED OUTPUT: What does completed look like? What has been produced? 8. QUALITY CHECK: 3–5 things to verify before marking this process done 9. COMMON MISTAKES: Top 3 errors to avoid 10. RELATED SOPS: Other processes this connects to (I'll fill these in later) Write in plain, direct language. Assume the reader is competent but unfamiliar with this specific process.
ℹ️ Build your SOP library in Notion. Create a database with one page per SOP. Tag each by category (Sales, Delivery, Finance, HR, Marketing). This becomes your business operations manual — and the foundation for training anyone new, whether that's a hire, a VA, or an AI agent running automated workflows.
Start with your 5 most frequently run processes. Run the SOP generator once per process. Review, correct anything AI got wrong based on your description, and save it. Five SOPs done in an afternoon.
Build a Client Onboarding Toolkit
30 min to build · 5 min per new client · Claude · First impressions on autopilot
The first 7 days of a client relationship set the tone for everything that follows. Clients who feel informed, welcomed, and confident in week one have higher retention, higher satisfaction scores, and generate more referrals. Most small businesses leave this to chance — this prompt builds an onboarding toolkit that runs consistently every time:
You are a client experience specialist helping me build a complete new-client onboarding toolkit. My business: [name and what I sell] Typical engagement: [e.g. 3-month project / ongoing monthly retainer / one-time service / product purchase] What clients typically worry about at the start: [e.g. timeline slipping, not knowing what's happening, getting what they paid for] What makes our onboarding currently inconsistent or incomplete: [e.g. we forget to send resources, communication drops off after signing] Build me a complete client onboarding toolkit with: 1. WELCOME EMAIL — A warm, confident email sent within 1 hour of signing that: - Confirms what they've bought and what comes next - Sets expectations for communication cadence and timeline - Gives them 1 thing to do that makes them feel engaged immediately - Closes with genuine enthusiasm (not corporate filler) Variables to customise: [CLIENT_NAME], [SERVICE_PURCHASED], [START_DATE], [PRIMARY_CONTACT] 2. ONBOARDING QUESTIONNAIRE — 8–10 questions to gather everything you need to start work. Questions should feel like a conversation, not a form. Cover: - What success looks like to them - Key contacts, approvers, decision-makers - Previous experiences (good and bad) with similar services - Practical logistics (preferred communication, availability, tools they use) - The one thing they most want us to know 3. WEEK 1 CHECK-IN MESSAGE — A brief, friendly message at end of day 3 that: - Confirms where things stand - Surfaces any confusion before it becomes frustration - Makes them feel like they made the right decision 4. END OF WEEK 1 SUMMARY — A structured week-1 wrap-up message with: - What was accomplished this week - What's coming next week - One specific thing to expect from them - A confidence-builder: why this is going well 5. 30-DAY ONBOARDING CHECKLIST — A master checklist of every onboarding touchpoint, in order, with timing and owner (me vs. client vs. automated)
💡 The single most important onboarding moment is the first hour. Clients make a judgment about whether they made the right decision within the first few hours after signing. A well-crafted welcome email sent within 60 minutes of signing does more for retention than anything you send in month 2. The AI can draft it in 90 seconds — there's no excuse for a 24-hour delay.
Build a Decision Framework Tool
20 min · Claude · Consistent, reasoned answers to your hardest recurring choices
Every business has a set of recurring hard decisions — whether to take on a client, whether to hire someone, whether to drop a service, whether to discount a price, whether a project scope has expanded beyond what was agreed. These decisions are often made inconsistently: sometimes by gut, sometimes by whoever's loudest, sometimes by whoever's most tired.
A decision framework prompt turns these recurring choices into a structured analysis — same inputs, same process, same reasoning quality, every time:
You are a business advisor helping me think through an important decision using a structured framework. My business: [name and what I sell] My current business stage / priorities: [e.g. growing revenue / protecting margin / building team / reducing owner hours] THE DECISION I NEED TO MAKE: [Describe the decision clearly — e.g. "Whether to take on a new client who wants our services at 30% below our standard rate because they say it will lead to more work later."] THE OPTIONS I'M CONSIDERING: Option A: [e.g. Accept the work at the discounted rate] Option B: [e.g. Decline and move on] Option C: [e.g. Counter-propose at a different structure] (add more options if relevant) RELEVANT CONTEXT: - Financial situation: [e.g. pipeline is healthy / slow month / we have capacity / fully booked] - Relationship with this party: [e.g. brand new / existing client / referred by a VIP] - Time / resource impact: [e.g. 15 hrs/week for 3 months / a single 2-day project] - What I'm leaning toward and why: [be honest — this helps identify confirmation bias] - What's making me hesitate: [what's the nagging doubt or opposing consideration?] - Have I faced a similar decision before? What happened? Run a structured decision analysis: 1. REFRAME THE DECISION — What is this decision actually about? (Often the stated decision masks the real one) 2. CRITERIA THAT MATTER — Based on my priorities, what are the 4–5 factors that should drive this decision? Weight each. 3. OPTION ANALYSIS — Score each option against the criteria. Show the reasoning, not just the score. 4. RISK ASSESSMENT — For each option: what's the best case? worst case? most likely case? 5. CONFIRMATION BIAS CHECK — Based on what I said I'm leaning toward, what might I be unconsciously weighting too heavily or discounting? 6. THE RECOMMENDATION — Given all of the above, what would you recommend and why? Be direct — don't hedge. 7. IF YOU PROCEED — If you take the recommended action, the first 3 steps to take in the next 48 hours.
ℹ️ Save a version of this prompt for each recurring decision type. Build a "Should I take this client?" version, a "Should I hire this person?" version, a "Should I drop this service?" version. Each one is pre-loaded with your relevant criteria and priorities so it takes 5 minutes to run, not 20.
The confirmation bias check is the most valuable part. AI is particularly good at identifying when you've described a situation in a way that's already tilted toward the outcome you want — and surfacing the reasoning on the other side that you may have unconsciously downplayed.
Organising Your Tool Library
Once you've built these 6 tools, the way you store and share them determines whether they actually get used. The simplest system that works for most small businesses:
- Create one Notion page or Google Doc called "AI Tool Library." Sections for each tool category: Sales, Operations, HR, Finance, Communication.
- Each tool is a saved, named prompt with the variables clearly marked in [brackets] and a note at the top: "When to use this" and "Inputs needed."
- For team use: share the Notion page. Anyone can copy a prompt, fill in the variables, and get a consistent output without needing to know how to prompt AI from scratch.
- For advanced use: create a Custom GPT in ChatGPT (Settings → My GPTs → Create) and load your most-used tool prompt as its system instructions. Share the link with your team — one click, fully loaded, no copy-paste required.
⚠️ AI tools drift if you don't maintain them. When your business changes — new services, new pricing, new brand voice — your tools need updating too. Set a reminder every quarter to review your tool library. Outdated prompts produce outdated outputs, which is worse than no tool at all because people trust what the AI produces without questioning it.
Playbook Summary — Your Internal Tools Checklist
- Operations audit completed — top 5 tool opportunities identified and ranked
- Quick win tool identified and built in first session
- Proposal generator built and tested on a sample client scenario
- Proposal generator saved to Notion/Google Doc tool library
- Pricing calculator prompt built with all pricing factors codified
- Pricing FAQ response tool built for handling common objections
- SOPs written for your 5 most frequently run processes
- SOP library organised in Notion with categories and tags
- Client onboarding toolkit complete — welcome email, questionnaire, week-1 sequence
- Onboarding checklist documented with timing and owner per step
- Decision framework prompt built and saved for at least 2 recurring decision types
- AI Tool Library page created and shared with your team
- Quarterly review reminder set to update tools as business evolves
🎉 You've Completed All 6 Playbooks
Your website is live, your content is running, your CRM is automated, your support is handled, your data is analysed, and your business engine is documented. You've built in 6 focused sessions what most businesses spend years trying to figure out. Now it compounds.
Back to Home All 6 PlaybooksFrequently Asked Questions
What counts as an internal tool?
Any prompt or AI workflow that automates a task your business does repeatedly — a proposal generator, a pricing calculator, an SOP template, a hiring rubric, a client communication sequence. If you do it more than once and it follows a consistent structure, it can become an AI tool.
Do I need to know how to code to build these tools?
No. Every tool in this playbook is a structured prompt — no code, no developer, no platform subscription required. You fill in inputs, AI returns output. For team sharing, a Notion page or Custom GPT link is all you need.
How do I share internal tools with my team?
The simplest method is a shared Notion page or Google Doc with all prompts saved and labelled. Each team member copies the prompt, fills in the variables, and pastes into Claude or ChatGPT. For a more structured experience, create a Custom GPT and share the link — one click, fully loaded.
What's the fastest internal tool to build first?
The proposal generator delivers the fastest ROI for most service businesses — reducing 2–4 hours of manual writing per proposal to 15–20 minutes, with a more professional, consistent output. If you're product-based, the pricing calculator typically wins.
How do I keep tools consistent and on-brand?
Include your brand voice guidelines, specific formatting instructions, and at least one example of output you like in the tool prompt. Run each tool 3–5 times with varied inputs to test it, then refine any parts that drift from your standard. Review the full tool library quarterly.























