Digital business field study · 2026
Starter Story online business study
We examined 349 public projects to see which online businesses report higher revenue and which are more practical for small teams using AI.
The report compares revenue and business models across 349 public projects. Three named cases explain how revenue can become profit; they do not represent the profitability of the full group.
- Background
- Starter Story is centered on businesses that report revenue. It can show visible business forms, but it does not represent all startup attempts.
- Purpose
- Compare reported revenue across the 349 projects, determine what the public evidence can establish about profit, and identify which business forms suit online delivery and AI-assisted development.
- Main sources
- Starter Story's public project database and category statistics, founder interviews, and current product sites.
- Project sample
- 349 public project profiles
- Reported monthly revenue
- $30K median among projects publishing an amount
- Population profitability
- Unknown · no consistent cost or net-profit data
Brief mode: the main findings, key figures, and open questions stay visible while extended tables and sources are condensed.
Key findings
The 349 projects describe revenue within this public set. Profit cases explain operating mechanisms only; they do not represent the wider population.
At least one-quarter of the projects report $100K or more per month
At least 86 of 349 projects (24.6%) report $100K+/month, and 151 (43.3%) report $30K+. Projects without a comparable public amount remain unknown, so these are conservative lower bounds across the full sample.
Projects publishing an amount have a $30K monthly-revenue median
The middle half report $9K–$113.5K per month, a spread of more than twelvefold. Starter Story already centers on revenue-generating businesses, so the figures describe differences among success cases, not expected income for a new project.
The public evidence cannot say how many of the 349 projects are profitable
The project database does not consistently disclose costs, net profit, or founder labor. It therefore cannot support a profitable-project count, average profit, or population margin. Named disclosures are used only as illustrations.
Of 349 projects, 128 (36.7%) belong to more than one category. Category counts should not be added together; category medians are better used to compare revenue scale than to estimate market share.
Business-model medians range from $20K to $200K a month
Higher reported revenue does not mean easier delivery: marketplaces and APIs also carry operating, reliability, and trust requirements.
At least 86 of the 349 projects report $100K a month
The shares below cover all 349 projects. Projects without a comparable amount remain unknown but stay in the total, making each result a conservative lower bound.
“At least” means projects without a comparable public amount are neither treated as zero nor used to inflate the rate. Because Starter Story centers on revenue-generating projects, these are not startup success rates.
The middle half spans more than twelvefold
The 25th and 75th percentiles are $9K and $113.5K/month. The wide spread makes a single mean or headline case especially misleading.
Micro-SaaS has twice the category median of simple apps
The platform reports a $40K monthly median for micro-SaaS and $20K for simple apps. A simpler build does not automatically lead to more revenue.
Marketplaces report more revenue and carry heavier operations
The marketplace category median is $200K/month, above APIs at $80K and micro-SaaS at $40K, but marketplaces must also manage two-sided liquidity, trust, disputes, and operations.
Micro-SaaS, productized services, and digital products are the three broadest product-form labels
Starter Story currently displays 2,997 revenue-generating projects. The micro-SaaS label covers 670 projects (22.4%), productized services 549 (18.3%), and digital products 455 (15.2%).
Each percentage is the category count divided by 2,997. One project may carry several labels, so the values cannot be added and should not be drawn as a pie.
Among online-delivery forms, APIs have the highest category median
Starter Story reports category medians of $80K for APIs, $40K for micro-SaaS, $30K for GPT apps, and $20K for simple apps. These categories also carry different reliability, security, platform-management, and customer-support requirements.
The platform defines these overlapping categories. The figures compare revenue scale; they do not show which type is easier to make successful or more profitable.
| Product-form label | Projects | Share of 2,997 projects | Reported monthly-revenue median | Plain-language meaning |
|---|---|---|---|---|
| Micro-SaaS | 670 | 22.4% | $40K | Subscription software for a narrow need |
| Productized services | 549 | 18.3% | $40K | Services sold with a fixed scope, price, and process |
| Digital products | 455 | 15.2% | $20K | Reusable content such as templates, ebooks, and courses |
| Marketplaces | 286 | 9.5% | $200K | Platforms matching buyers and sellers and charging for access or transactions |
| Niche blogs | 254 | 8.5% | $10K | Focused content monetized through ads, membership, or affiliates |
| Simple apps | 241 | 8.0% | $20K | Small software products that solve a few repeated tasks |
| GPT apps | 149 | 5.0% | $30K | Product-specific workflows and interfaces built on language models |
| Consumer iOS apps | 79 | 2.6% | $40K | Paid or subscription apps distributed through the App Store |
| APIs | 54 | 1.8% | $80K | Capabilities sold as interfaces that other software calls |
| Plugins | 19 | 0.6% | $30K | Extensions that run inside host platforms such as Airtable or Shopify |
These overlapping labels show product-form prevalence and reported revenue scale in the platform database. The percentages do not add to 100%, and revenue medians are neither profit nor success probability.
Revenue is not profit
The 349-project database has no consistent profit data. The named cases below show possible margin ranges and the costs that can absorb revenue.
- 85% stated margin
- Data Fetcher · $23K monthly revenue
- 60–70% stated margin
- Supergrow · later above $19K/month
- 26.3% margin from reported figures
- ChartDetector · about $11.5K April profit
Three cases with explicit profit figures range from about 26% to 85% margins
| Case | Reported monthly revenue | Profit result | Main costs |
|---|---|---|---|
| Data Fetcher | $23K | 85% stated margin, implying about $19.55K/month | About $2.5K hosting and $1K software; tax and founder labor excluded |
| Supergrow | Later above $19K | 60–70% stated margin, implying at least $11.4K–$13.3K/month | Affiliate share, software, and operations; profit amount uses the revenue floor |
| ChartDetector | $43.7K in April | About $11.5K profit; 26.3% margin from the reported figures | TikTok acquisition, Apple fees, and ad creative |
Figures are unaudited founder reports from different dates. They illustrate cost structures and margin ranges, not the average profit of the 349 projects.
DealA reported more than $250K spent for roughly $2K monthly profit and negative ROI. Antropy reported strong profit, but with a five-person agency team. “Fully online” and “good solo software bet” are different tests.
Eight representative online-delivery forms have very different operating burdens
These eight are representative patterns, not a complete classification of the 349 projects. They show what the products actually do and which build work AI can reduce.
| Representative form | What the product actually does | Build layer AI can accelerate | Main gap outside the build |
|---|---|---|---|
| Platform workflow extension | Data Fetcher imports external API or web data into Airtable, replacing manual transfer | UI, database, payments, external APIs | Store conversion, platform dependence, integration support |
| Focused task tool | Formula Bot turns a described need into an Excel formula or explanation: one input and one bounded output | Narrow input-output workflow plus inference or rules | Repeat use, inference cost, search-traffic durability |
| Mobile utility | ChartDetector identifies charts and Erly makes users do push-ups to silence an alarm; each repeats one small task on a phone | Mobile UI, subscriptions, analytics, store integration | Paid acquisition, store fees, cohort retention |
| AI customer-support tool | SiteGPT reads a company's website and answers visitor questions in a chat interface | Data ingestion, chat UI, model and vector APIs | Model cost, answer quality, churn, support |
| API or plugin | ScreenshotOne sells website screenshots through an API; Data Fetcher sells data import as an Airtable extension | Endpoint or host-platform integration | Reliability, security review, and platform changes |
| Marketplace | MentorCruise matches mentors and learners and handles listings, trust, and transactions | Standard web and transaction components | Two-sided liquidity, trust, disputes, operations, take rate |
| Productized service | Antropy sells delivery with a relatively fixed scope and process rather than quoting every engagement from scratch | Software can standardize intake, scheduling, and delivery | Labor per revenue unit, utilization, customer concentration |
| Sensitive-document automation | Bank Statement Converter extracts bank-statement data and turns it into structured output | Upload, extraction, conversion, billing | Security, document variance, exception handling, compliance |
- What AI can accelerate
- Interfaces, databases, payments, API integrations, and the first version of a known workflow.
- What revenue figures do not establish
- Demand, distribution, retention, security, reliability, support, and commercial profit.
Build cost is only one part of total cost; acquisition, platform fees, exception handling, and support can outlast the first build.
Seven important unknowns remain about profit, success rates, and operating effort
The public evidence can compare revenue scale, but the seven questions below lack consistent, combinable data.
- How many of the 349 projects are truly profitable?
- Unknown. The database does not consistently disclose costs, net profit, tax, or founder labor.
- What is the average net profit or margin?
- Not calculable. The small number of interviews use different dates and definitions and cannot be combined into a comparable average.
- What is the chance that a new project reaches these revenues?
- Not calculable. Starter Story centers on businesses already producing revenue and does not include every failed or zero-revenue attempt.
- Which form is easiest to make successful?
- Unknown. Categories overlap, while operating, capital, and labor requirements differ sharply.
- How much founder labor is really involved?
- Usually undisclosed. Support, exception handling, maintenance, and sales can materially change a “solo” business's workload.
- How much revenue or profit did AI create?
- Not isolated. The cases only show that AI participated in development; they do not show that it improved demand, retention, or profit.
- Will current revenue persist?
- Only a few cases disclose churn or cohort retention, so a monthly figure does not establish durability.
Main sources and how to read them
The main statistics come from Starter Story's public project database. Profit and operating details come from named founder interviews, with current product sites used to confirm that an offer remains live.
- The 349 projects
- Revenue-generating projects in the platform's public collections. They describe the distribution among visible success cases, not all startup attempts.
- Revenue figures
- Monthly revenue displayed by the platform or stated by founders. It is unaudited and is not net profit.
- Profit cases
- Named interviews that explicitly disclose profit, margin, or costs. They explain mechanisms and are not converted into a population rate.
- Category statistics
- Public project counts and revenue medians defined by the platform. A project may belong to several categories.
- Project databaseStarter Story public projects and category statistics
- Founder interviewData Fetcher: $23K/month micro-SaaS
- Founder interviewSupergrow: $65K in three days
- Founder interviewChartDetector: $50K/month mobile app
- Founder interviewFormula Bot case study
- Founder interviewBank Statement Converter case study
- Founder interviewProfit AI: spreadsheet to micro-SaaS
- Founder interviewErly: simple alarm app
- Founder interviewSiteGPT case study
- Founder interviewMentorCruise mentorship marketplace
- CounterexampleDealA: negative-ROI coupon site
- CounterexampleAntropy: profitable agency with a team
- Current productData Fetcher
- Current productProfit AI
- Current productFormula Bot
- Current productSiteGPT
- Current productBank Statement Converter
- Current productSupergrow
- Current productChartDetector
- Product documentationScreenshotOne screenshot API
Every business figure is a platform display or founder statement from a particular date. The report cannot verify net income, taxes, founder labor, or current churn, and it cannot calculate the probability that a new entrant reproduces a case.