AI Peekaboo: Revenue, team and how it was built
AI Peekaboo (AI search visibility tracking (GEO), Portugal and USA, Wyoming company): US$8K a month MRR, January 2026. Evidence grade D: reported, not verified. Team: 2 co-founders, Filipe Lins Duarte (CEO) and John Rice (CTO); no employees named in the sources. Launched September 2025; still running.Added 25 October 2026
- Outcome
- US$8K a month MRR, January 2026
- Evidence
- D reported, not verified
- Launch
- September 2025
- Status
- Still running
- Team
- 2 co-founders, Filipe Lins Duarte (CEO) and John Rice (CTO); no employees named in the sources
- Outside funding
- None found
- Group
- SEO and AI search
- Main channel
- Not stated
What is not verified
The figure of US$8,000 in monthly recurring revenue rests on one source: Filipe Lins Duarte's post on Hacker News of 22 January 2026, written to find a buyer, with no dashboard or document. No payment processor, buyer or journalist confirms it, and no later revenue figure exists. The team size at that date is also known only from the company's pages, which name two co-founders and no employees. A reader could look for a payment-processor profile (for example on TrustMRR), a marketplace listing for the sale, or a later interview with either founder. 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.
AI Peekaboo is a tool for generative engine optimization (GEO), the work of making a brand appear in the answers of AI assistants, in the way that search engine optimization (SEO) makes a page appear in Google's list of links. The customer enters prompts, the questions that a person types into an AI assistant, for example "best review platform for online stores". The software sends each prompt to each chosen AI model with live web search turned on, stores the answer, and computes an AI visibility score: the share of answers that name the brand. It also records the brand's position in the answer, the sentiment (whether the answer describes the brand positively or negatively) and the citations, the web pages that the model names as its sources. The AI models are not only the subject of the product: according to the company's help and home pages, every number in it is computed from answers that the models generate when the product queries them, and Google AI Overviews, which has no API, is read from the real search interface.
The company page names two co-founders, Filipe Lins Duarte (CEO, strategy and sales) and John Rice (CTO, product and technology), who it says have built together since 2024. A free beta ran in May 2025. The about page dates the launch to September 2025, and the terms of service, from Think Outside The Box Ventures, LLC, a Wyoming company, took effect on 2 September 2025. The company lists offices in Lisbon, Portugal, and in Minnesota, USA. On 22 January 2026 the founder posted on Hacker News: "We hit $8K MRR about 5 months after launching." MRR means monthly recurring revenue, the subscription money that comes in each month; US$8,000 a month is about US$96,000 a year if it stays the same. In the same post he wrote that "due to some misalignment in co-founder objectives" they were considering a sale. In October 2026 the about page still names both founders and calls the company independent; no source reports a sale.
The revenue figure rests on one post by the founder, grade D: no payment processor, buyer or journalist has confirmed it, and the post was written to find a buyer. The rest of the page comes from the company's own pages, live and archived, and from two articles that the company itself lists as a partnership announcement and as sponsored content. The page also shows four earlier companies that sold brands a measure of their visibility on platforms that they did not own.
How much revenue does AI Peekaboo 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, January 2026 | US$8,000 | D | Founder's post on Hacker News, 22 January 2026: "We hit $8K MRR about 5 months after launching." One source, written to find a buyer. | [1] |
| Run rate from that figure | about US$96,000 a year | calc | Calculated: US$8,000 times 12. | [1] |
| Time from launch to the figure | 4 calendar months (about 4.7 months, 143 days, from 1 September 2025) | calc | Calculated: September 2025 (about page; terms effective 2 September 2025) to 22 January 2026. The founder wrote "about 5 months". | [1, 2, 10] |
| Subscriptions needed for US$8,000 a month at January 2026 prices | 40 to 160 | calc | Calculated: US$8,000 divided by the brand plan prices of US$200 and US$50 a month. Agency plans on custom prices would lower the count. | [1, 9] |
| Reason given for a possible sale | "misalignment in co-founder objectives" | D | Same post. In October 2026 the about page still names both co-founders. | [1, 2] |
| Founders | Filipe Lins Duarte (CEO) and John Rice (CTO) | C | About page and the archived home pages of September 2025 and January 2026. Both are named in the February 2026 sponsored article. | [2, 8, 9, 12] |
| Launch | September 2025 | C | About page: "Launch day, September 2025." Terms of service effective 2 September 2025. A free beta was live on 22 May 2025. | [2, 7, 10] |
| Company | Think Outside The Box Ventures, LLC, a Wyoming limited liability company | C | Terms of service. Peekaboo is a trade name of this company. The site lists offices in Lisbon and in Minnesota. | [10, 2] |
| Outside funding | none found | C | No source names an investor. The January 2026 post asked for introductions to buyers of small software companies or to funds. | [1, 2] |
| AI models queried | ChatGPT, Gemini, Perplexity, Google AI Overviews, Google AI Mode; Claude API and Microsoft Copilot as add-ons | C | Help page: Gemini with search grounding, Perplexity with its own web search, ChatGPT and the Google surfaces through "dedicated infrastructure", Claude through Anthropic's API (claude-sonnet-5) with web search. | [6] |
| Price unit in 2026 | US$16.67 per 1,000 checks a month, minimum US$29 | C | Help page, last reviewed 14 August 2026. Yearly billing 15% off. | [5] |
| Agency plans at launch | US$100 a month for 2 brands, US$200 a month for 5 brands, daily tracking | C | Archived home page, 4 September 2025. | [8] |
| Claimed users | "3,000+ users" (September 2025 to January 2026); "250+ companies" (October 2026) | C | Company pages. The January 2026 page says "3,000+ brands analyzed", which counts free reports, not paying customers. | [8, 9, 4] |
| Agency partnership | Intellibright, a digital marketing agency in Austin, Texas | C | MarTech360, 19 December 2025, quoting the CEO. The current home page shows Intellibright's white-label client reports. | [11, 3] |
| Payments | Stripe | C | Terms of service and the subscription help page. | [10, 17] |
How to read this table: the grade D value is one founder post that no one else has checked. Grade C values are the company's own pages, live or in the Wayback Machine, and articles that the company lists as partnership news or sponsored content. A "calc" value was computed on this page from the rows above it.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Minimum charge | 24.65 | 29 | the floor for any configuration |
| Default on the pricing page | 25.50 | 30 | 1 brand, 40 prompts, 3 models, every 2 days: 1,800 checks a month |
| Example: daily on all 5 standard models | 85 | 100 | 1 brand, 40 prompts: 6,000 checks a month; white-label reports included from this price |
| Example: 100 prompts | 212.50 | 250 | 1 brand, 100 prompts, 5 models, daily: 15,000 checks a month |
| Large configurations | 600 and up | handed to sales instead of self-serve checkout; custom volume pricing above 150 brands | |
| Claude API and Microsoft Copilot | add-on | priced separately in the quote | |
| Starter (Brand), January 2026 | 50 | 1 brand, 40 prompts, every 2 days, 1,800 data points | |
| Peek (Brand), January 2026 | 100 | 1 brand, 40 prompts, daily, 3,600 data points | |
| Grow (Brand), January 2026 | 200 | 1 brand, 100 prompts, daily, 9,000 data points |
Later in 2026 the fixed plans gave way to a calculator: according to the help page (last reviewed 14 August 2026), US$16.67 per 1,000 "configured checks" a month, where one check is one prompt for one brand answered by one model once, with a minimum of US$29 a month. The pricing page says every plan starts at US$30 a month, which matches its default configuration. Yearly billing takes 15% off. The yearly column and the examples were computed on this page from that rate. White-label reports and share links unlock at US$100 a month, and configurations of US$600 a month or more go to a sales conversation. The January 2026 plans in the last three rows used 3 models; the same coverage under the new rate would cost about US$30, US$60 and US$150 a month (calculated). A white-label lead-generation funnel for agencies is a separate managed service with custom pricing.
When did AI Peekaboo launch, and how did it grow?
From a free beta to a sale offer on Hacker News in eight months, and still run by the same two founders in October 2026. Each entry carries its grade.
- 2024CFilipe Lins Duarte and John Rice start building together, according to the about page. [2]
- 22 May 2025CAn archived page shows a beta, "Completely free during early access", that reports a site's visibility in ChatGPT, Gemini and Claude, and promises "autonomous content agents" that publish posts on Medium and wiki pages. [7]
- 2 September 2025CThe terms of service of Think Outside The Box Ventures, LLC take effect. The about page dates the launch to September 2025. [10, 2]
- 4 September 2025CThe archived home page lists agency plans at US$100 and US$200 a month, claims "3,000+ users", and shows the two co-founders as the team. [8]
- 19 December 2025CIntellibright, an agency in Austin, Texas, announces a partnership to offer AI Peekaboo's visibility data to its clients. [11]
- 14 January 2026CThe archived home page lists brand plans at US$50, US$100 and US$200 a month. [9]
- 22 January 2026DOn Hacker News the founder posts "Selling my AI visibility SaaS – $8K MRR, 5 months in – looking for intros", citing differences in the co-founders' objectives. [1]
- 9 February 2026CForbes Portugal publishes sponsored content (Forbes Lab) in which both founders answer questions about brands in AI answers. [12]
- August 2026CThe help center describes a new price per check, a REST API and MCP access on every plan, and a free open-source content skill that drafts articles from the measured gaps. [5, 16, 3]
- October 2026CThe about page still names both co-founders and calls AI Peekaboo "An independent company". The pricing page claims "250+ companies". [2, 4]
Why did AI Peekaboo work?
Four causes, in the order they mattered. Each one is a reusable lesson.
Timing
The product launched in September 2025, in the months when marketers began to see visitors arrive from ChatGPT and Perplexity and asked what those assistants said about them. The free beta of May 2025 tested the idea before the paid launch. The founder said in January 2026 that monthly recurring revenue reached US$8,000 about five months after launch.
Agencies as the first market
The September 2025 page listed only agency plans with prices: several client brands in one account, white-label reports and unlimited seats. An agency that sells AI visibility to ten clients brings ten brands, and the December 2025 partnership with Intellibright made one agency a reseller of the data to its clients.
Price follows the cost of each answer
Each check is one call to one model, and model calls cost money. The 2026 price, according to the help page, is US$16.67 per 1,000 checks a month, so a customer who tracks more prompts, more models or a higher frequency pays more, and the margin per check stays the same.
Real interfaces, not a guess
The help pages say that the software asks live, web-search-enabled models instead of asking a model what it would probably say. Google AI Overviews, which has no API, is captured from the real search interface. The company presents this as the reason its numbers match what a buyer sees.
How it was built
These details come from the company's own pages, live and archived. They are grade C. [6, 3, 16, 17, 10, 19]
- Every tracked prompt is asked live against each chosen model with web search on. Gemini is queried with search grounding, Perplexity through its own web search, and Claude through Anthropic's API with web search.
- ChatGPT, Google AI Overviews, Google AI Mode and Microsoft Copilot are captured through what the help page calls dedicated infrastructure that reads the consumer interfaces; a fallback method runs when the main path fails.
- Results feed a dashboard, a Looker Studio connector, a REST API and an MCP server, so a customer can ask Claude Code or Cursor about the data.
- A free, open-source agent skill on the founder's GitHub account reads a customer's data through the API and drafts an article for a gap, pausing for approval before it writes.
- Subscriptions are billed through Stripe; the site pages are built with Next.js, as their page source shows.
- Free tools (an AI visibility report, a citation checker, an llms.txt generator) and comparison articles about more than twenty other SEO and AI visibility tools are listed on the site and in its sitemap.
Tools named in the sources
Each one opens its page in the Tools tab, with every story that uses it.
Who tried this before AI Peekaboo, and why did they fail?
Four companies in the Loot Drop database sold a measure of how a person or a brand appeared on platforms that they did not own. The values below are from that database and are not independently verified.
Klout
The promiseA score from 1 to 100 that measured how influential a person or a brand was on social networks.
Why it diedFacebook, Twitter and Instagram built their own analytics, rivals gave more detail, and changes to API access limited the data that Klout could collect.
ThinkUp
The promiseInsights from a person's Twitter and Facebook activity that went beyond counts of likes and retweets.
Why it diedThe platforms improved their own analytics and restricted API access, and larger social media tools offered more than analytics.
SumAll
The promiseOne dashboard that collected a small business's data from its social media and marketing platforms.
Why it diedIt could not set itself apart from larger analytics products and free tools such as Google Data Studio, and customers left over price.
Salorix
The promiseMachine learning that ranked social media conversations, so brands could see the themes and influencers in their market.
Why it diedIts investors withdrew their support at the same time as leaner competitors with newer AI methods arrived.
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
All four measured a brand's presence on platforms owned by someone else, and all four lost when the platforms built their own analytics or closed their data. AI Peekaboo carries the same risk: OpenAI, Google, Anthropic and Microsoft decide what their assistants answer, and the help page shows that ChatGPT and the Google surfaces are read from consumer interfaces, not from an official reporting service. Large SEO suites, such as the Ahrefs and Semrush products that the company's blog reviews, sell AI visibility reports too.
Two things differ. First, the platforms do not sell this measurement themselves yet, so there is no native analytics product to replace it, as there was for Klout and ThinkUp. Second, the company has no investors to satisfy: two founders, per-check pricing that passes the model cost to the customer, and agencies that resell the data. The open question is the founders themselves: the founder wrote in January 2026 that their objectives differed.
Can one person copy AI Peekaboo now?
The method, yes. The market, crowded.
The company's own blog has comparison and alternatives articles about more than twenty other SEO and AI visibility tools, so a new copy meets a crowded market. The method is still open: price each model call so that the customer pays for the coverage they choose, sell to agencies first because each agency brings many brands, and give them reports under their own name.
Where it is still open: Measurement where AI now decides what people see and no official report exists: AI shopping answers, AI answers in maps and local search, and assistants inside apps.
- 1Query the real AI interfaces with web search on, and store every answer and citation as evidence.
- 2Price per unit of model usage, with a low minimum, so that small customers can start and large ones pay for their volume.
- 3Sell to agencies with multi-brand accounts and white-label reports, and sign one agency partner early.
- 4Before you post a revenue figure to sell, connect a payment-processor profile so that a buyer can check it.
What could go wrong for AI Peekaboo?
The risks are the mirror image of the causes above.
What is not verified
The figure of US$8,000 in monthly recurring revenue rests on one source: Filipe Lins Duarte's post on Hacker News of 22 January 2026, written to find a buyer, with no dashboard or document. No payment processor, buyer or journalist confirms it, and no later revenue figure exists. The team size at that date is also known only from the company's pages, which name two co-founders and no employees. A reader could look for a payment-processor profile (for example on TrustMRR), a marketplace listing for the sale, or a later interview with either founder.
The co-founders disagreed
The founder wrote in January 2026 that the two founders' objectives differed and that they were considering a sale. In October 2026 both are still named on the about page and no sale has been reported, but the outcome of that disagreement is not public.
A beta promise of content agents
The free beta page of May 2025 promised "autonomous content agents" that would "publish and refresh content on Medium blog posts wiki pages and your website", with tasks such as "Build wiki page for your product". The paid product from September 2025 does not list this, and a help page of 2026 says Content Briefs "will not write your articles for you". Publishing promotional pages on third-party sites to influence AI answers would be against the rules of many of those sites.
Reddit tasks in the to-do list
The recommendations screen suggests tasks such as "Earn mentions in cited Reddit threads" and "Share customer proof in Facebook groups". The product only suggests them; it does not sell or post comments. How a customer carries out such a task, openly or not, is the customer's responsibility.
Claims on the site are larger than the paying base
The site claimed "3,000+ users" from September 2025, the month of launch, which counts free reports. In October 2026 it claims "250+ companies". Neither number is a count of paying customers.
The models decide the numbers
Every score depends on answers that the AI providers generate, and these answers change from one run to the next. If a provider blocks automated access to its consumer interface or raises its prices, the cost and the reliability of each check change with it.
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 Hacker News, 22 January 2026: Selling my AI visibility SaaS – $8K MRR, 5 months in – looking for intros (post by filipeduarte), read through the Algolia HN APIRead in full (story and text, items 46725725 and 46725726). Only source for the MRR figure, the time since launch and the reason for a sale.
- C AI Peekaboo about page, read 25 October 2026Read in full. Source for the founders, their roles, the launch month, the company name and the independence statement.
- C AI Peekaboo home page, read 25 October 2026Read in full, including the menus and the managed services section. Source for the current features and for the purpose check.
- C AI Peekaboo pricing page, read 25 October 2026Read in full. The calculator loads its prices by script; the page text gives the starting price, the rules and the customer claim.
- C AI Peekaboo help center: Plans and pricing (last reviewed 14 August 2026)Read in full. Source for the price per check, the minimum, the yearly discount and the thresholds used in the pricing table.
- C AI Peekaboo help center: How we get the data (last reviewed 14 August 2026)Read in full. Source for how each AI model is queried.
- C AI Peekaboo home page, Wayback Machine copy of 22 May 2025Read in full. Source for the free beta and the content agents it promised.
- C AI Peekaboo home page, Wayback Machine copy of 4 September 2025Read in full. Source for the launch-month agency prices, the user claim and the team section.
- C AI Peekaboo home page, Wayback Machine copy of 14 January 2026Read in full. Source for the brand plan prices in the month of the revenue figure.
- C AI Peekaboo terms and conditions (effective 2 September 2025)Read the sections on the contracting entity, payment and governing law.
- C MarTech360, 19 December 2025: Intellibright Partners with Peekaboo to Redefine Modern Search MarketingRead in full. A partnership announcement that quotes the CEO; linked from the company's about page as partnership news.
- C Forbes Portugal (Forbes Lab, sponsored content), 9 February 2026: Será que a IA sabe que a minha marca existe?Read in full (in Portuguese). Sponsored content, as the page and the company's about page say. Names both founders.
- C AI Peekaboo: White Label AI Lead Generation for AgenciesRead in full. Source for the managed service listed under "Managed services" in the menu, checked for the purpose gate.
- C AI Peekaboo: To-dos and recommendations feature pageRead the sections on the action plan. Source for the Reddit and Facebook task suggestions.
- C AI Peekaboo help center: What AI Peekaboo does not do (last reviewed 14 August 2026)Read in full. Source for the statement that Content Briefs will not write articles.
- C AI Peekaboo help center: The AI Peekaboo content skill on GitHub (last reviewed 14 August 2026)Read in full. Source for the open-source drafting skill.
- C AI Peekaboo help center: Managing your subscription (last reviewed 14 August 2026)Read the section on cancelling and invoices. Source for the Stripe billing portal.
- C AI Peekaboo sitemap, read 25 October 2026Read the list of pages. Source for the free tools and the comparison articles about other tools.
- C Loot Drop database export, 1,749 failed startupsLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.