◆ The Shop · Syllabus · Updated September 2026
Claude course syllabus: Claude for Managers and Owners
In short: The Shop is 8 lessons and 48 chapters, 234 minutes of reading in total. Lessons unlock in order and each one ends with a saved Work Order. Lesson 1 is free to read. The course is $397 once with lifetime updates and a 30-day money-back guarantee.
Lesson 01 · Free
Lesson 1: Drop in: what you're building
What you'll have at the end and how long it really takes, the one-workflow rule, and a written goal with a name and a date on it.
- Time
- 3 chapters · 13 min
- You build
- Your goal and the one workflow you'll automate first
- Claude plan
- Free Claude plan works
-
01
What you'll walk out with 5 min
This course is about four hours of reading. The biggest piece of doing is a six-hour Business Brain capture across Lessons 3 and 4, and then one workflow runs for a week in Lesson 5. At the end you will have a Business Brain document, a scored list of the workflows in your business worth automating, an honest readiness score with named gaps, one workflow running as draft-and-approve, and a 30-day plan you could hand to someone on your team Monday morning. Setup takes five minutes and costs nothing.
-
02
One workflow. Not four. 4 min
You are going to automate one workflow. Not four, not "let's see what it can do." One, running properly and boring, for thirty days. Most people pick their first workflow wrong, and they pick it wrong in a predictable direction: they choose whatever annoys them most. Irritation isn't volume. This chapter gives you the rule and a provisional pick. Lesson 2 gives you the scoring method to confirm it.
-
03
Put a name and a date on it 4 min
Before you build anything, write one sentence that says what "this worked" looks like in numbers, with a name and a date in it. Nobody skips this on purpose. Everybody skips it. Without it, day 30 becomes a vibe check, and vibe checks always pass. This chapter ends with a short template you fill in, and that filled-in template is your deliverable for this lesson.
Lesson 02
Lesson 2: What Claude can really do for your business (and what it can't)
What AI genuinely does and doesn't do, how to pick the first workflow by a number, what it costs, and an honest 30-question readiness score with named gaps.
- Time
- 6 chapters · 30 min
- You build
- Your automate-first shortlist and readiness score
- Claude plan
- Free Claude plan works
-
01
What Claude is great at, and where it'll burn you 5 min
AI drafts. A named human approves. Nothing reaches a customer without a yes. That sentence is the whole system, and this chapter is the honest half of it: what Claude genuinely does well for a business, or the department you run, and the list of things it will not do for you no matter what you paid.
-
02
The chore to hand off first (it's not the one you think) 5 min
Score every recurring chore on three scales, hours, repetition, and how teachable the rules are, then multiply. Pick the single highest number and leave everything else alone for 30 days. The workflow that annoys you most is almost never the workflow that costs you most, and the scoring exists to stop the felt problem from beating the measured one.
-
03
The real price tag, including the part nobody mentions 5 min
Doing this properly yourself costs roughly $50 to $125 a month in Claude seats, one six-hour context capture, and 20 minutes a week. Every path that looks cheaper is charging you in hours or in cleanup instead, and it's worth knowing which currency you're paying in.
-
04
The 15-minute readiness check 5 min
Thirty yes/no questions across five categories, one point per yes, plus a one-week time audit no checklist can do for you. The rule that makes it worth doing: if it isn't written down or you can't point to it, it's a no. Fifteen minutes. Score the way things actually are this week, not the way they'll be once you finally get to that project.
-
05
What your score is really telling you 5 min
Your total maps to a tier, and the tier's only job is to filter advice. The five subtotals matter more than the total. Find your lowest category, and that's your first move. Two categories low? Workflows goes first, because a documented workflow makes every other category cheaper to fix.
-
06
The three gaps nearly everyone has 5 min
Three gaps show up in nearly every business and every department that takes this assessment, regardless of size, industry, or how sharp the person running it is. Nothing is written down. Everyone's using a different tool, alone. Nobody owns it and nothing has a date. None of them cost money to fix. All three cost about an afternoon.
Lesson 03
Lesson 3: The Business Brain: teach Claude to sound like you
Why generic output is a missing-context problem, the five sections that fix it, how to capture your voice in ten minutes, where the document lives, and the skeleton to build it now.
- Time
- 6 chapters · 34 min
- You build
- Your Business Brain document
- Claude plan
- Free Claude plan works
-
01
Why Claude sounds generic (it's not your prompt) 6 min
Generic output is not a bug in Claude and not a problem you solve with a better prompt. Anthropic built the engine and left the part that knows your business, or the department you run, to you. That part is a document I call the Business Brain. Capture it once, and every workflow you ever build reads it first. Skip it, and none of the workflows will sound like you.
-
02
Five sections that make Claude sound like you 5 min
The Brain has five sections, and the order below is roughly the order they pay off. Voice: a real sample, not adjectives. What you sell: real numbers, including what's not included. Who you serve, and who you turn away. How work arrives: your sales motion, or your intake process if you run a department. Hard rules: draft only, never invent, flag what you don't know.
-
03
Steal your own voice in 10 minutes 6 min
Adjectives ask the model to interpret. Samples give it something to imitate. Interpretation produces the average of the internet. Imitation produces you. Find one real piece of writing that passes three questions, fill in a short template around it, and test it with five prompts. Ninety minutes, most of which is finding the sample.
-
04
Give it a home Claude reads every time 5 min
A Brain that lives in your head or in six Slack messages doesn't exist. It has to sit somewhere Claude reads automatically at the start of every conversation. Three places do that: the instructions on a Claude Project, a CLAUDE.md file at the root of a Claude Cowork workspace, or the same file at the root of a Claude Code project. Pick one, pin it, and adopt the one rule that keeps it honest: any change to a price, policy, or offer updates the Brain the same day.
-
05
Running a team? Here's your version 6 min
If you run a department instead of a business, the Brain is the same five sections scoped to your team. Voice becomes how your team writes to the people it serves. What you sell becomes what your team delivers. Who you serve becomes your internal customers and the requests you push back on. How work arrives becomes intake. Hard rules don't change at all. Same six hours, same payoff, and often an easier capture, because a department's menu is shorter than a business's.
-
06
Bench time: build your Brain now 6 min
Here is the skeleton. Copy it into the home you picked in chapter four, replace every bracket, and run one test prompt. A rough first pass takes 20 to 30 minutes and a rough first pass is fine. You'll improve it every week for the rest of the year anyway. Don't move to Lesson 4 until the output sounds like someone who works for you.
Lesson 04
Lesson 4: What you sell, who it's for, and how work lands
Every offer in a six-field block, one paragraph on who you serve and one on who you don't, a one-sentence positioning, and your sales motion or intake process written down.
- Time
- 6 chapters · 33 min
- You build
- Structured offer, one-sentence positioning, and your sales-motion capture
- Claude plan
- Free Claude plan works
-
01
What you sell, in words Claude can't get wrong 5 min
Offers, ICP, positioning, and sales motion travel together, and they get captured in that order. Offers first because they're the most concrete thing in your business: you either sell it at that price or you don't. Start by listing everything you sell, or everything your team delivers. The list is always longer than you think, and an incomplete menu doesn't make Claude cautious. It makes Claude wrong.
-
02
The six-field offer block that kills vague answers 6 min
For each offer, capture six fields in the same format every time: price, what's included, what's not included, who buys it, timeline, and notes. The "what's not included" field is the one most people leave blank and the most important field in the layer, because language models err generous. Then add the à la carte items and a short block of rules on how you present prices.
-
03
Who it's for, and who you turn away 5 min
One paragraph on who you serve, with four things in it: the revenue band, the industry, the role of the person who signs, and the problem they're solving when they find you. Then one paragraph on who you don't serve, and why. The second paragraph is often the most useful field in the entire Brain, because most of the waste in a pipeline isn't lost deals. It's time spent on deals that were never real.
-
04
Your whole pitch in one sentence 5 min
Compress your offers and your customer into one sentence: for [target customer], we are the [category] that [unique value], unlike [obvious alternative]. The last clause does the real work. Your buyer's obvious alternative is almost never a direct competitor. It's hiring in-house, the expensive firm, duct-taping it together, or doing nothing for another year. The sentence isn't finished until the alternative in it is the one your prospects genuinely weigh you against.
-
05
How work lands: write down the call you already run 7 min
Write down the conversation you already run. For an owner, that's the discovery call in order, one real proposal skeleton, the Day 1, 3, 7 follow-up cadence, and the two or three objection responses that have actually worked. For a manager, it's the intake questions in order, the acknowledgment, the status rhythm, and the pushbacks that got a yes. Same structure, and the same side effect: writing it down reveals the hole in it.
-
06
Load it in and pressure-test it 5 min
Add everything from this lesson to your Brain under the three headers you created in Lesson 3. Then test it with five prompts that touch money, fit, and follow-up, and fix the capture, not the output, until they pass. When you're done, the Brain is v1 complete. Everything Claude drafts from this point forward starts with your voice, your menu, your buyer, and your motion already loaded.
Lesson 05
Lesson 5: Your first run: one workflow, one week
Pick one workflow by the numbers, build it as draft-and-approve inside a Claude Project loaded with your Business Brain, and run it for a week with one honest metric.
- Time
- 7 chapters · 29 min
- You build
- One working draft-and-approve workflow, run for a week
- Claude plan
- Free Claude plan works
-
01
Claude drafts, you approve: the only safe first run 4 min
Claude drafts, a named human approves, and nothing reaches a customer without that yes. It is the only safe first step because a wrong draft costs you nothing, and it is the only setup that gets better every week instead of quietly worse.
-
02
Pick your line: the one workflow worth building 4 min
Score every recurring chore on volume, repetition, and low judgment, multiply the three, and pick the single highest number above 60. Leave everything else alone for 30 days. The task that annoys you most is almost never the task that costs you most.
-
03
Five minutes to your first draft 4 min
Open Claude, create a Project named after your business or department, paste your Business Brain into the project instructions, and run one test prompt. Money doesn't enter the picture today.
-
04
The five-part prompt that runs the whole thing 4 min
A workflow prompt has five parts: when to run it, what you're giving it, what it should produce, the rules, and the output format. The FLAG rule is what makes it safe. Save it somewhere findable or it dies from friction.
-
05
One week, one number 4 min
One number for the next 30 days: drafts approved per week. Track it with tally marks, not a dashboard. When the output is wrong, fix the Brain, not the prompt. Ten minutes every Friday.
-
06
How to tell it's actually working 4 min
Working means the approved count goes up, the heavy-edit rate goes down, and the person running the process can name what they do less of. At day 30 you make one of three calls: ship it, fix one thing, or kill it cleanly.
-
07
The manager's version: same pattern, your team 5 min
The pattern is identical for a team. Only the customer changes. Two workflows score highest for managers: meeting notes to decisions, and the weekly report. Build one of them exactly the way you built the customer-facing one.
Lesson 06
Lesson 6: Wire it in: connectors and the four skills every business needs
Wire Claude into the five tools your work actually runs on, learn to fix a connector when it breaks, and install the four skills every business and every department builds first.
- Time
- 7 chapters · 33 min
- You build
- Five connectors live and four skills installed
- Claude plan
- Uses Claude Team
-
01
Connectors: the five that matter 4 min
A connector is an authorized link between Claude and one of your tools, so it reads real data instead of guessing. Five matter: your office suite, Slack, your CRM, your books, and one that fits your business. Whoever connects each tool decides what Claude can see.
-
02
Hook them up in the right order 4 min
Get on the Team plan, turn on the small-business feature in workspace settings, connect the five tools in order as the owner, and run one test prompt per tool. Don't move on until all five tests pass against reality.
-
03
When a connector breaks: the fix that works 90% of the time 4 min
About 90% of connector problems are one of four things, and the fix sequence is the same regardless of tool: reauthorize, check who connected it, check plan tiers on both sides, test with the smallest possible ask. And when your tool isn't on the list, route its data into something Claude already reads.
-
04
Skills: the SOP you never wrote 5 min
A skill is a saved, reusable instruction set for one repeatable task. It's the standard operating procedure you never wrote down, finally written down, with "read the Business Brain first" as line one. Build one for anything you do three or more times a week with a clear pattern.
-
05
The four skills every business ends up building 6 min
Four skills are universal: a weekly report, a customer service drafter, meeting notes to action items, and a content repurposer. Every business I've installed Claude into uses all four within the first week. Together they recover three to five hours a week.
-
06
Install your first skill: the weekly report 5 min
Install the weekly report first. It's the lowest-stakes skill, it pulls from three connectors at once, and its output lands in an internal channel where mistakes cost nothing. Save it by name, schedule it for Monday at 7 AM, and run all four skills for a full week before building anything else.
-
07
Same four, built for your department 5 min
The department's four are the same four with different nouns: a status report instead of the CEO report, meeting notes to decisions, a delegation brief, and an onboarding doc. Every one reads the team's Brain first. Everything that leaves the team stops at a draft.
Lesson 07
Lesson 7: Pick your setup: B2B growth, DTC marketing, or Team
Choose the one pack that matches how your work makes money or gets measured, build its skills as a chain where each output feeds the next, and measure it with arithmetic instead of vibes.
- Time
- 7 chapters · 36 min
- You build
- The pack that matches your work, installed
- Claude plan
- Uses Claude Team
-
01
Which pack fits how you make money 4 min
Past this point, what you build depends entirely on how your work makes money or gets measured. Sell services, build the B2B growth pack. Sell products to consumers, build the DTC marketing pack. Run a department, build the Team pack. Pick one. Skip the other two.
-
02
B2B: discovery call to same-day follow-up 6 min
The first three B2B skills turn a discovery call into branded notes, the notes into a same-day follow-up, and the notes plus your offer menu into a first-draft proposal. Each output is the next skill's input, and every one stops at a draft.
-
03
B2B: contracts and invoices without the retyping 5 min
The last two B2B skills fill your attorney-approved contract templates from the accepted proposal and draft the first invoice in QuickBooks from the signed SOW. Claude fills fields. It never writes legal language, never sends an envelope, never sends an invoice.
-
04
DTC: an analyst for your ad and email numbers 6 min
The DTC pack is an analyst, not a pipeline. The first two skills read your weekly ad and email exports and tell you plainly what's winning, what's bleeding, and what to do about it. One rule above all others: Claude reads and recommends. It never spends.
-
05
DTC: product copy, and ads that match your stock 6 min
The last two DTC skills write on-brand product copy from your product data and cross-reference inventory against ad spend and the email calendar to catch stockouts and dead weight. Both are read-and-recommend. No connection with write access to prices, discounts, or purchase orders, ever.
-
06
The Team pack: meeting prep to onboarding 5 min
The Team pack is five skills that chain a piece of work from meeting prep to onboarding: meeting-prep brief, meeting notes to decisions, delegation brief, the weekly numbers digest, and the onboarding doc. Internal outputs can post on their own. Anything that leaves the team stops at a draft.
-
07
Do the math: is it paying for itself? 4 min
Measure the hours before you automate, then run the winner through one formula: hours per week times loaded rate times 4.3. Against a real high-volume workflow, the math is boring-good. The expensive mistake is never the tool. It's the quarter burned automating the wrong thing.
Lesson 08
Lesson 8: Keep it running: trust rules, tune-ups, and your year-one plan
Learn the seven ways these projects die and the permanent boundaries that keep yours safe, pin the Trust Rules to your Brain, set the maintenance calendar, and write the 30/60/90 that turns a setup weekend into infrastructure.
- Time
- 6 chapters · 26 min
- You build
- Trust Rules doc and your 30/60/90 plan
- Claude plan
- Uses Claude Team
-
01
The seven ways this goes wrong (none are the tech) 5 min
AI projects in small businesses fail in seven identical ways, and not one of them is a technology failure. Every one is a sequencing, context, governance, or ownership problem, which means every one has a two-week fix. Watch week three hardest.
-
02
Where Claude should never have the last word 4 min
AI can draft anything. It should not release anything where the cost of being wrong lands on someone who didn't consent to the risk: your customer, your employee, or a regulator. When you hit a case this list doesn't cover, ask who eats the mistake and whether you'd say out loud that a machine decided it.
-
03
Trust Rules: one sentence that keeps you safe 4 min
The Trust Rules are one sentence and its corollaries: Claude drafts, a human approves, and nothing reaches a customer without a named person clicking send. Paste the doc into your Brain today, then spend fifteen minutes teaching your team two things: whoever sends it owns it, and fix the Brain, not just the draft.
-
04
The monthly tune-up: 20 minutes 4 min
Your Brain was 80% right the day you wrote it and decays from there. Twenty minutes on the first Monday of every month keeps the whole system honest: fifteen on the Brain, five on subscriptions. Once a quarter, prune skills before you plant new ones.
-
05
Your first 30 days, step by step 5 min
One workflow, one owner, one weekly appointment, one number, thirty days, then a real decision. The plan isn't complicated. The discipline of doing only one thing is the entire difficulty.
-
06
Your year-one map and your 30/60/90 4 min
Year one is front-loaded setup, then long stretches of small adjustments. The businesses that fail treat this as a project with an end date. The ones that win treat it as infrastructure with a maintenance schedule. Write your 30/60/90 before you close this lesson.
Syllabus FAQ
- How long is each lesson in The Shop?
- Each lesson is 13 to 36 minutes of reading, and chapters run 4 to 7 minutes. The whole course is 234 minutes across 48 chapters, about 4 hours. The biggest piece of doing is separate from the reading: a six-hour Business Brain capture across Lessons 3 and 4, then one workflow run for a week in Lesson 5.
- Do the lessons have to be done in order?
- Yes. Each lesson unlocks when the one before it is finished, because each builds on the last: the Business Brain from Lesson 3 feeds the draft-and-approve workflow in Lesson 5, and Lesson 6 connects that setup to your tools.
- Which lessons need a paid Claude plan?
- Lessons 1 to 5 work on Claude's free plan, which allows up to five Projects. From Lesson 6 on, the setup uses the Claude Team plan for shared connectors and Cowork: $25 per seat per month billed monthly or $20 billed annually, for teams of 2 to 150. Prices as of September 2026.
- What do you build in each lesson?
- Lesson 1: Your goal and the one workflow you'll automate first. Lesson 2: Your automate-first shortlist and readiness score. Lesson 3: Your Business Brain document. Lesson 4: Structured offer, one-sentence positioning, and your sales-motion capture. Lesson 5: One working draft-and-approve workflow, run for a week. Lesson 6: Five connectors live and four skills installed. Lesson 7: The pack that matches your work, installed. Lesson 8: Trust Rules doc and your 30/60/90 plan.
- Is any of The Shop free?
- Yes. Lesson 1, Drop in: what you're building, is free to read with no sign-in: 3 chapters, about 13 minutes. The rest of the course is $397 once with lifetime updates and a 30-day money-back guarantee.