ColorBliss: Revenue, team and how it was built
ColorBliss (AI coloring page generator that makes printable line art from a prompt, a photo or a sketch, USA): US$2K MRR, May 2024. Evidence grade D: reported, not verified. Team: 1 founder working alone, by his own account; no employees reported. Launched 29 September 2023; still running.Added 27 October 2026
- Outcome
- US$2K MRR, May 2024
- Evidence
- D reported, not verified
- Launch
- 29 September 2023
- Status
- Still running
- Team
- 1 founder working alone, by his own account; no employees reported
- Outside funding
- None reported; the founder wrote that his family had about a year of runway when he quit his job
- Group
- Photos and portraits
- Main channel
- Search and AI answers
What is not verified
The headline US$2,000 of monthly recurring revenue rests on one founder post from about May 2024, which this page could read only as eBiz Facts quoted it on 25 July 2024 and Starter Story on 25 October 2024. No payment processor, tax filing or buyer has confirmed any revenue figure, and the founder's later posts, in September 2025 and February 2026, describe the business without numbers. A reader could look for a payment-verified profile, for example on TrustMRR (ColorBliss takes payments through Stripe), or for a newer interview with dated figures; his Indie Page hides ColorBliss's revenue. The headline number is grade D: one claim or one estimate, with no second source. Aiolum shows it because the story is useful, and marks it so that it is not mistaken for a verified figure.
ColorBliss is an AI coloring page generator: a web app that sends a short text description, a photo or a sketch to an image model and returns clean black outlines on white that a child or an adult can color by hand. An image model is an AI model that draws a new picture from a prompt or a reference picture; the founder's tools page names OpenAI, Replicate and fal.ai as his providers. Users buy credits, units of use in which one standard page costs one credit, through plans that the company's pricing page lists at US$9 to US$99 a month.
Ben Robertson, a software developer in South Carolina who had worked in customer success at Gatsby Inc and Netlify, quit his job in September 2023 and launched ColorBliss alone at the end of that month. He wrote that it made US$92.50 in its first month and cost US$179.55, mostly for ads. In a post that eBiz Facts quoted on 25 July 2024, he said that it took 93 days to reach US$100 of MRR, monthly recurring revenue, 94 more days to reach US$1,000, and 26 more days to reach US$2,000, which was 213 days after launch, about May 2024.
That US$2,000 figure, which the founder stated in one post, is his latest dated revenue figure, and no payment processor has confirmed it, so this page is marked "Reported, not verified" with evidence grade D. In September 2025 he wrote that he had set out to support his family on 20 hours of work a week and that, two years in, "the theory is holding", without a number. The page also shows earlier companies that sold creative content or tools to parents and failed.
How much revenue does ColorBliss make?
Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.
| Fact | Value | Grade | How we know | Sources |
|---|---|---|---|---|
| Monthly recurring revenue, about May 2024 | US$2,000 | D | The founder's post as quoted by eBiz Facts on 25 July 2024: "And now, just 26 days later, I made it to $2k!" Starter Story quoted the same post on 25 October 2024 as an Indie Hackers post titled "Hit 2k MRR!". Both quote one statement; X and Indie Hackers were not opened. | [1, 2] |
| Run rate from that figure | US$24,000 a year | calc | Calculated: US$2,000 times 12. | [1] |
| Steps to US$2,000 MRR | 93 days to US$100, 94 more days to US$1,000, 26 more days to US$2,000 | D | The same post: "It was 93 days until I hit $100 MRR." "It took 94 more days to grow to $1k MRR." Calculated: 93 + 94 + 26 = 213 days. | [1, 2] |
| Date of the US$2,000 MRR | about May 2024 | calc | Calculated: 213 days after a launch in October 2023, as the post says, falls between 1 and 31 May 2024; counted from the launch post of 29 September 2023 it is 29 April 2024. The post itself was not dated in the sources. | [1, 5] |
| First month, 13 September to 20 October 2023 | US$92.50 of income, US$179.55 of expenses, net minus US$87.05 | C | Founder's blog, month one update. Expenses: Facebook ads US$101, Pinterest ads US$36.14, image generation US$38.95, domain US$3.46. | [4] |
| Revenue in October 2023 | about US$240 | C | Founder's blog, September 2025: "I believe ColorBliss did $240 in October." | [3] |
| Launch and first sale | first version at the end of September 2023; first sale on about 3 October 2023 | C | The launch post is dated 29 September 2023. In the week of 2 October the founder launched on Product Hunt, where it ended 14th for the day, and ran Facebook ads. The X post says "I launched ColorBliss in October." | [5, 4, 3, 1] |
| Cost and price at launch | about 2 cents per image; US$9 for 20 credits | C | Founder's launch post. Starter Story gives the same price. | [5, 2] |
| Where sales came from | mostly organic search | D | The founder's post as quoted by eBiz Facts: "most sales have come from organic search", with some paid ads and email. An interview in October 2025 says SEO remains his core growth channel (grade C). | [1, 8] |
| Users | more than 40,000 monthly users (October 2025); 153,000+ sign-ups and 1,723,702 pages created (October 2026) | C | AI Empires interview of 24 October 2025, and the counters on the company's home page read 27 October 2026. | [8, 9] |
| Team | 1 person | C | AI Empires: "ColorBliss was built entirely by Ben himself". His tools page lists "The tools I use to build, run, and grow ColorBliss as a solo founder". The privacy policy is written in the first person singular. The tools page also mentions contract work that he does. | [8, 6, 11] |
| Working time and income | no more than 20 hours a week; the founder says it supports his family | C | Founder's blog, 18 September 2025: he wanted to test whether 20 hours a week could support his family, "(Two years in, and the theory is holding!)". No figure is given. | [3] |
| Outside funding | none reported | C | The founder wrote that he wanted a break from venture capital style businesses and that his family had a safe runway of about a year. | [3] |
| Company | Marginalia Ventures LLC | C | Footer of the company's terms and privacy pages: "ColorBliss is a product of Marginalia Ventures LLC." | [12, 11] |
| A newsletter's later figure, not used on this page | "September 2024: $6,300 monthly recurring revenue" | D | Startup Spells, 2 January 2026. The words link to the founder's Indie Page, which shows "$6.3k/month" next to a progress bar; the page data gives this as an MRR goal of US$6,250 and marks ColorBliss's revenue as not shared, in October 2024, May 2025 and October 2026 alike. The same newsletter's range of US$30,000 to US$100,000 a month is its own estimate with no method. | [16, 15] |
| Reviews shown by the company | 4.7 from 242 reviews on the reviews page; 4.91 from 233 in the page data; "from 243 reviews" on the home page | C | The reviews page lists Trustpilot (89), Capterra (22), Twitter (5), Product Hunt (3) and LinkedIn (1) as sources, and its rating breakdown shows 211 five-star and 21 four-star reviews and none lower. | [13, 9] |
| Time from launch to the headline figure | 7 months | calc | Calculated: October 2023 to May 2024. | [1, 5] |
How to read this table: a grade D value is a founder statement that only one post contains, quoted by eBiz Facts and Starter Story, or a newsletter's figure with no second source. A grade C value comes from the founder's own blog, an interview, or the company's pages read on 27 October 2026. A "calc" value was computed on this page from the rows above it. No value is grade A or B.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Free account | 0 | 0 | up to 12 coloring pages |
| Starter | 7.00 | 9 | 250 credits a month, personal use |
| Hobby | 12.42 | 18 | 500 credits a month, drawings and advanced editing, personal use |
| Artist | 24.75 | 35 | 1,000 credits a month, bulk generation of 50 pages, commercial use |
| Business | 83.08 | 99 | 5,000 credits a month, API access after approval, commercial use |
| Credit packs | 25 to 80 one-time | 500, 1,000 or 2,000 credits; expire after six months without activity |
Yearly billing is US$84, US$149, US$297 or US$997 a year. A standard page from a text prompt costs 1 credit; a regular-quality photo conversion or the Ghibli style costs 3, the detailed style 5, character sheets and high resolution 10. Hobby and higher plans roll unused credits over, and only Artist and Business allow commercial use, such as selling books on Amazon KDP or Etsy. The pricing page names Stripe for payments and offers a 7-day money-back guarantee, while the home page FAQ says that generated pages cannot be refunded and that each plan includes extra retry credits instead.
When did ColorBliss launch, and how did it grow?
From a 4 a.m. keyword search to a coloring page business that, its founder says, supports his family. Each entry carries its grade.
- Early September 2023CBen Robertson leaves Netlify, which had acquired Gatsby Inc, his employer of three years. He plans to work at most 20 hours a week. [3]
- 13 September 2023CAwake at 4 a.m., he researches coloring page searches in SEMrush, buys the domain ColorBliss.art and makes the first commit. [3, 4]
- 29 September 2023CLaunch post: a custom coloring sheet generator and free printable pages, with credits at US$9 for 20, and a Facebook ad campaign aimed at parents of preschoolers. [5]
- Week of 2 October 2023CProduct Hunt launch, 14th for the day, and the first two sales, both from strangers. [4]
- 20 October 2023CMonth one update: US$92.50 of income and US$179.55 of expenses. Pinterest ads convert better and cost less than Facebook ads. [4]
- About January 2024DUS$100 MRR, 93 days after launch, according to the founder's post. [1]
- About April 2024DUS$1,000 MRR, 94 days later, according to the same post. Between US$1,000 and US$2,000 he added features and changed pricing. [1]
- About May 2024DUS$2,000 MRR, 26 days later, 213 days after launch, according to the same post. [1, 2]
- June 2024CHe buys colorbliss.com, which had been offered for about US$2,500 in September 2023. [3]
- 25 July 2024DeBiz Facts publishes the founder's figures, and Starter Story repeats them on 25 October 2024. [1, 2]
- 18 September 2025CTwo-year post: he writes that 20 hours a week still supports his family. [3]
- 24 October 2025CAI Empires interview: more than 40,000 monthly users, built by one person, SEO as the main channel. [8]
- 6 February 2026CHe writes that conversions slowed in January after a very good December, and that his onboarding emails had not been sent for a month after a migration. [7]
- 27 October 2026CThe site shows 1,723,702 pages created, 153,000+ sign-ups and plans from US$9 to US$99 a month. [9, 10]
Why did ColorBliss work?
Four causes, in the order they mattered. Each one is a reusable lesson.
Search demand found before the product
The founder wrote that searches such as "mushroom coloring pages" and "Bowser coloring pages" had low difficulty but thousands of searches, and that a large free coloring site lived on ads. He made free pages for such searches first and put the generator on top of them. The home page now lists more than 8,800 free printable pages.
A product the general tools do not make well
General image models draw shaded color pictures. ColorBliss returns clean outlines at print resolution and adds the workflows that coloring needs: photo conversion, word art, the same character on every page and bulk generation of 50 pages for book makers.
Cheap images, paid credits
The founder's launch post said each image cost him about 2 cents while 20 credits sold for US$9. Plans now sell 250 to 5,000 credits a month, and expensive modes such as photo conversion cost more credits, so heavy use is paid for.
Small paid tests, then search
The founder's first-month report said Facebook ads brought traffic but no paid users, while Pinterest ads cost about US$1.95 per thousand ad views and converted better. Product Hunt, an early listing on There's An AI For That and planned outreach to parent bloggers were his ways to get the first links to the site; later, he said, organic search brought most sales.
How it was built
These details come from the founder's blog and tools page (grade C), Starter Story (grade D) and the company's own pages (grade C). [6, 5, 11, 2, 9]
- The first version took a couple of weeks of part-time work in September 2023. The founder worked out the prompts for line art with Stability Dreambooth, according to his launch post.
- The app dashboard is built with Next.js, the marketing site with Astro, and content is managed in WordPress. Supabase is the database and login service; backend jobs moved to Laravel after traffic spikes, deployed with Laravel Forge; hosting is on Netlify.
- Images come from OpenAI's API, Replicate and fal.ai, with requests spread across providers; the founder moved to Replicate and fal after OctoAI was acquired by NVIDIA. A prompt improver runs on Cerebras for speed.
- Stripe takes the subscription payments. Bento and Resend send email, PostHog measures product use, and Ahrefs replaced SEMrush for search research.
- He writes code in Cursor and tracks features in Linear. The privacy policy says uploaded photos are deleted after the page is generated and that OpenAI does not train on them.
Tools named in the sources
Each one opens its page in the Tools tab, with every story that uses it.
Who tried this before ColorBliss, and why did they fail?
Four companies in the Loot Drop database sold creative content or creative tools to parents and consumers, and failed. The values below are from that database and are not independently verified.
Zuuka
The promiseA library of interactive children's books for parents and teachers.
Why it diedApple and Amazon added interactive children's books to stores that already had the customers, and Zuuka could not pay for the content and marketing to keep up.
Airy Labs
The promiseLearning games that make parents proud of their children's screen time.
Why it diedEach game needed game designers and learning experts, while parents would not pay much for children's apps when free ones existed.
Snipshot
The promiseEdit photos in the browser without installing software.
Why it diedLarger free tools such as Picnik offered more, and it never found a clear way to make money.
FiftyThree
The promiseA digital sketchbook for the iPad with its own stylus.
Why it diedApple's own Pencil and stronger drawing apps took away what made it different, and the hardware raised its costs.
The failed startups come from Loot Drop, 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 these cards carry no evidence grade.
What changed
Zuuka and Airy Labs needed paid artists, writers or learning experts for every book or game, and parents would not pay enough to cover that work. Snipshot and FiftyThree built creative tools and lost when larger companies gave away the same features.
ColorBliss has no artist on staff: each page is drawn by a rented image model, and the founder's launch post put the cost at about 2 cents. That makes it possible to give away thousands of free pages for small searches and to answer a child's exact request in seconds. The risk that remains is the one that ended Snipshot and FiftyThree, since general AI image tools from large companies can also draw a coloring page when asked.
Can one person copy ColorBliss now?
Yes, for a narrow output that people search for.
One person built the first version in a couple of weeks with rented image models, and the free pages that bring search traffic cost almost nothing to make. The niche now has competitors, and the founder still depends on Google rankings and on outside model providers. Only the founder's own statements describe the revenue.
Where it is still open: The same pattern for other printables with many variations: worksheets for one school subject, stencils, embroidery or cross-stitch patterns, and craft templates.
- 1Look for a printable or visual item that people search for in many small variations, and check the search volume before you build.
- 2Generate free example pages for those searches, and put the paid generator on every page.
- 3Sell credits, and charge more credits for the modes that cost more to generate.
- 4Test small ad budgets on several networks, measure paying users per network, and stop the ones that bring none.
What could go wrong for ColorBliss?
The risks are the mirror image of the causes above.
What is not verified
The headline US$2,000 of monthly recurring revenue rests on one founder post from about May 2024, which this page could read only as eBiz Facts quoted it on 25 July 2024 and Starter Story on 25 October 2024. No payment processor, tax filing or buyer has confirmed any revenue figure, and the founder's later posts, in September 2025 and February 2026, describe the business without numbers. A reader could look for a payment-verified profile, for example on TrustMRR (ColorBliss takes payments through Stripe), or for a newer interview with dated figures; his Indie Page hides ColorBliss's revenue.
A larger figure that is probably a goal
Startup Spells wrote on 2 January 2026 that ColorBliss had US$6,300 of monthly recurring revenue in September 2024, and linked to the founder's Indie Page. That page shows "$6.3k/month", and its data stores this as an MRR goal of US$6,250, with ColorBliss's revenue not shared. This page therefore does not use the figure, nor the newsletter's own estimate of US$30,000 to US$100,000 a month.
Review counts that do not match
The home page says "from 243 reviews", the reviews page says 4.7 from 242, its rating breakdown adds up to 232 and shows no review below four stars, and the page data claims 4.91 from 233. The reviews page names Trustpilot, Capterra, X, Product Hunt and LinkedIn as sources for 120 of them; its source filter names none for the others. The founder's own family coloring book appears under his name in the gallery of user creations.
Characters that belong to others
The free library has pages of Spider-Man, Hello Kitty, Yoshi and Batman, the FAQ invites users to name popular characters, and a premium mode is called the Ghibli style. Commercial plans allow selling books on Amazon KDP. The terms forbid content that infringes intellectual property, but a rights holder could still ask for pages or books to be removed.
Children and family photos
The site answers "Is it safe for kids to use?" with yes, while the privacy policy says that the service is intended for adult use and that users under 13 need a parent's consent. Parents upload family photos, which go to OpenAI's API; the policy says the originals are deleted after generation and that OpenAI does not train on them.
Dependence on search and on model providers
The founder names organic search as the main source of sales, so a change in Google's ranking of AI-generated pages would hit the business directly. The images come from OpenAI, Replicate and fal.ai, and he already had to move once, when OctoAI was acquired by NVIDIA. His Indie Page also lists other products of his: Holly Support, Churnpill, Color Anything and a warrior cat name generator.
Sources
The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.
- D eBiz Facts, 25 July 2024, by Fardeen Khan: His AI Coloring Page Software: $2000/Month in 213 DaysRead in full in the site's RSS feed (page 2), downloaded by the pipeline on 27 October 2026; the robots.txt group for all agents allows /feed/ and posts. It quotes the founder's post on X, which was not opened. Source of the headline figure.
- D Starter Story, 25 October 2024: How Ben Robertson Grew ColorBliss to $2K MRR in Just 7 MonthsRead in full. The page is not marked "Estimated from public sources"; it lists the founder's Indie Hackers post "Hit 2k MRR!" and his blog as sources. The Indie Hackers post was not opened. Source for the price at launch, the tools and the Product Hunt traffic.
- C Ben Robertson's blog, 18 September 2025: ColorBliss is two years old!Read in full. The founder's own account of the start, the first sales, the October 2023 revenue and his working rules. The site's robots.txt has no rules.
- C Ben Robertson's blog, 20 October 2023: ColorBliss - Month One UpdateRead in full. Source for the first-month timeline, income, expenses and ad results.
- C Ben Robertson's blog, 29 September 2023: Launching Colorbliss.artRead in full. Source for the launch date, the first price, the image cost and the first model.
- C Ben Robertson: Uses, the tools behind ColorBliss, read 27 October 2026Read in full. The founder's list of the tools he uses to build, run and grow ColorBliss.
- C Ben Robertson's blog, 6 February 2026: Conversions slowed way down for ColorBliss in JanuaryRead in full. A short note without figures.
- C AI Empires (There's An AI For That), 24 October 2025: Turning FOMO Into 1M Users, section How ColorBliss Prints Ideas on DemandRead the section on ColorBliss, a summary of a video interview with the founder; the video was not watched. The robots.txt group for all agents disallows only /login.
- C ColorBliss home page, read 27 October 2026Read in full, including the structured data, for the purpose check and the counters. The robots.txt disallows only /auth/, /account/ and /admin/.
- C ColorBliss pricing page, read 27 October 2026Read in full. Source for the plans, credit costs and credit packs.
- C ColorBliss privacy policy, last updated 6 January 2026Read in full. Source for the AI provider, photo handling, analytics tools and the age rule.
- C ColorBliss terms of service, last updated 23 March 2026Read in full. Source for the commercial license, the content rules and the company name.
- C ColorBliss reviews page, read 27 October 2026Read the summary, the rating breakdown, the source filters and a sample of reviews, for the purpose check.
- C ColorBliss affiliate program page, read 27 October 2026Read in full, for the purpose check: it states a 20% recurring commission and a calculator based on that commission, and makes no claim about what users earn.
- C Indie Page: Ben Robertson, read 27 October 2026, with Wayback Machine copies of 9 October 2024 and 21 May 2025Read in full, including the data embedded in the page, which gives the MRR goal of US$6,250 and marks ColorBliss's revenue as not shared. The robots.txt allows every path.
- D Startup Spells, 2 January 2026: ColorBliss' Programmatic SEO Strategy: How 8,700 AI Pages Captured Long-Tail SearchRead in full. Cited only for its US$6,300 figure and its estimate, which this page does not use; its links point to the Indie Page and to a post on X that was not opened.
- C Loot Drop database export, 1,749 failed startupsLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.