Willo: Revenue, team and how it was built
Willo (AI agents that build and run a small business website, USA): US$6.8K MRR, November 2026. Evidence grade A: disclosed by a company or a filing, or verified by a payment processor. Team: Co-founder Jason Francis and his brother Zach Francis, who presented it together in May 2026; headcount not published. Launched March 2026; still running, revenue falling.Added 7 November 2026
- Outcome
- US$6.8K MRR, November 2026Revenue falling
- Evidence
- A disclosed by a company or a filing, or verified by a payment processor
- Launch
- March 2026
- Status
- Still running
- Team
- Co-founder Jason Francis and his brother Zach Francis, who presented it together in May 2026; headcount not published
- Outside funding
- None found in the sources
- Main channel
- Not stated
Willo sells AI agents, programs that use a large language model to plan steps and then act on their own, such as writing code or sending an email. A customer types a business idea. In about ten minutes, by the home page's own timeline, the agents scan competitors, choose a positioning, build and deploy a website, write a list of 20 first tasks and put up a contact form and a signup form. Paid plans add a backend (a database, user logins and APIs), email sending, payments through Stripe Connect, a Stripe service that lets a platform set up payment accounts for its own customers, and the "Growth Engine", which keeps working between chats and waits for the customer's approval before it publishes or sends anything.
The company is Willo, Inc., and its terms are under Oregon law. Jason Francis is named on the home page as co-founder. A podcast episode of 18 May 2026 presented him and his brother Zach Francis building a business live with the product. The founded date on the TrustMRR profile is 22 March 2026, and the first payment came in April 2026. TrustMRR reads the company's Stripe account through an API key. On 7 November 2026 it showed MRR (monthly recurring revenue, the sum of all active subscriptions for one month) of US$6,823 from 118 active subscriptions, and US$8,803 of revenue in October 2026, the best month so far.
The money figures on this page are grade A, because a payment processor is the source. The same profile shows that the last 30 days brought 18.4% less revenue than the 30 days before, and that MRR fell 12.8% in 30 days. Team size and funding are not published. The page also shows earlier companies that promised to build a website or an app for people who cannot code, and why they failed.
How much revenue does Willo 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 |
|---|---|---|---|---|
| MRR, 7 November 2026 | US$6,823 | A | TrustMRR, metric snapshot, revenue synced 7 November 2026 from Stripe through an API key. A copy of the profile downloaded on 6 November 2026, synced 5 November, showed US$6,747 and 120 subscriptions. | [1] |
| Active subscriptions, 7 November 2026 | 118 | A | TrustMRR, same snapshot. | [1] |
| Average MRR per subscription | about US$58 a month | calc | Calculated: US$6,823 divided by 118 is US$57.82. This lies between the Builder plan (US$49) and the Growth plan (US$99). | [1, 3] |
| Run rate from the MRR | US$81,876 a year | calc | Calculated: US$6,823 times 12. This is the headline figure in US dollars a year. | [1] |
| Revenue in October 2026 | US$8,803 | A | TrustMRR monthly revenue timeline. The best month so far; 14.5% more than September (calculated). | [1] |
| Revenue by month, April to September 2026 | US$153, US$1,551, US$3,356, US$3,573, US$5,592, US$7,689 | A | TrustMRR monthly revenue timeline. Every month higher than the month before. | [1] |
| Revenue in the 30 days to 7 November 2026 | US$7,351 | A | TrustMRR: 18.4% less than the 30 days before. MRR fell 12.8% in the same 30 days. The last two days in the window show US$0.00, and TrustMRR says today's value may be partial. | [1] |
| All-time revenue, 7 November 2026 | US$31,773 | A | TrustMRR. The same as the last 12 months, because the first payment came in April 2026. | [1] |
| Earlier MRR reading by IdeaWave, 24 October 2026 | US$7,469 | A | IdeaWave, a tracker that copies the same Stripe feed, so it is not an independent second source. It shows US$7.5K and stores 7469 in the page data. | [4] |
| Founded date | 22 March 2026 | C | Entered by the founder on the TrustMRR profile; profile text is user-written. | [1] |
| Company | Willo, Inc., terms under the laws of the State of Oregon | C | Terms of Service, last updated 19 October 2026. | [5] |
| Team | co-founder Jason Francis; his brother Zach Francis presented the product with him; headcount not published | C | Home page signature "Jason Francis, Co-Founder, Willo". The podcast page speaks of "the Francis brothers" and their 15 years as entrepreneurs. No source gives a number of employees. | [2, 6] |
| Outside funding | none found | C | No funding round appears in any source read for this page. | [1, 2, 6] |
| AI model provider | Anthropic API | C | Home page FAQ: "Willo accesses AI through the Anthropic API". | [2] |
| Prices | Free, then US$25, US$49, US$99 and US$199 a month | C | Pricing page. Credits are spent only on tasks that produce output. | [3] |
| Affiliate commission | 20% recurring | C | Pricing page. The podcast of May 2026 said it was an affiliate partner, not a paid sponsor. | [3, 6] |
| Time from launch to the November 2026 figure | 8 months | calc | Calculated: March 2026 to November 2026. | [1] |
How to read this table: a grade A value comes from the company's Stripe account as TrustMRR and IdeaWave read it. A grade C value is the company's own statement on its website or in a podcast. A "calc" value was computed on this page from the rows above it.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Free | 0 | 40 credits one time, no backend, does not run on its own | |
| Starter | 25 | 150 credits a month, up to 3 businesses | |
| Builder | 49 | 300 credits a month, up to 6 businesses, custom domains, custom agents | |
| Growth | 99 | 600 credits a month, up to 12 businesses | |
| Scale | 199 | 1,300 credits a month, up to 25 businesses, priority support | |
| Credit packs | 11 to 180 one-time | 50 to 1,000 credits, US$0.22 to US$0.18 per credit |
The pricing page shows monthly prices only; no yearly billing is offered on it, so the yearly column is empty. Credits are a shared pool across all businesses of one account, chat and read-only tools cost no credits, and Builder and higher plans roll unused credits over for one month. Credit packs cost US$11 for 50 credits up to US$180 for 1,000 credits. The Free plan card says 2 businesses, while the comparison table on the same page says 1. Affiliates and agencies earn 20% of revenue, every month, from customers they refer.
When did Willo launch, and how did it grow?
From a founded date in March 2026 to 118 paying subscriptions in eight months. Each entry carries its grade.
- 22 March 2026CFounded date given on the TrustMRR profile. [1]
- April 2026AFirst verified revenue: US$153. [1]
- 14 May 2026CThe first blog posts appear, among them "What is an AI Business Builder Platform", which uses Willo as the example. [8]
- 18 May 2026CThe Edge of Show, episode 482: Zach and Jason Francis build a business live with Willo, with market research, a brand, a landing page, GitHub repositories and Stripe payments in under 10 minutes. [6]
- June to August 2026ARevenue of US$3,356, US$3,573 and US$5,592 a month. [1]
- 19 August 2026CListed on PeerPush, a product directory. [7]
- October 2026ABest month so far: US$8,803. The Terms of Service are updated on 19 October. [1, 5]
- 24 October 2026AIdeaWave records MRR of US$7,469. [4]
- 7 November 2026AMRR of US$6,823 from 118 subscriptions. Revenue in the last 30 days is 18.4% below the 30 days before. [1]
Why did Willo work?
Four causes, in the order they mattered. Each one is a reusable lesson.
Bundling
The home page lists website, hosting, SEO, analytics, customer forms and logins, a database and payments "all in one place". A small business owner who would otherwise pay a website builder, a host, an email tool and a developer pays one subscription. The comparison pages in the site footer name the products it replaces: ChatGPT, Claude, Lovable, Shopify, freelancers and agencies.
Credits for work that ships
Chat is free on every plan. Credits are spent only when a task produces output, such as a page, a deploy or an email. This ties the price to the model cost of real work, and the shared credit pool across up to 25 businesses sells the higher plans to people who run several small sites.
Agents that keep working, with approval
On paid plans the Growth Engine keeps proposing and doing tasks between chats, on a schedule the customer approves. The home page says that changes to the site, every email send and the decisions that shape the business wait for the customer's yes, and that every action is logged; its FAQ also names a "Decision Register".
A free start
The Free plan gives 40 credits, a website and hosting with no credit card. The research, brand, website and roadmap are made before the customer pays, so the first paid step comes after the customer has seen a working result.
How it was built
These details come from the company's home page, pricing page and the page source of the home page. They are grade C. [2, 3, 1]
- The agents use Anthropic's models through the Anthropic API, as the home page FAQ says.
- Each customer's website is deployed automatically from a GitHub repository and hosted by Willo; the pricing table calls this "Auto-deployed website (GitHub + hosting)".
- Customer data is separated by a tenant ID and Row-Level Security policies in the database, so a query from one business cannot return another business's rows (home page FAQ).
- Payments for the customer's own business run through Stripe Connect, and Willo says it takes no platform fee on them.
- Builder and higher plans let customers connect their own coding agents, Claude Code, Cursor or Codex, over MCP (Model Context Protocol, a standard way for an AI agent to call outside tools).
- Willo's own subscriptions are paid through Stripe, which TrustMRR reads to verify the revenue.
Who tried this before Willo, and why did they fail?
Four companies in our database of failed startups promised to build a website, a store or an app for people who cannot code, and failed. The values below are from that database and are not independently verified.
The Grid
The promiseAI that designs your website, so you do not pay a designer.
Why it diedThe AI of that time made clumsy designs, while rivals sold polished templates that were easy to use.
Builder.ai
The promiseCustom apps for about US$20,000, built by an AI-driven assembly line instead of expensive developers.
Why it diedApps sold for US$20,000 cost about US$35,000 to build, and much of the work said to be done by AI was done by offshore developers.
LocalOn
The promiseA website toolkit that small local shops could use with no coding.
Why it diedShopify and Wix offered more, and LocalOn ran out of money before small businesses noticed it.
Dukaan
The promiseAn online store for Indian small merchants in 30 seconds, with no code.
Why it diedMost merchants were too small to pay a software subscription that covered the cost of serving them.
The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.
What changed
The Grid tried to design websites with AI in 2014, before the models could do it well. Builder.ai sold AI-built software while people did much of the work, so each app cost more to make than its price. LocalOn and Dukaan sold no-code websites and stores to small merchants, who paid too little to cover the cost of serving them, against Wix and Shopify.
Two things differ. First, the language models of 2026 write working code, text and email from a plain description, so the agents do the work that Builder.ai paid people to do, and the cost per customer is model usage, which Willo passes on as credits. Second, Willo charges from US$25 a month and lets one account run up to 25 businesses, so its customers are people who build several small online businesses, not only one shop. The risk that killed LocalOn remains: large builders such as Shopify and Wix, and AI builders such as Lovable, sell similar products.
Can one person copy Willo now?
The product, hard. The pricing model, yes.
Copying an agent system that researches, builds, hosts and markets a business needs real engineering, and the field already has funded companies. The parts that a small team can copy are the pricing and the trust design: free chat, credits only for work that ships, one pool of credits across many projects, and an approval queue that keeps the customer in control.
Where it is still open: Narrower versions for one trade or one kind of business, which Willo itself starts with pages for electricians, plumbers, HVAC firms, landscapers and cleaners.
- 1Give the first result away: build the site and the plan before asking for a card.
- 2Charge credits only for work that produces output, so the price follows the model cost.
- 3Put every send and every publish behind the customer's approval, and log each action.
- 4Connect a payment processor to a public revenue tracker from the first month, so the growth can be checked.
What could go wrong for Willo?
The risks are the mirror image of the causes above.
Revenue fell in the last 30 days
Monthly revenue rose every month from April to October 2026, when it reached US$8,803. On 7 November 2026 TrustMRR showed US$7,351 for the last 30 days, 18.4% below the 30 days before, and MRR 12.8% lower than 30 days earlier. Two of those days are still at US$0.00, and TrustMRR says today's value may be partial, so the next sync may change the figure. The sources do not say why revenue fell.
Testimonials that could not be checked
The home page shows video testimonials under the heading "Founders on Willo, in their own words" from eight customers with their business names. Two of them give a first name and a last initial only ("Adam R." of MaxYield and "Bill S." of Kno-Me), and a written quote is signed "Erik Mendelson · BlastLister". This page did not check any of these people or businesses.
Team and funding not published
The home page names Jason Francis as co-founder, and a podcast of 18 May 2026 presented him with his brother Zach Francis and spoke of their 15 years as entrepreneurs and of another product of theirs, Helm SEO. No source gives the number of employees or any funding round. The podcast was an affiliate partner of Willo, as the show said.
Depends on one model provider
The home page FAQ says that "Willo accesses AI through the Anthropic API" and that Willo has no control over Anthropic's data policies. A price rise, a rate limit or a policy change at Anthropic would reach Willo's costs and product directly. The docs list in the sitemap includes a page named "bring your own key", which this page did not read.
Agents act in the customer's name
The agents send email campaigns and run custom tasks; one example on the home page is "New lead comes in — enrich it, draft the reply, book the call". The home page says that every email send waits for the customer's approval and that the signup form sends nothing until the customer confirms. It does not say which data sources the lead enrichment uses.
Competition
Polsia and Buildpad, both on this site, also sell AI agents that start or run a small business for a customer, and Willo's own footer has comparison pages against Polsia, Lovable and Shopify.
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.
- A TrustMRR: Willo, markdown profile, revenue synced 7 November 2026Read the markdown profile, downloaded 7 November 2026. The Verification Sources section names Stripe through an API key. Source for MRR, subscriptions, the monthly revenue timeline and the 30-day changes. A copy downloaded on 6 November 2026 gave the 5 November snapshot. The founded date and the description are founder-written.
- C Willo home page, read 7 November 2026Read in full, including the FAQ answers in the page data. Source for the product steps, the co-founder's name, the Anthropic API, the data separation, Stripe Connect, the approval rules and the testimonials.
- C Willo pricing page, read 7 November 2026Read in full. Source for the plans, the credit rules, the credit packs and the affiliate commission.
- A IdeaWave: Willo revenue, $7.5K MRR, as of 24 October 2026Read in full, downloaded by the pipeline on 6 November 2026. A tracker that copies the same Stripe feed; not an independent second source.
- C Willo Terms of Service, last updated 19 October 2026Read the sections on the parties and the governing law, downloaded by the pipeline on 6 November 2026. Source for Willo, Inc. and Oregon law.
- C The Edge of Show, episode 482, 18 May 2026: Launch a Business in 8 Minutes with Willo AI (CastFox episode page)Read the episode description in full; the audio was not heard. Source for the Francis brothers, the live demo, Helm SEO and the affiliate partnership.
- C PeerPush: Willo, AI teams that build and run your businessRead in full, downloaded by the pipeline on 6 November 2026. Source for the listing date, 19 August 2026.
- C Willo blog indexRead the list of posts and their dates, downloaded by the pipeline on 6 November 2026. The oldest posts are dated 14 May 2026.
- C Database of failed startups, local exportLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.
Run this company? Send corrections or a verified figure to aiolum@livyden.com