# RomeoSEO

Aiolum story: https://aiolum.com/story/romeoseo/ · added 2026-11-03 · updated 2026-11-03

**In short.** RomeoSEO (AI SEO agent that writes and fixes pages for approval, United Kingdom): US$500 a month in recurring revenue, October 2026. Evidence grade D: reported, not verified. Team: 1 founder; no other person named on the site. Launched July 2026; still running, revenue growing.

AI SEO agent that writes and fixes pages for approval · United Kingdom · AI-native · solo founder · still running, revenue growing · group: SEO and AI search

- Outcome: US$500 a month in recurring revenue, October 2026 (evidence grade of the headline number: D; headline in US dollars: 6000)
- Launch: July 2026; months from launch to the outcome: 3
- Team: 1 founder; no other person named on the site
- Outside funding: None found in the sources
- Main factor: Audience
- Meters, 1 to 4: evidence strength 1, copyable pattern 2, capital needed 1
- Tags: AI-native, Solo founder, Bootstrapped, Running, B2B, Search marketing

Links:

- Visit romeoseo.com: https://romeoseo.com/

**The product.** An AI agent for a company's own website. Every day it picks one keyword and writes one article for it, with images and internal links, and it picks one existing page that is losing Google traffic and rewrites its text, title or links. Nothing goes live until the site owner approves it. It connects to Google Search Console, GitHub, Framer and WordPress, and it can also be used from Claude, ChatGPT and Cursor.

**Why it worked.** The founder first built the system for his own dictation app, Voicy, and says on the home page that it took that site to 30,000 Google clicks a month. He then sold the same system to other site owners at one flat price per site. In his own words, most of the first customers came from two articles that he posted on X.

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

RomeoSEO is an AI agent for SEO, search engine optimization: the work of making a website's pages appear higher in Google results and, more recently, in answers from AI assistants such as ChatGPT. It reads the site's own Google Search Console data (a free Google service that shows for which searches each page appears, how often and how many people click), chooses one new article to write and one weak page to improve each day, and prepares both for the owner's approval. On sites kept in GitHub the change arrives as a pull request, a proposed code change that the owner reviews and merges.

Kourosh Ghaffari, who also runs the dictation app Voicy, writes on the About page that he founded RomeoSEO in London in 2026. On 23 October 2026 he posted on X that RomeoSEO had reached US$500 MRR from 6 customers, three months after he started it, as quoted by the founder directory arrfounder.com. MRR means monthly recurring revenue, the subscription income of one month. On 27 October he posted "$10,000 ARR", annual recurring revenue, and arrfounder now lists the product at US$868 a month.

Both figures are the founder's own posts, relayed by a directory that says it does not check them, so this page grades the headline D, reported and not verified. The home page still asks new users to join a waitlist, and access opens by invitation. The page also shows earlier companies that promised small businesses more web traffic or easier content marketing, and how they failed.

## Pricing

RomeoSEO prices on the home page, read 3 November 2026 (grade C)

| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Monthly, early-bird |  | 69 | per site; 30 new pages and 30 improved pages a month, monthly call with the founder |
| Monthly, regular price (announced) |  | 149 | per site; same content |

One plan, billed monthly per site, with no yearly option shown. The home page calls US$69 an early-bird price that will rise to US$149 a month later. Onboarding starts with a 3-day free trial that needs a card. The Internet Archive copy of 11 October 2026 shows the same US$69 price without the US$149 note.

## How much revenue does RomeoSEO make? The numbers, graded

A founder-reported US$500 MRR from 6 customers on 23 October 2026, then a founder claim of US$10,000 ARR four days later, for an invite-only AI SEO agent at US$69 a month per site

| Fact | Value | Grade | Note | Sources |
|---|---|---|---|---|
| MRR, 23 October 2026 | US$500 from 6 customers | D | The founder's post on X, as quoted by arrfounder.com: "$500 MRR with Romeoseo.com" and "With Romeo it took 6 customers". This is the headline figure. Only arrfounder quotes it; the X post itself was not opened. | [1] |
| Run rate from the 23 October figure | US$6,000 a year | calc | Calculated: US$500 times 12. | [1] |
| ARR claim, 27 October 2026 | US$10,000 | D | The founder's post on X, as quoted by arrfounder.com: "YO $10,000 ARR WITH @romeoseocom", with "The goal is $2k MRR by the end of the year". Not used as the headline: see the next two rows. | [2] |
| Arrfounder's listing, updated 2 November 2026 | US$868 a month, US$10K a year | D | Directory profile. US$868 times 12 is US$10,416 (calculated), so the directory derives its monthly figure from the ARR claim. Its disclaimer says the data is disclosed by founders, mainly on X, and not checked. | [3] |
| Change from 23 to 27 October 2026 | from US$500 to about US$833 a month, plus 67% in four days | calc | Calculated: US$10,000 divided by 12 is US$833; the rise of US$333 equals about 4.8 more sites at US$69. The two posts do not contradict each other, but no source explains the rise, so this page uses the smaller, earlier figure. | [1, 2, 4] |
| Revenue per customer, 23 October 2026 | about US$83 a month | calc | Calculated: US$500 divided by 6. Above the US$69 price per site, which fits customers with more than one site; 7 sites would be US$483 (calculated). | [1, 4] |
| Price | US$69 a month per site, early-bird; US$149 announced | C | Home page, 3 November 2026. 3-day free trial with a card. The Internet Archive copy of 11 October 2026 shows US$69. | [4, 8] |
| Access | waitlist; access by invitation | C | Home page, 3 November 2026: "Access is opening in stages. Join the waitlist and we'll email you when it's your turn." The pricing box says "Access by invite". The archive copy of 11 October 2026 shows the same waitlist. | [4, 8] |
| Monthly output per site | 30 new pages and posts, 30 improved pages | C | Home page pricing box. The home page says the agent works one article and one fix a day, and that approved work is published automatically. | [4] |
| Founder and place | Kourosh Ghaffari, London, 2026 | C | About page: "I founded RomeoSEO in London in 2026". He writes that he did SEO for six years before. | [5] |
| Team | 1 founder | C | The home page names the founder as the person on the monthly customer call; no other person is named on the site. The X post of 23 October speaks of "my" products. | [4, 5, 1] |
| Where the first customers came from | two articles on X | C | Founder's post of 23 October 2026, as quoted by arrfounder: "I got most of my customers from 2 X articles that did really well." He plans outbound sales, more articles and SEO. | [1] |
| Outside funding | none found | C | No source mentions investors or a funding round. | [4, 5] |
| Time from start to the headline figure | 3 months | calc | Calculated: July 2026 to October 2026. The founder wrote "It took 3 months to get here" on 23 October. | [1] |

How to read this table: a grade D value is a founder post that only a directory quotes, with no second source. A grade C value is a statement on the company's own pages. A grade B value is a record kept by an outside archive. A "calc" value was computed on this page from the rows above it. No value is grade A.

## When did RomeoSEO launch, and how did it grow? Timeline

From the founder's own SEO system to a paid product with a waitlist, in about three months. Each entry carries its grade.

- **July 2026** (grade C [1, 5]): The founder starts building RomeoSEO, three months before his 23 October post. The agent comes from the SEO system that he built for his app Voicy.
- **August 2026** (grade C [7]): First articles on the RomeoSEO blog.
- **11 September 2026** (grade C [10]): Terms of service and privacy policy published in their current form, with Stripe named as the payment processor.
- **30 September 2026** (grade C [9]): The founder announces RomeoSEO on X, as recorded on this site's Voicy story.
- **11 October 2026** (grade B [8]): The Internet Archive copy of the home page shows a waitlist and US$69 a month per site.
- **23 October 2026** (grade D [1]): The founder posts US$500 MRR from 6 customers, quoted by arrfounder.
- **27 October 2026** (grade D [2]): The founder posts US$10,000 ARR and a goal of US$2,000 MRR by the end of 2026, quoted by arrfounder.
- **3 November 2026** (grade C [4]): The home page still shows the waitlist, with the early-bird price and a planned US$149 regular price.

## Why did RomeoSEO work?

### A tool the founder used first

The founder built the agent to write the blog of his own app, Voicy, and the home page says that site reached 30,000 Google clicks a month with it. Selling it to others started from a working system and a result he could show.

### B2B prices reach a target with fewer customers

In his 23 October post the founder said his consumer app costs US$8.49 a month and would need 59 customers for the same MRR, while RomeoSEO needed 6. A flat price per site makes each new customer worth as much as several consumer sign-ups.

### Small, approved daily output

The product does one new article and one page fix a day and waits for the owner's approval. The home page says "You are not buying volume nobody reads", and the free programmatic SEO guide on the site warns against Google's scaled content abuse and doorway page policies.

### An audience that already follows the founder

The founder posts his revenue numbers on X. He wrote on 23 October that two articles on X brought most of the customers, and that he cannot repeat that every day.

## How it was built

These details come from the company's privacy policy, terms of service and home page. They are grade C.

- OpenAI models for analysis and for writing text and images, and DataForSEO for keyword, ranking and search-results data, according to the privacy policy.
- Read-only access to Google Search Console; the connection secrets are stored encrypted.
- Changes go out as GitHub pull requests through a GitHub App, or as drafts in Framer and WordPress. The terms say RomeoSEO never merges a pull request into the default branch itself.
- Hosting with managed databases, Cloudflare services and R2 storage, Sentry for errors, PostHog for analytics, Resend for email, Cal.com for the optional monthly call, and Stripe for billing.
- An MCP server, the Model Context Protocol that lets AI apps call outside tools, so customers can ask about their Google results or start work from Claude, ChatGPT or Cursor.
- 47 free SEO skill files for AI agents under the MIT license, published on the site and on GitHub, which bring visitors to the site.

Sources: [10, 4, 6, 7]

## Who tried this before RomeoSEO, and why did they fail?

Four companies in our database of failed startups promised small businesses more web traffic or easier content marketing, and failed. The values below are from that database and are not independently verified.

The failed startups come from our database of dead startups, retold here in our own words; Its entries are AI-assisted summaries of public sources, so they carry no evidence grade.

### LocalOn (2013 to 2016; funding US$500K; cause: Competition)

- Promise: Websites, social media and local SEO for small shops, without technical skills.
- Why it died: Shopify and Wix offered the same and more. It ran out of money before small businesses noticed it.
- Record: 

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

- Promise: A platform for companies to collect, curate and publish content for their customers.
- Why it died: It stayed a curation tool while competitors sold full content-marketing suites.
- Record: 

### Marketfox (2017 to 2020; funding US$2M; cause: Competition)

- Promise: Automated email, push messages and segments for online shops.
- Why it died: Mailchimp and other large tools had more money and more features.
- Record: 

### Argyle Social (2009 to 2013; funding US$2.5M; cause: Competition)

- Promise: Social media marketing for lead generation, connected to Pardot and Marketo.
- Why it died: Larger competitors with bigger marketing budgets took the market while it made small updates.
- Record: 

## What changed

The four predecessors sold software that still needed a person to write the content, and each lost to larger suites with more features. Their cause of death in the database is competition every time.

RomeoSEO sells the work itself: a language model writes the article and the fix, and the owner only approves. That makes a one-person company possible, and the founder says the first customers paid US$69 a month per site for it. Competition is still the main risk, now from other AI SEO tools and from AI assistants that can do part of the work directly.

## Can one person copy RomeoSEO now?

**Verdict: Yes, if you already have a result to show.**

The method is to build a tool for your own product, measure what it did, then sell the same tool to people with the same problem at a flat business price. The founder says he had six years of SEO work and a traffic result on his own site before he started.

1. Automate a task for your own product first and record the result with real data.
2. Sell it per site or per seat at a business price, not a consumer price.
3. Keep the output small and approved by the customer, so quality stays visible.
4. Post your numbers and your method where your buyers read, and plan a second channel before the first one fades.

**What is still open.** Agents that do one piece of recurring marketing or operations work for small companies and ask for approval before they act.

## What could go wrong for RomeoSEO?

- **What is not verified.** Both revenue figures rest on the founder's posts on X: US$500 MRR from 6 customers on 23 October 2026 and US$10,000 ARR on 27 October 2026. Only the directory arrfounder.com quotes them, and its disclaimer says it does not check them; this page did not open the X posts. There is no payment-processor profile (TrustMRR has no RomeoSEO page on 3 November 2026). A reader could look for a Stripe-verified profile or an interview with a dashboard.
- **Two figures four days apart.** US$500 MRR on 23 October is US$6,000 a year; the US$10,000 ARR claim on 27 October means about US$833 a month (calculated). No source explains the rise. This page uses the earlier, smaller figure as the headline.
- **New users join a waitlist.** On 3 November 2026 the home page offers "Join the waitlist" and says access is "by invite". Paying customers exist by the founder's account, but a new reader cannot sign up at once.
- **The home page describes two schedules.** Most of the home page says the agent checks every page every day, but one answer says "You run an audit when you want one" and that automatic weekly runs are on the roadmap.
- **Many AI pages on the founder's other site.** This site's Voicy story records the founder's statement that Voicy has more than 1,000 pages and that his agent writes its blog posts. RomeoSEO itself sells 30 approved pages a month per site and warns against scaled content in its own guide.

## Sources

1. Arrfounder news, 23 October 2026: Romeoseo Hits $500 MRR — https://arrfounder.com/news/romeoseo-hits-500-mrr (grade D). Read in full. Quotes the founder's X post (https://x.com/BuildingWaves/status/2113569186370720083) word for word; the X post itself was not opened. Source for the MRR, the 6 customers, the 3 months and the two X articles.
2. Arrfounder news, 27 October 2026: Romeoseo Hits $10K ARR — https://arrfounder.com/news/romeoseo-hits-10k-arr (grade D). Read in full. Quotes the founder's X post (https://x.com/BuildingWaves/status/2115126629575831721); the X post itself was not opened. Source for the ARR claim and the US$2,000 MRR goal.
3. Arrfounder product page: Romeoseo, updated 2 November 2026 — https://arrfounder.com/product/romeoseo (grade D). Read in full. A founder directory that lists US$868 a month and US$10K a year and says the data is not checked.
4. RomeoSEO home page, read 3 November 2026 — https://romeoseo.com/ (grade C). Read in full. Source for the product, the waitlist, the price, the trial, the integrations and the founder's statement about his own site.
5. RomeoSEO: About RomeoSEO, meet the human behind Romeo — https://romeoseo.com/about (grade C). Read in full. Source for the founder, London, 2026 and his six years of SEO work.
6. RomeoSEO: Use Romeo in Claude, ChatGPT and Cursor — https://romeoseo.com/mcp (grade C). Read in full. Source for the MCP connector.
7. RomeoSEO blog, free SEO skills and the programmatic SEO plan skill — https://romeoseo.com/skills/programmatic-seo-plan (grade C). Read the skill page in full, the skills index and the blog index with its post dates (https://romeoseo.com/blog). Source for the warning against scaled content and doorway pages, the 47 skills under MIT license, and the first blog dates.
8. Internet Archive copy of romeoseo.com, 11 October 2026 — https://web.archive.org/web/20261011001212/https://romeoseo.com/ (grade B). Read in full. Shows the waitlist and US$69 a month per site.
9. Aiolum: Voicy story — https://aiolum.com/story/voicy/ (grade C). Read the facts table and timeline. This site's earlier page, which read the founder's X profile on 14 October 2026; source for the 30 September announcement and the 1,000 Voicy pages.
10. RomeoSEO privacy policy and terms of service, last updated 11 September 2026 — https://romeoseo.com/privacy (grade C). Read both in full (terms at https://romeoseo.com/terms). Source for the service providers, the integrations, Stripe billing and the pull-request rule.
11. Database of failed startups, local export — /method/#dead (grade C). Local export. AI-written summaries, not independently verified. Source for the failed predecessors.