Turning a discovery call into a scope and proposal used to take me four hours by hand. With the Claude prompt pack I built, the same job takes about 40 minutes, checks included.
I am a former marketing ops contractor in Columbus, Ohio, and that pack was the first of 13 AI digital products I built from a spare room.
You clicked a screenshot of a $20K-plus month from AI digital products. I kept that number in a spreadsheet with a column for money after fees, and month 14 read $20,563.36. Weeks 1-2 were $0.
The first sale was $59 in week 3, then month 3 closed at $438.24 after fees, month 6 at $4,120, month 9 at $11,860.
Fourteen months, 18 hours a week around a full-time job, one launch at a time. Pick one of the 13 from this list and by the end you will know its buyer, its price, its fee and the week it first sold.
Then start it this month on one or two hours of weekly upkeep.
The figures in this article are one illustrative example, not typical results.
| Product | Price | Monthly sales or subscribers | Monthly gross | Payment fees | After payment fees |
|---|---|---|---|---|---|
| 1. Claude Prompt Pack for Client Onboarding | $59 | 38 | $2,242 | $104.88 | $2,137.12 |
| 2. ChatGPT Prompt Pack for Real Estate Agents | $29 | 54 | $1,566 | $84.24 | $1,481.76 |
| 3. Midjourney Prompt Pack for Product Photography | $39 | 32 | $1,248 | $62.72 | $1,185.28 |
| 4. AI Content Calendar for E-commerce Brands | $49 | 27 | $1,323 | $63.72 | $1,259.28 |
| 5. AI Social-Media Audit Worksheet for Salons | $39 | 33 | $1,287 | $64.68 | $1,222.32 |
| 6. Custom GPT Bundle for Etsy Sellers | $97 | 24 | $2,328 | $102.72 | $2,225.28 |
| 7. AI Launch Toolkit for Course Creators | $197 | 15 | $2,955 | $124.20 | $2,830.80 |
| 8. AI Content Ops Notion Template | $39 | 36 | $1,404 | $70.56 | $1,333.44 |
| 9. Monthly Claude Prompt Pack for Recruiters | $19/mo | 58 | $1,102 | $67.28 | $1,034.72 |
| 10. Living AI Prompt Library for Wedding Planners | $197 | 10 | $1,970 | $82.80 | $1,887.20 |
| 11. AI for Bookkeepers Mini-Course | $129 | 22 | $2,838 | $122.32 | $2,715.68 |
| 12. AI Product-Photo Background Pack | $27 | 49 | $1,323 | $72.52 | $1,250.48 |
| 13. Podcast Show-Notes Prompt Pack | $19 | 0 | $0 | $0 | $0 |
1. Claude Prompt Pack for Client Onboarding
A discovery call ends and the real work starts: the notes nobody finished, the scope written from scratch, a proposal tweaked for the fifth time, and the kickoff email at 11 p.m. That was my old way, and it ate two evenings a client.
My new way was 42 Claude prompts and six editable proposal templates, sold at $59 to solo consultants and small agency owners. Each prompt does one job: discovery notes, scope, proposal, kickoff email, revisions. A one-page instruction sheet sits on top.
Two hours a week went to updates, and privacy came first. Neither I nor the file ever touched real client material: no client files, passwords, or medical or financial records.
Month 14: 38 sales, $2,242 gross, $104.88 in fees, $2,137.12 kept. That fee is 4% plus $0.40 on each sale, the whole cost line. The first $59 landed in week 3, after two weeks at $0.
💡 Tip: Write the 42 prompts out in a plain index and ship on a direct Dodo checkout page at $59. This is the product the last section sends a reader with a service buyer back to.
2. ChatGPT Prompt Pack for Real Estate Agents
Prompt to copy
You are a real estate agent. Write listing copy for a 3-bed, 2-bath home with a renovated kitchen. Three short paragraphs, warm but plain, no hype adjectives, and leave every price and contract term blank for me to fill.
That one prompt, plus 49 more for open-house follow-up, buyer FAQs and social captions, went into a $29 pack for busy agents. A spreadsheet index listed every prompt by task so an agent could find one in seconds. First sale came in week 4.
Month 14: 54 sales, $1,566 gross, $84.24 in fees, $1,481.76 kept. The $1.56 fee leaves $27.44 a sale. Fifty-four buyers at $27.44 earned $655.36 less than 38 at $56.24. Volume lost, and lower prices keep the lesser crowd.
No rates, contract terms or disclosure language went into any prompt. Every output is a draft the agent checks before a client sees it.
💡 Tip: Keep rates, contract terms and disclosures out of every prompt. Each output is a draft the agent checks before it reaches a client.
The first four products by the numbers
Month 14 ledger, after 4% + $0.40 payment fees
3. Midjourney Prompt Pack for Product Photography
A small seller needs a white background hero shot, a flat lay and a seasonal backdrop, and a phone photo against the spare-room wall is a poor start. Copy sells on saved typing, so an image pack has to sell on evidence.
The product was 60 Midjourney prompts for hero shots, flat lays, lifestyle scenes and seasonal backgrounds, plus a style guide with blanks for camera, light and aspect ratio.
Month 14 read 32 sales, $1,248 gross, $62.72 in fees, $1,185.28 kept, one hour a week. First sale in week 5.
The catch is that I am not a photographer. Every sample on that listing was my own phone shot in the spare room, run through a prompt. None copied a brand's product, logo or one photographer's exact style, and that limit showed in the samples.
The fix was three before-and-after prompt examples, sitting right in the listing. Sellers already had product photos and no studio. Thirty-two sales at $39 put this third from the bottom on gross, and the ones who bought knew exactly what they were getting.
4. AI Content Calendar for E-commerce Brands
A sequence they can follow without inventing the order.
Three prompt packs had shown that people pay for one job at a time, done well. The next request was for the whole month, mapped out in advance.
Step one: open a blank Notion page and build a table with 90 daily rows, one column for the channel and one for the post topic. Name it the 90-day calendar.
Step two: add a second page holding the 25 prompts, grouped under captions, product education and email, and paste the review checklist above them.
Step three: fill month one completely as the sample, so a buyer sees a finished week instead of an empty grid.
Step four: write the product page around a 1-3 person e-commerce team. Mine went live in week 6 and sold 27 copies at $49 in month 14. That is $1,323 gross, $63.72 in fees, $1,259.28 kept, for 2 hours a week.
The catch is that the calendar produces drafts, not facts. Every claim about a product has to be checked against the brand's own materials before anyone posts it.
That is why the review checklist is a working page of the product, not a line at the bottom.
It is also why 27 sales, the lowest count of my first four, still beat a $39 image pack on gross: the buyer paid for a sequence, not a list.
One filled example month is the whole sales argument. A new buyer copies week one in about 20 minutes, and 27 buyers at $49 out-earned 32 at $39. The order is the thing being sold.
💡 Tip: Sell it with one filled example month so a buyer copies week one in 20 minutes.
🛠️ How a $59 prompt pack goes from blank page to first sale
Pick one job
Choose a single workflow a named buyer already does by hand, like turning a discovery call into a scope.
Write the prompts
Draft 40-50 prompts in a plain document, each one specific to that job and that buyer.
Test on real input
Run every prompt on a real example and log what came back and what needed fixing.
Build the file
Lay out the prompts in an index, add templates and a one-page instruction sheet.
Set the checkout
Publish on a direct Dodo page at $59, with 4% + $0.40 taken per transaction.
Send it to five buyers
Message five people who do the job and ask one to test the pack for free.
Fix and relaunch
Use the first buyer's notes to sharpen the weakest prompts, then sell.
5. AI Social-Media Audit Worksheet for Salons
My listing promised no bookings and no income figure anywhere on the page, because a salon owner burned once stops reading. The design stayed plain enough to rerelease each quarter.
Free templates are everywhere; fillable audit pages and salon-specific prompts were the only things that sold. Two of three Columbus owners pointed straight at the audit page.
Thirty-three copies at $39 made $1,287 gross in month 14. Payment fees took $64.68, leaving $1,222.32, on one hour a week. First sale came in week 8, the slowest start to that point.
The ledger line: $39 price, $1.96 fee per sale, $37.04 kept, 33 sales, $1,222.32 net. The five-minute walkthrough video rides on top at no extra cost and gives a poster who gets no bookings a reason to pay.
💡 Tip: Package the worksheet with a five-minute walkthrough video at $39, aimed at salons that already post and get no bookings.
6. Custom GPT Bundle for Etsy Sellers
A custom GPT is a saved chatbot with fixed instructions, and Etsy sellers with 20 to 200 listings already gather where they shop for tools.
My bundle held five of them: listings, buyer questions, photo planning and a launch outline, plus a 10-page setup guide and sample inputs. The seller sees the offer before paying.
Prompt to copy
You are helping me write the instruction block for a custom GPT that writes Etsy listing titles, descriptions and tags for a seller with [20-200] listings in [category]. Ask me five questions about their products and buyer first. Then write the instruction block in under 400 words: the role, the listing structure to follow, three sample inputs with expected outputs, and a rule that bans invented materials, measurements or shipping times. Plain text only.
ChatGPT returned a usable instruction block in about 40 seconds; my review ran 15 minutes, mostly cutting invented material claims and adding a line that no output promises a ranking.
The catch: GPT behavior shifts and store rules change, so I retested all five before each sale. Every link is separate, and one free sample GPT is testable.
Twenty-four sales at $97 came to $2,328 gross, $102.72 in fees and $2,225.28 kept, on two hours a week. The $4.28 fee is the heaviest bite of any product under $100 here, and 24 sales still ranked third on net. First sale: week 10.
💡 PRO TIP
The fee that does not scale down
At $19 a sale, the fixed $0.40 takes a bigger percentage than at $197. Run the arithmetic before you price below $30: 58 monthly transactions added $23.20 in fixed fees alone.
7. AI Launch Toolkit for Course Creators
Fifteen buyers paid $197 for a launch kit and I recorded $2,955 on it in month 14. The $97 test had shown price barely slowed a sale. So I wanted to know how high a bundle could climb without a single recorded course behind it.
This is the AI Launch Toolkit for course creators: a 30-day launch calendar, 18 email prompts, a webinar outline, 12 social prompts and a checklist.
The buyer already has an audience and no launch system. First sale came in week 12, and two hours a week keeps it current.
The payment fee took $8.28 a sale, leaving $2,830.80 from $3,055.20, and a 4% + $0.40 deduction on each of 15 transactions is what that came to.
In my month-14 sheet that was the smallest paying count in the whole catalog and the highest net of all thirteen products.
The catch: the toolkit includes affiliate swipe files. An affiliate link is a tracked link that pays you a commission when someone buys through it, and every one of those, plus any sponsored recommendation, needs a clear FTC disclosure right next to it.
Mine sits inside the file, not in a page footer.
Ledger: 15 sales at $197, $124.20 in fees, $2,830.80 after fees, first sale week 12-15 buyers at $197 beat 54 buyers at $29. I sold it as a $197 implementation bundle, not a course, with one filled launch calendar pasted in as the sample.
💡 Tip: Sell it as a $197 implementation bundle, not a course. Put one filled launch calendar on the page as the sample.
8. AI Content Ops Notion Template
The Notion template made $1,333.44 after fees in month 14, from 36 buyers at $39. One hour a week. It does that because the buyer duplicates it once and then opens it every Monday morning to run their content, long after the receipt is forgotten.
The $197 toolkit that came before it only paid off for someone with a trusting audience already lined up.
This one steps back to $39 for the solo consultant or coach who wants the working system inside a tool they already live in, and it can sell without a list.
Inside is a content calendar, an idea bank, 24 prompts, a repurposing checklist and a weekly review page.
Setup is a four-minute video I recorded once, showing how to duplicate the dashboard into your own workspace.
I placed that video above the buy button and it still runs unchanged.
The catch is Notion's own plan limits, which change. The template sells as a duplicate link, and my page points buyers to Notion's current plan details instead of stating limits I would have to keep rechecking.
None of my pages quote a plan limit at all.
Free Notion templates sit beside it on the same shelf, and 36 people still paid.
What sold was the weekly review page and the 24 prompts wired together, a routine rather than a calendar, at $1.96 of fee per sale and first sale in week 7.
9. Monthly Claude Prompt Pack for Recruiters
Fifty-eight recruiters pay $19 a month for the prompt pack, which cleared $1,034.72 in month 14 after $67.28 in fees, and it first sold in week 9. A one-time sale happens once per customer.
This traded that ceiling for a monthly charge, and it traded certainty for churn: the share of subscribers who leave each month.
The work each month is 15 fresh Claude prompts for candidate outreach, job descriptions and interview notes, plus an archive, a short changelog and a request form. Writing a month's set by hand takes about 3 hours.
Claude drafts all 15 in under a minute, and my check on the outreach wording takes 50 minutes. Updates ship on a fixed date, or the subscription loses its reason to exist.
Fifty-eight is a month-14 snapshot, not a promise. The $0.40 fixed fee across 58 transactions is $23.20 of the $67.28, and the $1.16 fee is 6.1% of a $19 sale, the heaviest percentage in the catalog.
I launched with one free sample month so a recruiter could test the format. Two times: 3 hours by hand against about 5 minutes with Claude, plus 50 minutes of checks.
💡 Tip: Launch at $19 a month with one free sample month so recruiters can test the format before subscribing.
Product 7 at a Glance
💰 Price
$197
📦 Monthly sales
15
💵 Net after fees
$2,830.80
⏱️ Weekly hours
2
📅 First sale
Week 12
🧾 Fee per sale
$8.28
Why it stands out
It has the smallest paying count in the catalog and the highest net, because the $197 price carries the fixed fee.
10. Living AI Prompt Library for Wedding Planners
For me, yes, if the updates actually ship. A subscription dies quietly when a deadline slips.
A living library can rot the same way, but it has a year of runway written into the sale.
The buyer is a wedding planner with 5 to 30 events a year. For $197 she got 120 prompts for inquiry replies, timelines, vendor emails and social posts, plus a quarterly update file.
First sale came week 14, the longest wait in my catalog outside the course.
The catch: the living promise holds only if updates land on a fixed quarterly schedule. One missed quarter turns a $197 library into a $197 file, and refunds follow.
Month 14 read 10 sales, $1,970 gross, $82.80 in fees at 4% plus $0.40, and $1,887.20 after fees for one hour a week.
That is the best net per weekly hour in the ledger. I sell it as one payment with the first year of updates included, then offer updates as a separate paid list.
Static pack vs. living library
Static pack
- Refund trigger: prompt breaks after a tool update
- Fix each copy by hand, sale by sale
- No update hours, no update revenue
- Buyer keeps using last year's prompts
- Sold as a one-time file, no email list
Living library
- Refund trigger: one missed quarter
- Fix once, push the file to every buyer
- 4 update weeks a year, about 8 hours each
- Upsell list of buyers who renew
- First sale week 14, then steady
11. AI for Bookkeepers Mini-Course
The temptation with a course is to record everything you know and call it a curriculum. That is how a weekend turns into three months.
What worked was narrow: six lessons, each under 10 minutes, so buyers finish. The whole thing was recorded in the spare room over one weekend.
Before, a bookkeeper writing month-end client emails pulled each one from scratch. After, she opens the 18 prompts, picks the summary template, pastes the account names and edits. Same job, a fraction of the blank-page time.
One rule first. No fake testimonials and no AI-written reviews. The FTC rule in force since October 2024 bans fake reviews and testimonials, including AI-written ones, and buying or selling them. My page carries none.
The buyer is a bookkeeper or small accounting firm. Month 14 read 22 sales at $129, $2,838 gross, $122.32 in fees, $2,715.68 after fees, 2 hours a week.
First sale week 16, the latest in the catalog. I sell it with one free lesson, and the 18 prompts ride along as the workbook.
💡 Tip: Sell at $129 with one free lesson and the 18 prompts included as the workbook.
The five terms behind every ledger line
12. AI Product-Photo Background Pack
The tedious part is not the product photo. It is the background: a bedsheet taped to a wall, then an hour in an editor matching shadows.
I sold 40 AI-generated backgrounds in square and vertical crops to sellers who shoot on a phone, plus a five-step compositing guide. First sale came week 11.
By hand, one clean hero shot with a swapped background ran about 25 minutes. With the pack, the background is picked in 2, and the composite and fixes take 10. Check each image against the tool's current terms.
Two times: 25 minutes by hand against 12 with the pack, 10 of them mine.
$27 is the lowest price in the catalog, and the $1.48 fee is 5.5% of it, against 4.2% on the $197 toolkit.
Still, 49 sales made $1,323 gross and $1,250.48 after fees for 1 hour a week. Six free samples sit in the listing, and the pack sells as a low-risk add-on to the $39 Midjourney prompt pack.
13. Podcast Show-Notes Prompt Pack
A show-notes pack for podcasters was my thirteenth product: 25 prompts for episode titles, show notes, guest questions and clip captions, priced at $19.
I built it in week 17 and dropped it in week 18, and its ledger line has read $19 and 0 sales every month since.
The ledger: 6 build hours, 0 weekly hours, $0 gross, $0 fees, $0 after fees, and no first paid week ever. That $0 is not bad luck. It is the test doing its job.
I had no podcast audience to sell into, and the prompts were the kind any free list could replace. No named buyer with a channel, and nothing specific enough to make $19 worth it.
The failure cost 6 build hours and nothing else. The same week pointed at the $59 onboarding pack, already selling 38 copies a month, was the better trade, so the week went there instead.
Total: $0 earned against 6 hours spent, which is what a $19 test is for. A product with no named buyer and no distribution gets skipped.
That is the control case the other twelve answered: each one had a person to sell to and a reason the price held.
💡 Tip: No named buyer and no distribution channel means skip the build. Put the week into the $59 onboarding pack instead.
Frequently Asked Questions
Do I need to code to build any of these?
No. Every product here was built in Notion, a document, a PDF editor or a checkout page. If you can write a clear prompt and paste it into a template, you can build the first four.
How long before I see a first sale?
In my ledger the first sale came in week 3 for the $59 onboarding pack, and as late as week 16 for the mini-course. The average across the first ten products was around week 9.
What does the checkout actually take?
My direct checkout charges 4% plus $0.40 per transaction. A marketplace can take 20-40% per prompt sale, which is why I validated on cheap listings first and moved to direct checkout once a product proved itself.
Can I sell these on Etsy or Gumroad instead?
Yes, and it is a reasonable place to start. Etsy has 86-95 million active buyers, and Gumroad charges a flat 10% plus $0.50 per sale on top of processing. Use a marketplace until you have 20-30 sales, then build your own checkout.
Is my chatbot data safe to use for client work?
On free and personal plans, chats may be used to improve the model unless you switch that off in settings. Never paste passwords, client files, or medical or financial records into a chat window.
Should I expect $20,563.36 in my first month?
No. That figure is a month-14 portfolio total across 13 products, and it took 18 hours a week for 14 months to reach. Month 3 ended at $438.24 after fees, and most of the growth came after the sixth product.
Which of the 13 to build when you have a service buyer, a marketplace audience, or a skill to teach
If you have a service buyer, start with product 1, the $59 Claude prompt pack for client onboarding. Write 10 prompts for one workflow you already do by hand, test them on a real project, and publish on a direct checkout page.
If you have a marketplace audience, start with product 6, the $97 custom GPT bundle for Etsy sellers. Build one GPT, give it a free sample link, and retest it before every sale.
If you can teach a professional skill, start with product 11, the $129 bookkeeping mini-course. Record one lesson under 10 minutes, write the prompts as a workbook, and put a free lesson on the page.
The $20,563.36 is a month-14 portfolio total, not a launch-week promise. Week 1 brought $0 for me. Launch one product this month and let the ledger tell you what to fix.