For Business Owners — 2026 Edition

Claude for
Business Owners

The practical setup guide — projects, file architecture, skills, agents, financial security, connectors, and team policy. Written by an owner, for owners.

01Why structure beats prompting
02Project organisation
03File & folder architecture
04Skills files
05Your personal voice file
06Claude as your CFO advisor
07Data security — traffic lights
08Connectors — email & cloud
09Building agents
10Team AI usage policy
11Beginner best practices
12Your 90-day plan
Section 01

Why structure beats prompting

Most business owners are prompting their way through AI. The ones pulling ahead are building systems around it.

The gap between a business that uses Claude well and one that doesn't isn't the quality of their prompts. It's whether they've built the scaffolding that makes Claude work consistently, at scale, across a team.

Without structure, every team member starts from scratch in every conversation — re-explaining your brand, re-describing your customers, re-stating your tone of voice. Claude is equally capable in conversation one and conversation one hundred. But without persistent context built into projects and knowledge files, every session costs you setup time that compounds across a team.

This guide covers what that scaffolding looks like in practice: how to organise projects, where files should live, how to give Claude standing knowledge about your business, how to handle sensitive financial data safely, and how to build the first automations that actually save meaningful time.

The right plan — non-negotiable

Personal accounts are a data risk — move to the Team plan before anything else

On a personal Claude.ai account, your conversations may be used for model training. The Claude Team plan changes this: your data is not used for training, nothing is retained after the session ends, and you get admin controls over all users. Everything in this guide assumes the Team plan. If any team member is still on a personal account, fix that first.

Four things that separate owners who get results from those who don't

01 — Foundation
Organised projects with the right access

A project structure that reflects who should see what. Shared projects for the team. Role-specific projects for sensitive functions. Private projects for the owner.

02 — Knowledge
Skill files written once, applied everywhere

Your brand voice, product knowledge, and processes written as structured documents. Every project draws from them. Update once — it improves everywhere simultaneously.

03 — Security
Clear rules about what goes in and what doesn't

A simple traffic light system the whole team follows. Financial data and sensitive files handled with a per-session pattern that gives Claude access without storing it permanently.

04 — Governance
A one-page policy everyone has read and signed

A written AI usage policy. A quarterly review. Human approval before anything AI-generated goes to a customer. Simple, but most businesses skip it entirely.

Section 02

Project organisation

Claude projects are persistent workspaces with their own instructions, knowledge files, and conversation history. Getting this structure right is the single highest-leverage thing you can do.

A project is a dedicated role. Its system prompt tells Claude what it is in this context. Its knowledge files give it the information it needs. Every conversation in the project inherits both — so your team is always working with a Claude that already knows your business.

Claude has no sub-folders — your naming convention is your folder system

Projects appear as a flat list. There is no way to group them into folders. With multiple functions, customers, and team members, this list grows quickly. A strict naming prefix convention is the fix — it creates visual grouping and makes the list scannable. Set the convention on day one and enforce it. Changing it later when you have 15 projects is painful.

The three-tier model

Tier 1 — Shared team projects All staff · visible in the Team tab
Brand & Voice
Tone, guidelines, DO/DON'T rules. The foundation project every other project references.
Customer Service
Scripts, FAQs, escalation paths, response templates, complaint handling.
Content & Comms
Blog posts, social content, newsletters, proposals — all on-brand.
Product Knowledge
Specs, FAQs, comparison notes, vendor context. What your team needs to answer anything.
Supplier Manager
Vendor evaluation, procurement, RFQ drafting. Point to specific supplier folders per session.
Operations
Process documentation, internal comms, HR templates, onboarding guides.
Tier 2 — Role-specific projects Restricted by function · not cross-visible
Sales Workspace
Meeting prep, account notes, follow-up emails, pipeline context. Sales team only.
Major Account — [Name]
One project per key account above your threshold. History, contacts, pricing context, meeting prep.
General Sales
All other accounts. Upload relevant account file per session. No dedicated project needed below the threshold.
Tier 3 — Restricted · named individuals only Access controlled via cloud storage permissions — not Claude settings
CFO Advisor
Financial analysis, cashflow, P&L, margins. Each authorised person creates their own private project. Cloud storage permissions are the gatekeeper.
How restricted access works
Claude has no "share with specific members" option. Workaround: each authorised person creates an identically-configured private project pointing at the same restricted folder.
Adding someone
Grant them access to the restricted cloud folder. They create their own private project with the same system prompt. Cloud storage does the access control.
Tier 4 — Fully private · one person only Never shared · invisible to the team
Strategy
Business direction, competitive thinking, board prep, partnership development.
Personal Voice
Your personal writing style file. Uploaded per session into shared projects when needed. Never in the shared skills folder.
Individual workflows
Any team member's own private workspace. Fine to have — but must reference company skill files from cloud storage, not local copies.
Recommended naming convention

Use a consistent prefix per tier. Example: CO — Brand & Voice · CO — Customer Service for shared projects. SALES — [Client name] for accounts. No prefix for private and restricted projects — they don't appear in the Team tab anyway. Agree the convention on day one. Everyone uses it.

Section 03

File & folder architecture

Your files are Claude's long-term memory. Two storage locations with different rules. Getting this distinction right is the most important architectural decision in your Claude setup.

There are two places a file can live. One makes it permanently available to everyone in the project. The other gives Claude access for one session only, controlled by your cloud storage permissions. Sensitive data always uses the second.

Location A — Project Files tab
Permanently embedded in the project

Claude reads this at the start of every conversation in the project. Anyone with project access can see it. Use for skill files, brand guidelines, product knowledge — non-sensitive, always-needed context only.

Location B — Cloud storage only
Uploaded per session · controlled by folder permissions

The file lives in your cloud storage with its own access permissions. Uploaded at the start of a session when needed. Gone from Claude when the session ends. Never embedded in the project. Financial data, margins, sensitive pricing — always here.

Cloud storage folder structure

Create one root folder — /company-claude/ or similar — in SharePoint, Google Drive, or Dropbox. Build this structure inside it:

/skills/
brand-voice.md
product-kb.md
sales-playbook.md
customer-service.md
cfo-framework.md
supplier-criteria.md
Read: all staff. Write: owner only.
/projects/
brand-guidelines.pdf
/sales/ ← role-restricted
  margins.xlsx
  account-pricing.xlsx
/suppliers/ ← per session
/templates/
Role-based. Sensitive files: per session only.
/finance/
kpi-framework.md
budget-template.xlsx
cashflow-model.xlsx
— — — — — —
Actuals: upload per
session, never store
Authorised access only.
/outputs/
/approved/
/drafts/
YYYY-MM-DD naming
Author tagged
Version history on
= free audit trail
All staff read/write. Reviewed content only.

Which location for which file — quick reference

File type Project Files tab Cloud storage only
Brand voice, tone of voice rules ✓ Always here
Product knowledge, specs, FAQs ✓ Always here
Sales playbook, objection handling ✓ Always here
Margin data, account-level pricing ✗ Never here ✓ Per session
Sales pipeline, deal values ✗ Never here ✓ Per session
Supplier-specific data, price lists ✗ Never here ✓ Per session
P&L, cashflow, management accounts ✗ Never here ✓ Per session
Customer personal data (GDPR) ✗ Never here ✗ Not in Claude
The danger of private copies of company skill files

Every team member can create private projects and attach their own files. This is fine for personal workflows. It becomes a problem when someone downloads a copy of product-kb.md or brand-voice.md and attaches it locally — that copy will drift out of date silently. When you update the source in cloud storage, their private copy doesn't update. The rule: if it describes how your company works, it lives in cloud storage and is always referenced from there. If it describes how one person works, it can be private.

Section 04

Skills files

Write your knowledge once. Make it available in every project. Stop re-explaining your business to an AI that can't remember between sessions.

A skill file is a Markdown (.md) document that gives Claude deep, standing knowledge about one topic. Think of it as the briefing document you'd hand a brilliant new contractor on their first day — written once, applied every time.

Priority skill files — build these first

brand-voice.md
Your most important file. Tone of voice, DO/DON'T examples, sentence rhythm, words you never use, how you talk to different audiences, your positioning statement. Attach to every customer-facing project.
product-kb.md
Product or service knowledge base. Key offerings, specs, common questions and correct answers, pricing tiers (no actuals — ranges only), vendor or partner context. Pull from your existing materials. Attach to customer service and sales projects.
sales-playbook.md
Common objections and how to handle them. Buyer personas. Key differentiators vs competitors. Your typical sales stages. Attach to sales and account-specific projects.
customer-service.md
Your support process. How returns, complaints, warranties, and escalations work. Standard response templates. Tone for difficult situations. Keeps every interaction consistent regardless of who handles it.
cfo-framework.md
How you measure your business. KPI definitions, budget categories, financial terminology you use internally. Framework only — no actual figures. Permanent file in your CFO Advisor project. Numbers upload per session.

Skill file structure — use this every time

1. Purpose line — one sentence: what this file is and when Claude should apply it.
2. Core rules — 5–10 specific things Claude must always do or never do. Specific beats vague every time.
3. Before / after examples — 3–5 pairs showing the wrong version and the right version side by side. This is the single most powerful thing you can include.
4. Reference data — facts, lists, templates Claude needs to do the job. What would a new hire look up in week one?
5. Last updated date — one line at the bottom. Revisit every 6 months.

Weak rules — vague, ignored

"Be professional and friendly"
"Write in our brand voice"
"Keep responses concise"
"Be helpful to customers"

Strong rules — specific, enforceable

"Never use exclamation marks in emails"
"Lead with the outcome, not the feature"
"Max 20 words per sentence in customer copy"
"Never use 'innovative' — use specific proof"

Section 05

Your personal voice file

A separate category from company skill files. Private to you. Stored differently. Used when Claude needs to write as you specifically — not as the company.

Your personal voice file is distinct from the company skill files in Section 04. It captures how you write — your sentence rhythm, your level of formality, your characteristic phrases, what you never say. It belongs to you, not to the business. It should never be in the shared /skills/ folder.

your-voice.md
Private — not in the shared /skills/ folder. Stored in a private location only you can access. Permanently attached to your private projects (Strategy, CFO Advisor). Uploaded per session into shared projects when you need Claude to write in your voice specifically.

How to build it in 20 minutes

1
Find 6–8 examples of writing you're proud of

Emails, LinkedIn posts, proposal intros, anything that sounds like you at your best. Recent is better than old.

2
Paste them into a new Claude conversation and use this prompt

Prompt to use
"Analyse these writing samples and extract a detailed personal style guide. Cover: sentence length and rhythm, level of formality, how I open and close messages, balance of technical vs plain language, vocabulary I favour, things I never write, and 5 concrete before/after examples. Format the output as a Markdown file I can use to instruct an AI to write in my voice."
3
Review and refine — it should feel accurate when you read it

Add anything it missed. Remove anything wrong. Save as your-voice.md to a private location — not the shared skills folder.

4
Attach permanently to your private projects

Strategy, CFO Advisor — these always write in your voice. For shared projects (Presentation Builder, Content), upload it per session when you specifically need it.

Using your voice file in a shared project

Per-session upload is not just for sensitive financial data — it is also the right pattern for personal context that belongs to one person but occasionally needs to inform a shared project. Drag your-voice.md into the conversation at the start of that session. It is available for that conversation only. Other team members using the same project in their own sessions will not have it — which is correct. The file stays private; the output is yours.

Revisit every 6 months

Your voice evolves. A voice file written today will feel slightly off in 18 months. Schedule a 30-minute refresh — paste new writing samples, run the same prompt, compare to the existing file, update what's changed.

Section 06

Claude as your CFO advisor

Claude can be an exceptional financial thinking partner. The setup is simple. The one rule is that actual financial figures are never stored permanently in any project.

The risk is not Claude — it is the project files. A file permanently attached to a project is readable by anyone with access to that project. Keep the framework permanent. Keep the numbers per session. That is the entire security model.

Setting up your CFO Advisor project

1
Create a private project — not visible to the team

Do not add it to the Team workspace. It should appear under "Your projects" only. If a finance manager also needs access, they create their own identically-configured private project.

2
Write a strong system prompt

Something like: "You are an experienced CFO advisor. I will share financial data for analysis. Be direct. Identify risks, flag weak assumptions, and always clarify what additional information you would want before making a real recommendation."

3
Attach only cfo-framework.md as a permanent file

This contains your KPI definitions and measurement framework — no actual numbers. It makes Claude immediately familiar with how you run your business without any financial exposure.

4
Upload actual financial data fresh at the start of each session

Drop in the PDF or spreadsheet at the start of the conversation. When the session ends, it is gone from Claude's context. Never use permanent project attachments for live financials.

5
Save outputs to your restricted cloud folder

Financial analysis, scenario models, board summaries — save to your authorised-access-only cloud folder, not the general outputs folder the team can see.

What Claude can do for you financially

Analysis
Actuals vs budget review

Upload management accounts. Ask Claude to identify the top variances, likely causes, and questions to raise with your accountant before the next meeting.

Modelling
Scenario planning

"If revenue drops 20% next quarter, where do we have flexibility?" Claude runs the logic. You provide the numbers. Faster than a spreadsheet for structured thinking.

Communication
Board and bank summaries

Turn raw financial data into a clear narrative for a board update, bank review, or investor brief — in your voice, with the right level of detail for the audience.

Challenge
CFO-style critique

"Act as a CFO reviewing this plan. What are the three weakest assumptions?" Structured critical thinking on demand, before you commit to anything.

Claude is a thinking partner, not a regulated advisor

Claude's financial analysis is excellent for internal decision-making, stress-testing, and communication preparation. For decisions with legal or tax implications — acquisitions, pension changes, statutory filings — validate with your accountant or solicitor. Claude gets you 80% of the way there, faster. The professional provides the final 20% with legal accountability.

Section 07

Data security — the traffic light system

Three categories. Every team member learns them in five minutes. Most data security failures happen because people don't know the rules — not because they're careless.

Status Category What this means Examples
Safe Public or product knowledge Already public, or shareable with any customer. Safe as permanent project files in any shared project. Product specs, brand guidelines, website copy, FAQs
Safe Internal operational content Internal processes and templates with no personal data and no financial figures. Safe for shared projects. Sales scripts, process docs, proposal templates, onboarding checklists
Caution Anonymised business data Fine with names and personal details removed. Use a role-specific project, not a shared team project. Meeting notes (anonymised), market analysis, internal strategy drafts
Caution Customer account data Only in a role-specific or account-specific project. Never in a shared team project. Anonymise where possible. Account history, pricing agreements, customer correspondence
Never* Financial statements & figures Private project only. Uploaded per session. Never attached permanently. Outputs saved to authorised-access-only folder. P&L, cashflow, management accounts, payroll, customer-level margins
Never Personal data (GDPR / CCPA) Any data covered by privacy law. Without a Data Processing Agreement, using this in Claude is a compliance risk. Employee HR records, home addresses, personal emails, ID documents
Never Passwords & credentials No exceptions. No workarounds. Not even for "just a second." System passwords, API keys, banking logins, SSO tokens

* Financial figures are permitted in the private CFO Advisor project via per-session upload only — see Section 06.

Cloud storage permissions — configure these on day one

/skills/ — Read all, write owner
Skill files are your company's AI knowledge. Treat them like master brand documents. Everyone reads; only the authorised owner edits.
/projects/ — Role-based
Sub-folders by function: /sales/, /ops/, /support/. Each group sees only what they need. Not cross-visible between functions.
/outputs/ — All staff read/write
Reviewed and approved AI-generated content only. Version history is your free audit trail — never disable it.
/finance/ — Authorised only
Named individuals only. No general team access. The strictest folder in the structure. Adding access = a deliberate decision, not a default.
Section 08

Connectors — email & cloud storage

Claude can connect to your email and cloud storage. The capability is real. The behaviour is less obvious than most people expect. Read this before enabling anything.

When you enable a connector — Microsoft 365, Google Workspace, or similar — Claude authenticates using your credentials. It can only reach what you personally have permission to access. It does not bypass your cloud storage folder permissions. The connector sees what you see, nothing more.

The risk is not Claude breaking into locked files. The risk is what project you are in when the connector is active. Any content Claude surfaces from your email or cloud storage appears in the conversation — and all conversation history is visible to everyone with access to that project.

The implicit query problem — the thing most people miss

Four scenarios — connector active in the same project
"Check my emails for the last 3 months for information about [Supplier X]"
Explicit instruction. Claude uses the connector. You know what it will do before it does it. This is the safe pattern — always use explicit instructions when you want connector access.
~
"Search for information about [Supplier X]" — connector active
Ambiguous. Claude will likely use all available tools including email and cloud storage — because it is trying to be maximally helpful. It may not tell you it is checking your email. Results from skill files, cloud storage, and email threads could be blended in one response without clear attribution. You don't know what it accessed.
"Search for information about [Supplier X]" — no connector active
Claude works only with project knowledge files and session-uploaded documents. Predictable and bounded. This is the recommended default state for all shared projects.
Connector active in a different project
Connectors are scoped per project, not globally. If you enable a connector in your private Strategy project but not in your shared Customer Service project, asking a question in the Customer Service project will not trigger email access. The connector is on per project context — not everywhere simultaneously.

The fix — add this to any project where a connector is enabled

Add to project system prompt
"Do not use the email connector, cloud storage connector, or any external tool unless the user explicitly instructs you to in this conversation. Default to project knowledge files and session-uploaded documents only. When you do use a connector, state clearly which source you are drawing from before presenting results."

Connector rules for business owners

Private projects only — to start
Use connectors in your private projects first (Strategy, CFO Advisor). Results stay private because the project is private. Build comfort before considering shared project use.
Read-only only — never write
Enable read access only. Never enable "send on behalf" or write access via any connector. An accidental send from a Claude session is not recoverable.
Add the explicit-only instruction
Add the system prompt instruction above to any project where a connector is enabled. Changes Claude from "use all available tools when helpful" to "only use connectors when explicitly told to."
No connectors in shared projects yet
Until your team is comfortable with explicit-only usage and understands that connector results appear in shared conversation history, keep connectors out of shared projects entirely.
The interim workflow that works today

No connector needed. Open the relevant project. Copy the email thread or document you want Claude to work with. Paste it into the conversation. You control exactly what context goes in. Outputs save to cloud storage as normal. This is safer, more predictable, and available immediately with no configuration.

Section 09

Building agents

An agent is Claude taking a sequence of actions to complete a goal. Not answering a question — doing a job. Here is how to think about it and when to build one.

The difference between using Claude and building an agent: using Claude is asking a question and getting an answer. An agent is giving Claude a goal, tools, and letting it figure out the steps. Build your project and skill file foundation first. Agents are what you add on top once that foundation is solid.

The four levels — where are you now?

Four levels of Claude deployment
L1
Conversation — you ask, Claude answers
No persistent context. Starting from scratch every session. Where most businesses begin. Useful but not scalable — you're doing all the setup work every time.
L2
Projects + skill files — Claude always has the right context
Structured projects, knowledge files, consistent team usage. This guide gets you here. This is the right foundation for most SMBs — and the prerequisite for everything below.
L3
Simple agent — Claude handles a multi-step workflow
Claude processes an input, takes several steps, produces a structured output with minimal human involvement. Examples: classify an email then draft a reply. Summarise a meeting then extract action items with owners.
L4
Autonomous agent — Claude plans and executes complex tasks
Claude browses, writes, updates systems, and sends communications with human checkpoints at defined points. Requires careful design, clear guardrails, and proven L2–L3 foundations first.

Three realistic starter agents for business owners

Starter A
Meeting summariser

Paste a transcript → Claude extracts decisions, action items, owners, deadlines → structured summary ready to paste into your project management tool.

Starter B
Email handler

Paste a customer email → Claude classifies it → drafts an on-brand response → flags if it needs human review before sending. Review the draft, send or edit.

Starter C
Content repurposer

Paste a blog post → Claude produces a LinkedIn post, a 3-email newsletter sequence, and 5 social captions — all in your brand voice, in one pass.

Four rules before you build any agent

1
Define the goal precisely
"Handle customer emails" is not a goal. "Read the email, classify it as enquiry / complaint / order, draft a reply using our templates, flag anything mentioning a refund for human review" is a goal. Write it in full before building anything.
2
Define what it cannot do
Every agent needs explicit limits written before anything else. "Do not send without human approval." "Do not access files outside /outputs/." "Do not make promises about pricing." Write these first.
3
Test with fake data before going live
Build 10 test scenarios covering normal cases and edge cases. Run each one. When the agent gets something wrong, update the system prompt — not the test. The test revealed a gap; the prompt is where you fill it.
4
Go live with a human review gate
Every output reviewed before any action is taken. After 20 consistent passes you can consider removing specific review gates — one at a time. Not before.
Section 10

Team AI usage policy

One page. Everyone reads it and signs it. You review it every 90 days. This is what responsible AI adoption looks like for a small business.

Below is a template you can adapt. Replace [Company] with your name. The rules should work for most businesses without changes — but read them carefully and adjust anything that doesn't fit how you operate.


[Company] — AI Usage Policy

Version 1.0 · All staff · Review every 90 days

What Claude is at [Company]

Claude is a secure AI assistant used for writing, research, analysis, customer service, and business advisory. It is accessed via Claude Desktop on the Team plan. Our data is not used to train Claude's models. Conversations do not persist after a session ends.

Rules everyone must follow

Use the right project
Always start in the project designed for your task. Don't use a shared project for something that belongs in a role-specific or private one.
Review before sending
Any Claude output going to a customer, supplier, or published externally must be reviewed by a human before it leaves. No exceptions.
Save approved work to /outputs/
Approved Claude outputs are saved to the shared /outputs/ folder with a date-prefixed filename. This is your record and version history.
Reference cloud storage skill files
Private projects are fine. But they must reference the company skill files from cloud storage — not locally-saved copies. Local copies drift out of date silently.
Report mistakes immediately
If you accidentally share sensitive data in Claude, report it immediately. The Team plan's zero-retention policy means it won't persist — but we document it regardless.
No financial figures in shared projects
Revenue, margins, P&L, payroll, and bank data are never entered into any shared project. Private CFO Advisor only, per session only.
No GDPR / CCPA personal data
Customer personal data, employee HR records, home addresses — never entered into Claude without a Data Processing Agreement in place.
No passwords or credentials
Never paste a password, API key, system credential, or banking login into Claude. No exceptions, no workarounds.
No connectors in shared projects
Email and cloud storage connectors must not be enabled in shared team projects. Connector results appear in shared conversation history visible to all project members.
No personal accounts for company work
All company work is done through the company Team plan account. Personal Claude.ai accounts must not be used for company tasks.

I have read and understood this policy.    Name: _______________________    Date: ___________

Section 11

Beginner best practices

The habits that separate consistent results from inconsistent ones. Not advanced techniques — fundamentals most business owners skip in the first 90 days.

The four-part prompt structure

Most people include one or two of these. The difference in output quality when you include all four is immediate and significant.

Four-part prompt structure
R
Role
Tell Claude what it is in this conversation. "You are an experienced B2B sales writer." One line shifts tone, vocabulary, and approach dramatically. Not needed if your project system prompt already sets the role.
T
Task
What do you actually want? Specific beats vague every time. "Write a 200-word follow-up email" is far better than "write an email." Include format, length, and purpose explicitly.
C
Context
What does Claude need to know to do this well? The customer's situation, the previous conversation, the product being discussed, any constraints. Don't make Claude guess — it will, and it will sometimes guess wrong.
O
Output format
How should the result be structured? "Give me three options in bullet form." Or: "Write as flowing paragraphs — no headers." If you don't specify, Claude chooses — and it may not match what you had in mind.

Six habits that compound over time

Habit 1
Give examples, not just instructions

Show Claude what right looks like. Paste an example of a good email. Say "write it like this." Examples outperform descriptions every time.

Habit 2
Ask for multiple options

For creative or strategic work, ask for 3 versions. You will nearly always pick parts of multiple — but you make a better decision than from one option.

Habit 3
Iterate, don't restart

If the first output isn't right, tell Claude exactly what to change. "Make the opening more direct." Refine the conversation — don't start over.

Habit 4
Say "think step by step"

For complex decisions or analysis, add this phrase. It measurably improves output quality on difficult tasks by making Claude show its reasoning.

Habit 5
Verify facts independently

Claude can be confidently wrong about specific facts, dates, and technical details. Always verify anything you will publish or quote to a client.

Habit 6
Save your best prompts

When a prompt consistently produces excellent results, save it in your skill files or a prompt library document. Your best prompts are assets.

The five traps to avoid

Accepting the first output
Claude's first response is almost never its best. One refinement prompt routinely produces significantly better results. The first output is a draft, not a delivery.
Vague prompts
"Write me a good email about our product" produces average results. Specific context, length, audience, and purpose produce something you could nearly send as-is.
Trusting facts without checking
Use Claude for structure and language. Check facts yourself — especially anything you'll publish, quote to a client, or use to make a business decision.
Building agents before the foundation is solid
Get your projects and skill files working consistently for 60 days before building any agent. Automation built on a shaky foundation just makes mistakes faster.
Using Claude for everything
Claude is exceptional at language, analysis, and synthesis. It is not the right tool for real-time data, regulated advice, or decisions requiring legal accountability. Know the boundaries.
Section 12

Your 90-day plan

Don't try to do everything at once. This sequence gets you from zero to structured in three months without disrupting the day-to-day.

The businesses that stall on AI adoption usually make one of two mistakes: they do too much too fast and lose control, or they stay in experiment mode indefinitely and never build real systems. This sequence threads the needle.

Month 1 — Foundation
Get the basics right
✓ Confirm all users on Team plan
✓ Create cloud storage folder structure
✓ Set folder permissions correctly
✓ Build brand-voice.md
✓ Build product-kb.md
✓ Set up Tier 1 shared projects
✓ Set up private CFO Advisor project
✓ Build your personal voice file
✓ Share and sign the usage policy
✓ Run 30-min team walkthrough
Month 2 — Build
Add depth and coverage
✓ Write sales-playbook.md
✓ Write customer-service.md
✓ Test each skill file before deploying
✓ Set up role-specific projects
✓ Set up major account projects
✓ Establish /outputs/ review workflow
✓ Run first CFO Advisor session
✓ Collect team feedback
✓ Identify one workflow to automate
Month 3 — Review & Expand
Refine and build forward
✓ First quarterly policy review
✓ Refresh any stale skill files
✓ Design your first simple agent
✓ Test agent on 10 scenarios
✓ Go live with review gate on
✓ Review /outputs/ quality patterns
✓ Consider connector use in private projects
✓ Onboard any new team members
✓ Set date for Month 6 review

The order matters

Foundation before skills. Skills before agents. Agents before connectors. Every layer depends on the one below it being solid. An agent built on a poorly-configured project with vague skill files will produce inconsistent results at best. At worst, it will automate your chaos.

The businesses that get the most from Claude in year one are not the ones who moved fastest. They are the ones who built the foundation properly, let the team get comfortable, and then expanded deliberately. The compound effect of good skill files and well-configured projects grows quietly over time — and after six months the gap between structured and unstructured AI adoption inside a business is significant.

The single most important thing to do this week

Write your brand-voice.md file and circulate the team usage policy. Those two things — knowing how to sound and knowing the rules — immediately change how your team uses Claude. Everything else can wait a week. Do those two things first.


Share freely · Adapt for your business · No attribution required

Accurate as of April 2026. Check claude.ai for current plan details.

Claude for Business Owners

2026 Edition · Practical Setup Guide


If you found this guide useful, I'd really appreciate a follow.

I share practical AI and business content on both platforms.

Follow on LinkedIn Follow @xichong