August 17–24: ¥42,000/month in document work, $130 from a fox, and a $900 flop
A ranked weekly scan of 13 first-person side-hustle reports, from ¥42,000/month service work and a $130 Etsy fox to zero-sales tests and a nearly $900 ad loss.
The week in three numbers
From August 17 at 09:00 to August 24 at 09:00 (-05:00), the clearest reports came from small, specific offers rather than broad "passive income" bets. One operator reported ¥42,000/month for writing rental-management notices. One Etsy shop reported about $130/month from personalized fox storybooks. Another operator reported losing nearly $900 on ads before switching to technical documentation templates.
The figures below are self-reported. Sales, profit, platform payout, cumulative totals, seasonal totals, and losses stay separate. A report published this week can describe an older operation, so each entry says whether the result came from a fresh test or an existing business.
Quick ranking
The order uses barrier first, then the reported result. "Low" means a reader can test the offer with little cash and ordinary consumer tools. "Medium" means the reader needs a specific skill, client trust, or a substantial setup. "High" means the test can consume meaningful capital before demand is proven.
| Rank | Experiment | Reported result and horizon | Startup cash | Barrier |
|---|---|---|---|---|
| 1 | Personalized Etsy storybooks | About $130/month currently; close to $400 combined in November–December; about $14 after fees in month three 1 | $0.20 per Etsy listing; AI and PDF-tool costs not disclosed | Low to medium: templates help, but personalization and quality control remain manual |
| 2 | AI portfolio template | ¥0 sales in August; the note account had 1,488 views at the reported checkpoint 2 | Free test described; downstream tool costs not disclosed | Low: one existing work sample and 30 minutes |
| 3 | Plumbing-expertise AI publishing experiment | ¥0 note sales in week three, with 202 views, 28 likes, and 7 followers 3 | Not disclosed | Low: publishing and a narrow sales funnel |
| 4 | Food delivery shift | ¥0 from one hour and zero deliveries on August 20; this is a shift result, not monthly revenue 4 | Not disclosed | Low to medium: app access, local demand, and transport are required |
| 5 | Amazon KDP puzzle books | ¥0 royalties for the month, with measurement starting that week 5 | Not disclosed | Low cash, medium time: publishing a catalog is the work |
| 6 | Practical Excel templates | ¥0 sales for the month, with measurement starting that week 5 | Not disclosed | Low to medium: the file must solve a specific work problem |
| 7 | Redbubble print-on-demand | $14 in six months after more than 600 designs 6 | Not disclosed | Low cash, high time: the reported catalog took hundreds of designs |
| 8 | Rental-management AI response writing | ¥42,000/month from two local management companies; about 12 hours/month 7 | Not disclosed | Medium: client trust, tone control, and sensitive business text |
| 9 | Customer-list cleanup | ¥34,100/month from three clients as of July 2026; 9–11 hours/month 8 | Not disclosed | Medium: spreadsheet accuracy, confidentiality, and client access |
| 10 | Résumé editing on Coconala | ¥33,600 gross/month in June; ¥26,208 after Coconala's stated 22% fee 9 | About $20/month for the reported Claude plan | Medium: each order took about 53 minutes and requires careful handling of personal data |
| 11 | Technical documentation templates | $250–$400/month for the last quarter; the initial build took about 60 hours 6 | Not disclosed | Medium: the product depends on technical-writing knowledge and a niche marketplace |
| 12 | Yahoo Shopping home-appliance resale | ¥45,019 total profit for the month; the same report lists ¥93,860 confirmed gross profit for card resale 5 | Not disclosed | High: this was an existing resale operation with inventory and operating capital |
| 13 | Dropshipping | Nearly $900 lost on ads for an unwanted kitchen gadget; this is an ad loss, not revenue 6 | Ad spend is reflected in the loss; total startup cash not disclosed | High: paid acquisition came before demand was proven |
Low-cash experiments: a small sale still needs a mechanism
A recurring fox did more work than a catalog
The Etsy operator sells printable personalized storybooks. A buyer sends a child's name, and the operator swaps the name into an existing story, generates pages with a recurring fox character, and delivers a PDF. Canva supplies the page templates, APOB AI generates the fox artwork, and iLovePDF merges the pages.
The current result is about $130/month, mostly handled in a Sunday evening batch. The path to that result included zero sales in each of the first two months, about $14 after Etsy's cut in month three, and close to $400 across November and December. The holiday total is seasonal and cumulative; it is not a monthly run rate. Etsy listing fees are $0.20 per listing, while the other tool costs are not disclosed. 1
The operator started with six story themes and existing templates. A reader's first move is to make one reusable story, one consistent character, and one clear personalization field before creating a catalog. The recurring character created repeat demand, while group scenes and name-correction requests created manual rework.
The portfolio test separated attention from buyers
A truck driver learning AI and web marketing published a 30-minute template for describing one existing work sample in six lines: audience, problem, what was made, scope, and what was checked. The August result was zero buyers and ¥0 sales, even though the note account had 1,488 views and 289 likes at the reported checkpoint. The free test is described as free; related paid products cost ¥500 and ¥980, and broader tool costs are not disclosed. 2
The first move is concrete: open one finished piece from the day job or a personal project and write the six lines without calling it a client result. The evidence supports a better application description. It does not yet support paid demand.
The AI experiment changed its funnel after three weeks at zero
A plumbing business owner is testing note publishing, social media, plumbing expertise, and affiliate links over 90 days. At week three, the experiment had 202 views, 28 likes, 2 comments, 7 followers, and ¥0 note sales. The author planned to change the path from problem to cause to self-help solution, then an optional tool recommendation and a boundary for when professional help is needed. Work hours and cash cost are not disclosed. 3
A 9-to-5 reader can copy the test structure without copying the niche: publish one useful answer, attach one relevant paid action, and record sales separately from views and likes. The week-three result is a funnel test with zero sales, not proof that the topic has no demand.
Three more low-cash tests stayed at zero or near-zero
The food-delivery report records one hour with the app on, zero deliveries, and ¥0 sales. The platform name and monthly revenue are not disclosed. A reader who wants to test delivery should treat the first shift as a demand check in one defined area and record the hour, orders, and gross payout separately. 4
The same zero-result discipline appears in the home-appliance seller's new Amazon KDP puzzle-book and practical Excel-template tests. Both showed ¥0 for the month, and both began measurement that week. The source does not disclose startup cash, hours, or the number of products listed. The first move for either test is one narrowly defined product and a dated sales log, rather than a large catalog built on assumed demand. 5
Redbubble produced a different warning. A technical writer uploaded more than 600 designs and made $14 over six months. The cash cost is not disclosed, but the time cost was large. A reader should set a design-count and time limit before testing print-on-demand; a six-month cumulative figure should never be presented as monthly income. 6
Skill and service work: the payment came from a narrow problem
Rental-management notices turned a familiar writing skill into a retainer
The operator is a Japanese retail-store manager who uses Claude to draft tenant-facing rental-management notices and then makes the human edits. Two local management companies pay ¥42,000/month in total. One client pays ¥28,000/month for roughly 30 notices plus individual responses; the other pays ¥14,000/month. The work takes about 12 hours/month. Cash startup cost is not disclosed. 7
The report was published on August 23, but the service is an existing operation: the first conversation happened in October 2025, ten free sample notices opened the first contract, and a referral brought the second client. A reader's first move is to choose one local business with repetitive customer messages and offer a small, reviewable sample. The public page is partly paywalled, so the full prompt set and workflow are not available for reproduction.
Customer-list cleanup sold accuracy and confidentiality
The same operator uses Claude and Google Sheets to match customer names, clean addresses, and consolidate records. Three clients produced ¥34,100/month as of July 2026, with 9–11 hours of work per month. The public breakdown is ¥15,600, ¥12,000, and ¥6,500; one client has about 4,800 records. Cash startup cost is not disclosed. 8
The service began with an ¥8,000 spot project in September 2025 and became recurring work. The operator found clients through a local merchant association, CrowdWorks, and a referral. A reader's first move is to offer a fixed-size cleanup sample using fictional or consented data, then write down the handling and deletion rules before touching client records. The six-step workflow and parts of the NDA and deletion procedure sit behind a ¥498 paywall.
Résumé editing reached 12 orders, with the platform fee visible
The operator listed résumé editing on Coconala in February 2026 and received the first order on day 17. By June, the service reached 12 orders at ¥2,800 each, or ¥33,600 gross for the month. Coconala's stated 22% fee left ¥26,208 before other expenses. Each order took about 53 minutes. The reported Claude plan cost roughly $20/month. 9
The operator had no formal HR background and handled the work at night after managing a retail store. A reader's first move is to define a narrow editing promise, create one anonymized before-and-after sample, and state exactly what the service will review. The detailed workflow and personal-data rules are partly paywalled.
Existing operations and the cost of shortcut thinking
Technical documentation templates beat 600 generic designs
A technical writer spent about 60 hours creating specialized documentation templates for small engineering firms and listing them on a niche marketplace. The templates brought in $250–$400/month for the last quarter without promotion or ads. The marketplace and cash startup cost are not disclosed. 6
This is an existing product built from a professional skill, not a one-week zero-to-one result. A reader's first move is to turn one repeated workplace document into a narrow template and ask three people who use that document what would make it worth paying for. The same operator's Redbubble result shows why the time spent making a product matters as much as the platform's entry fee.
Resale profit came with inventory and a prior operation
The Yahoo Shopping home-appliance resale operation reported ¥45,019 total profit for the month, up ¥20,414 from the previous week. The same report separately listed ¥93,860 confirmed gross profit for card resale for the month. The figures are from an established operation using daily profit-and-loss monitoring; startup cash and weekly hours are not disclosed. 5
A reader should treat this as a mature-asset benchmark, not a promise for a first week. The first move is to record purchase cost, fees, shipping, returns, and realized profit for a small batch before buying more inventory.
The $900 loss arrived before product-market proof
The technical writer also tried dropshipping and lost nearly $900 on ads for a kitchen gadget that attracted no meaningful demand. The post gives no full startup-cost total, hours, or sales figure. The loss is a warning about sequence: paid acquisition can magnify a weak product before a seller has evidence that anyone wants it. 6
A safer first move is a demand check that costs almost nothing: describe the product to a small, relevant audience, collect concrete purchase intent, and set a hard ad ceiling before spending. The experiment still needs a real offer and a real customer; a view or click is not a sale.
Four tests that fit around a full-time job
- Choose one narrow deliverable. Write one type of notice, clean one kind of spreadsheet, edit one kind of résumé, or personalize one story format.
- Set a time ceiling. The reports include a 30-minute portfolio test, 53-minute résumé orders, 9–11 hours/month for list cleanup, and a 60-hour initial template build. Put the ceiling beside the revenue target before starting.
- Ask for demand before scaling production. The strongest service reports began with a free sample, a spot project, a marketplace listing, or a referral. The weakest reports spent months on catalogs or ads before a buyer appeared.
- Track money separately from attention. Views, likes, followers, downloads, and signups can rise while sales stay at zero. Record gross sales, platform fees, realized profit, cash costs, and the exact time horizon in separate columns.
The next weekly issue will keep the same test: a reported number, a stated horizon, a visible starting condition, and a first move that can survive a 9-to-5 schedule. These are public first-person reports, not a worldwide census.
References
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
Side Hustle Lab
Weekly digest of side-hustlers' new experiments and flops worldwide, ranked by barrier-to-entry and payout, with startup costs and real monthly revenue
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.
Related content
- Sign in to comment.