# Inspired.so

Aiolum story: https://aiolum.com/story/inspired-so/ · added 2026-10-27 · updated 2026-10-27

**In short.** Inspired.so (GPT-3 idea generator for blog posts and marketing copy, Not stated in the sources): Sold for US$4K, 2021. Evidence grade D: reported, not verified. Team: 1 founder, building alone in about 60 hours. Launched May 2021; sold 2 months after launch.

GPT-3 idea generator for blog posts and marketing copy · Not stated in the sources · AI-native · solo founder · sold · group: Marketing and social content

- Outcome: Sold for US$4K, 2021 (evidence grade of the headline number: D; headline in US dollars: 4000)
- Launch: May 2021; months from launch to the outcome: 2
- Team: 1 founder, building alone in about 60 hours
- Outside funding: None reported
- Main factor: Early prompts for a new model
- Meters, 1 to 4: evidence strength 1, copyable pattern 2, capital needed 1
- Tags: AI-native, Solo founder, Bootstrapped, Exit, B2B, Copywriting

**The product.** A web app, built in Bubble on OpenAI's GPT-3 model, that suggested ideas for blog posts and for marketing and website copy. Its page called it an AI-powered inspiration engine for copywriters and entrepreneurs. It was sold with its domain, inspired.so, before it had any paying customers.

**Why it worked.** In 2021 GPT-3 was new, access to it needed an invitation, and good prompts took many hours of testing to find. The founder had done that work for content-marketing prompts and put them in a finished Bubble app with payments and analytics. By the founder's account, a buyer who was building a larger AI writing product chose to buy the working app instead of paying a developer to build one.

Evidence grades: A = disclosed by a public company, a filing or a payment processor, or verified by one; B = two or more independent reports agree, or a journalist confirmed it; C = a founder claim that no one else verified; D = reported, not verified: one founder claim or a third-party estimate with no second source; calc = computed on this page from graded facts.

## Summary

Inspired was a web app that wrote ideas for blog posts and for marketing and website copy with GPT-3, a large language model from OpenAI that continues a text from a written instruction, the prompt. Its page described it as "an AI-powered inspiration engine that helps copywriters and entrepreneurs come up with ideas for marketing and website copy". It was built in Bubble, a no-code tool in which an app is put together from visual blocks and workflows instead of written code, and it called the OpenAI API, the paid interface through which a program sends prompts to OpenAI's models.

Fernanda Graciolli, a neuroscientist turned web developer who had learned Bubble since 2017, built it in 2021 in about 60 hours and shared the work in public for feedback, while starting a contract job. They never launched it. They told the No-Code Exits newsletter in November 2022 that they listed the app and the domain inspired.so on MicroAcquire, a marketplace for small online businesses, at US$5,000, received about 15 inquiries in two weeks, and sold it for US$4,000. The app had no revenue and about 10 users, and the deal closed 48 days after the first day of building.

Every number on this page comes from that one interview, so the sale is graded D: reported, not verified. The buyer is not named. The story is small, but it shows something specific to 2021: for a few months, tested prompts for a new model, wrapped in a working app, were an asset that someone would pay for.

## Pricing

Inspired.so prices: none were published in any source this page could read. The rows show the asking and sale prices of the app itself (grade D) and a later listing of the domain (grade C)

| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| App subscription, 2021 |  |  | price not published in any downloaded source |
| Asking price on MicroAcquire, 2021 | 5,000 one-time |  | the app and the domain inspired.so |
| Sale price, 2021 | 4,000 one-time |  | paid by a buyer who is not named |
| Domain inspired.so offered for sale, November 2025 | 888 one-time |  | a later listing on spaceship.com |

The archived pages of 2021 load the Bubble app by script and contain no prices. The founder said the app was fully functional, including payments, but gave no price, and its revenue was zero. The first price column holds one-time prices.

## How much did Inspired.so make, and what was it sold for? The numbers, graded

A GPT-3 app built in about 60 hours and sold with its domain before it had revenue, on the founder's own account

| Fact | Value | Grade | Note | Sources |
|---|---|---|---|---|
| Sale price, 2021 | US$4,000 | D | The founder in No-Code Exits, 10 November 2022: "I listed it for $5,000 and sold it for $4,000". The buyer took the app and the domain. This is the headline figure. | [1] |
| Asking price on MicroAcquire | US$5,000 | D | Same interview. About 15 inquiries in two weeks; the founder said the domain "was also a big asset". | [1] |
| Sale price as a share of the asking price | 80% | calc | Calculated: US$4,000 divided by US$5,000. | [1] |
| Revenue at the sale | US$0 | D | Founder: "Revenue: 0$". The founder decided to focus on a contract job instead of launching or growing the app. | [1] |
| Users at the sale | about 10 | D | Founder: "Users: around 10 users". | [1] |
| Time spent building | about 60 hours | D | Founder: "Time spent: around 60 hours". | [1] |
| Sale price per hour of work | about US$67 | calc | Calculated: US$4,000 divided by 60 hours. | [1] |
| Days from the first day of building to the closed deal | 48 | D | Founder: "It was 48 days from the day I started building it to the day I closed the deal." | [1] |
| Date of the sale | not stated; no later than 25 June 2021 | calc | Calculated: the app was live in the Internet Archive on 8 May 2021, so building began before then, and the deal closed 48 days after building began. If building began after the domain was registered (still a parked page on 18 April 2021), the sale fell between 5 and 25 June 2021. | [1, 2, 3] |
| The product's own description, May 2021 | "Inspired is an AI-powered inspiration engine that helps copywriters and entrepreneurs come up with ideas for marketing and website copy." | C | Description tag of inspired.so in the Internet Archive copy of 8 May 2021, page title "Blog Post Inspiration by AI". The same text is in the copies of 14 June and 26 July 2021. | [3, 4, 5] |
| Built with Bubble | yes | B | The archived page source loads Bubble page data, and the founder says the app integrated GPT-3 with Bubble. | [3, 1] |
| Time from the first archived copy of the live app to the latest possible sale date | about 2 months | calc | Calculated: 8 May 2021 to no later than 25 June 2021 is 48 days, about 1.6 months, rounded to 2. The founder's 48 days from the start of building to the sale give the same. | [1, 3] |
| The domain in November 2025 | offered for sale at US$888 | C | Internet Archive copy of inspired.so, 7 November 2025: a for-sale page listed with spaceship.com. In April 2024 the domain showed a parking page. | [7, 8] |

How to read this table: a grade D value is a statement by the founder in a single newsletter interview, with no second source. A grade C value is read from archived copies of the product's pages. A grade B value is recorded by the Internet Archive itself. A "calc" value was computed on this page from the rows named in its note.

## When did Inspired.so launch, and how did it grow? Timeline

From a parked domain to a sold app in about seven weeks, then a domain that was offered for sale again. Each entry carries its grade.

- **2017** (grade D [1]): The founder, a neuroscientist turned web developer, starts learning Bubble and tries to launch a few products that do not take off.
- **18 April 2021** (grade C [2]): inspired.so shows a parking page: "This domain was recently registered at Namecheap."
- **2021** (grade D [1]): The founder spends many hours on content-marketing prompts in OpenAI's playground, then connects GPT-3 to a Bubble app, building in public for feedback.
- **8 May 2021** (grade C [3]): First archived copy of the live app, titled "Blog Post Inspiration by AI".
- **May or June 2021** (grade D [1, 3]): Listed on MicroAcquire at US$5,000 with the domain; about 15 inquiries in two weeks; sold for US$4,000, 48 days after building began, the founder says. The date is calculated, not stated.
- **14 June and 26 July 2021** (grade C [4, 5]): The archived app still carries the same title and description.
- **9 December 2021** (grade B [6, 1]): The Internet Archive records an error page (status 409) at inspired.so. The founder says the buyer had built the app into a larger product with a new design.
- **10 November 2022** (grade D [1]): No-Code Exits publishes the founder's account of the build and the sale.
- **3 April 2024** (grade C [8]): inspired.so shows a parking page.
- **7 November 2025** (grade C [7]): The domain is offered for sale at US$888 on spaceship.com.

## Why did Inspired.so work?

### Prompts before the crowd

The founder said they spent a lot of time in OpenAI's playground learning to write prompts for content marketing, at a time when GPT-3 was new. That knowledge, not the interface, was what the buyer used: the founder said the buyer kept the prompt engineering and the Bubble integration and changed the design completely.

### A finished app, not a prototype

The app was fully functional, with payments and analytics, before it was listed. The founder's advice to sellers is to fix bugs, finish the design and make it perform well before listing, because buyers who pay the most want a working app.

### The domain as part of the price

The founder said the short domain inspired.so was a big asset in the sale. It came with the app.

### Build or buy

The founder believes the buyer chose a ready-made app over paying a developer, and said they would have charged much more than the sale price of US$4,000 to build it for hire.

## How it was built

These details come from the founder's interview (grade D) and the archived page source (grade B).

- Bubble, a no-code app builder, for the whole app: pages, workflows, payments and analytics. The archived page source loads Bubble page data.
- OpenAI's GPT-3 through its API for the text. The prompts were developed and tested in OpenAI's playground first.
- Built in about 60 hours of the founder's time, while starting a contract job.
- Built in public: the founder shared the work while building and used the feedback to refine the app and the prompts.
- Sold through MicroAcquire, a marketplace for small online businesses. The founder also got the buyer an invitation to GPT-3.

Sources: [1, 3]

## Who tried this before Inspired.so, and why did they fail?

Four companies in the Loot Drop database tried to give writers a better tool and failed, mostly because larger products added the same feature. The values below are from that database and are not independently verified.

The failed startups come from Loot Drop (https://www.loot-drop.io/), a free database of dead startups, retold here in our own words; Loot Drop describes its entries as AI-assisted summaries of public sources, so they carry no evidence grade.

### WriteWith (2007 to 2010; funding US$20K; cause: Competition)

- Promise: Real-time group writing in the browser.
- Why it died: Google Docs offered the same core feature with a larger ecosystem, and the company had too little money to keep up.
- Record: https://www.loot-drop.io/startup/1237-writewith

### Draft (2011 to 2014; funding US$120K; cause: Competition)

- Promise: A simple online editor for writers, with version control and editing by others.
- Why it died: It did not grow beyond its first features, and Google Docs added similar ones.
- Record: https://www.loot-drop.io/startup/1605-draft

### Pressly (2011 to 2018; funding US$2.9M; cause: Competition)

- Promise: A platform for brands to collect, curate and publish content.
- Why it died: Larger marketing suites built curation in, and competitors had broader integrations.
- Record: https://www.loot-drop.io/startup/462-pressly

### Taleship (2018 to 2020; funding US$0; cause: No Market Need)

- Promise: A shared writing app in which writers create stories together.
- Why it died: Too many features spread too thin, with free tools all around and no way to make money.
- Record: https://www.loot-drop.io/startup/1918-taleship

## What changed

The four predecessors built tools for writers and lost when a larger product added the same feature. Inspired did not try to survive that race. Its founder sold it within weeks, at the moment when tested GPT-3 prompts were still scarce and a buyer wanted them inside a larger product.

That is also the limit of the story. The founder said that a buyer paid US$4,000 for the app and the domain, not for customers or revenue. The archived pages show that the app was gone from the domain by December 2021, and a for-sale page listed the domain itself at US$888 in November 2025.

## Can one person copy Inspired.so now?

**Verdict: The 2021 window, no. The pattern, sometimes.**

The 2021 window has closed: GPT-3 is no longer new, and the buyer's reasons, scarce tested prompts and access to the model, no longer apply. The pattern behind the sale still appears each time a new model or API is released: whoever learns to get good results from it first, and wraps that in a finished app, holds something that a builder of a larger product may buy for a few weeks.

1. When a new model or API appears, spend the hours to find prompts or settings that work for one specific job.
2. Finish the app completely, with payments and analytics, before you list it.
3. Put the domain and any access you can transfer into the sale.
4. List it on a marketplace for small businesses and expect a price close to what a developer would charge to build it, not a multiple of revenue.

**What is still open.** New model releases with few experienced users: video, voice and agent APIs, where tested prompts and working integrations are still scarce.

## What could go wrong for Inspired.so?

- **What is not verified.** The headline figure, a sale for US$4,000, rests on one source: the founder's interview in the No-Code Exits newsletter, published on 10 November 2022. The same interview is the only source for the US$5,000 asking price, the zero revenue, the 10 users, the 60 hours and the 48 days. The buyer is not named, no copy of the MicroAcquire listing was found in the downloaded sources, and the founder's own article about the build is on Indie Hackers, which this site does not fetch. A reader could look for the marketplace's record of the deal, the buyer's own account, or the founder's 2021 posts quoted in another publication.
- **No business was sold.** By the founder's account the app had no revenue and about 10 users. The buyer paid for working software, tested prompts, the domain and help with access to GPT-3. The founder said that a developer hired to build it would have charged much more than the price paid.
- **The sale date is not stated.** The interview says 2021 and 48 days from the start of building to the closed deal, but gives no date. The archived copies of the site place the sale no later than 25 June 2021. A reader commenting on the newsletter post also asked when the sale took place.
- **The asset did not last.** By December 2021 the address returned an error in the Internet Archive, by April 2024 it showed a parking page, and in November 2025 the domain was listed for sale at US$888, less than a quarter of what the app and domain sold for together in 2021.
- **Prompts are easy to copy.** Anyone with API access could write similar prompts. The founder said the buyer kept only the prompt engineering and the Bubble integration. Other builders could write similar prompts within weeks.

## Sources

1. No-Code Exits newsletter (Joshua Davis), 10 November 2022: How Fernanda built a GPT-3 No-Code App that got acquired — https://nocodeexits.substack.com/p/how-a-gpt-3-no-code-app-was-built (grade D). Read in full (downloaded 27 October 2026). The founder's interview. The only source for the prices, the revenue, the users, the hours, the 48 days and the buyer's reasons. robots.txt of nocodeexits.substack.com allows /p/.
2. inspired.so, Internet Archive copy of 18 April 2021: parking page — http://web.archive.org/web/20210418123032/http://www.inspired.so/ (grade C). Read in full. "This domain was recently registered at Namecheap."
3. inspired.so, Internet Archive copy of 8 May 2021: Blog Post Inspiration by AI — https://web.archive.org/web/20210508181743/https://inspired.so/ (grade C). Read the title, the description tag and the page source. The visible text loads by script.
4. inspired.so, Internet Archive copy of 14 June 2021 — https://web.archive.org/web/20210614143752/https://inspired.so/ (grade C). Read the title, the description tag and the page source.
5. inspired.so, Internet Archive copy of 26 July 2021 — https://web.archive.org/web/20210726222741/https://inspired.so/ (grade C). Read the title, the description tag and the page source.
6. Internet Archive index of inspired.so, 2017 to 2025 — https://web.archive.org/cdx/search/cdx?url=inspired.so&matchType=domain (grade B). Read in full. Lists the archived copies and their status codes, including the error of 9 December 2021.
7. inspired.so, Internet Archive copy of 7 November 2025: domain for sale — https://web.archive.org/web/20251107132538/http://inspired.so/ (grade C). Read in full. A for-sale page listed with spaceship.com at US$888.
8. inspired.so, Internet Archive copy of 3 April 2024: parking page — https://web.archive.org/web/20240403215259/https://www.inspired.so/ (grade C). Read the page source: a domain parking script and no content.
9. Loot Drop database export, 1,749 failed startups — https://www.loot-drop.io/ (grade C). Local export. AI-written summaries, not independently verified. Source for the failed predecessors.