Jake Shimota woke up one morning to a notification that a stranger had paid him $50.
His first reaction: "this has to be a mistake."
The site was ugly. He hadn't told anyone about it. In his own words, it looked like "Russian hackers are going to steal your identity."
But the payment was real. And five years later, that ugly side project runs a freemium business model that brings in $30,000 to $35,000 a month teaching people Excel.
He has no team and no funding, and half the tech stack is still vanilla JavaScript he copied off Stack Overflow in 2020.
Jake told the whole story on Starter Story this week, and it's one of the cleanest examples I've seen of a boring idea executed patiently. Here's the breakdown.
The numbers: $30K a month from teaching Excel
Excel Exercises (excelexercises.com) does exactly what the name says. It teaches Excel functions and keyboard shortcuts through hands-on exercises in the browser, instead of videos.
Jake pulled up his dashboard on camera. The last 4 weeks:
- A little over $30,000 in revenue
- About 50,000 visitors to the site
- Around 2,500 active paying subscribers
The cost side is almost comical. His Google Cloud bill is usually under $100 a month. SendGrid for email is $35 a month. Google Analytics is free. Then Stripe takes its cut.
That's it. That's the whole P&L.
The timeline matters too. Jake started building this in January 2020 and worked on it nights and weekends for years, always next to a full-time job or grad school. When Starter Story first profiled the business in late 2024, it was doing about $6K a month. He quit his job two years ago, went full time, and has grown it roughly 5x since.
The site now claims over 250,000 registered users and 50+ partner organizations.
Six years from first commit to $30K a month. Nobody clips that part for TikTok.
"Duolingo but for Excel"
The idea came from annoyance, not a brainstorm.
Jake studied finance and worked a suit-and-tie office job. Every new hire at his firm had to sit through a mandatory Excel training: a long collection of videos that he describes as about as fun as it sounds. He zoned out, fell asleep, and retained almost nothing.
Then one day he was doing a Duolingo lesson and it clicked: Duolingo, but for Excel.
He had zero tech background. He taught himself to code in his parents' basement watching YouTube videos and digging through Stack Overflow, built an iPhone app for fun first, then started on Excel Exercises as his first web project.
The first working version took months. The hardest part was building tables that actually look and feel like a real spreadsheet. If the grid doesn't behave like Excel or Google Sheets, users get confused and bounce.
He never set up a separate development environment. The half-broken site was just live on the internet the whole time. Which is exactly why a stranger could stumble in and pay him $50 before he thought it was ready.
A few more sales trickled in over the following weeks. That was his validation: people believed in the concept enough to pay despite the execution.
How the freemium business model works here
Jake's setup is simple: a few free lessons up front, then a hard paywall. No trial period, no credit card upfront. You sample the method for free, then you pay to continue.
His reasoning is the part worth stealing. Most people have never used a hands-on Excel simulator, so they don't understand what they'd be buying. The free lessons are a sample that proves it's different from watching videos. Once someone feels the difference, the paywall converts.
Run the numbers from his dashboard: 50,000 visitors and 2,500 active subscribers. You don't need a viral product for that math to work. You need traffic that arrives already wanting to learn Excel, and a sample that sells the method.
The other thing a hard paywall bought him: a business that runs itself. Everything from signup to billing runs on business automations, so Jake can go on vacation and the site keeps collecting subscriptions.
That's one person serving 2,500 customers, with no support team.
$0 to $10K: a 2,000-word homepage did the heavy lifting
Jake posted on Reddit a couple of times early on. But the real growth came from SEO, and his version of it breaks the usual advice.
He wrote blog articles about individual Excel functions, sure. What actually grew the site was the landing page itself. He wrote 1,500 to 2,000 words directly on his homepage: what this is, how you'll learn Excel, why it beats watching videos.
Marketing gurus told him it was wrong. Too much text on a homepage, bad design, trim it down.
He kept it, because that page got him indexed, and it pulled the highest-intent traffic he had. People searching "learn Excel" landed on a page that answered every objection, then hit lesson one.
→ The page that ranks and the page that sells don't have to be two different pages.
If your product is simple to explain and people already search for the problem, a long, honest landing page can be your entire top of funnel. It's the same pattern I keep seeing in boring SaaS businesses: unsexy niche, searchable pain, one page that does the convincing.
$10K to $30K: paid ads and the boss who signs up 50 seats
Two things took Jake from $10K to $30K a month.
First, paid advertising. He calls it the biggest turning point, not just in growth but in how he thinks. When all your traffic is organic and your costs are under $200 a month, every new subscription is nearly pure margin and you don't have to think hard. Paid ads force you to know your acquisition cost and your customer lifetime value, or you bleed.
Second, a product-led motion he didn't plan. Someone takes his training, likes it, goes to work and tells their boss. The boss emails Jake: "can I sign up 50 people?"
That's how the university and corporate deals started, without a sales team or a single outbound email. The product recruits its own champions, and each champion brings 50 seats instead of one.
If you sell anything a person uses at work, this is the quiet growth lever: make the individual experience good enough that employees pitch it internally for you.
The pricing fix that unlocked real growth
This is my favorite part of the episode, because it's a mistake almost everyone makes with Stripe.
Stripe shows you one LTV number. All subscriptions, all prices, all time, blended into a single average. Jake watched his average LTV to CAC ratio sit at about 3 to 1, exactly where the textbooks say it should be, and assumed everything was fine.
But growth felt slower than it should have been. So he exported every payment into a spreadsheet and segmented monthly versus annual customers.
The blended number was hiding two different businesses:
- Annual customers: really profitable
- Monthly customers: barely breaking even, with a 4 to 5 month payback period
His fix: double the monthly price, leave annual alone. That's when growth "really took off", in his words. Suddenly monthly subscribers paid back their acquisition cost fast, annual became the obvious deal, and he could spend more on ads than competitors who were still staring at a blended average.
I wrote a full piece on SaaS pricing strategy if you want to go deeper, but the one-line version of Jake's lesson: averages lie, segment your LTV before you touch your ad budget.
What you can steal from Jake
His advice for finding your own version of this, straight from the episode:
→ Build from intimate knowledge. Jake sat through the boring Excel training himself, so he knew the user's mindset because he was the user. His counterexample: people who see someone else win with a looksmaxxing app and build a looksmaxxing app they know nothing about.
→ Ship before you're ready. His exact bar: "if you're not embarrassed by the first iteration, then you've waited too long." The ugly live site earned him $50 of validation that a private repo never would have.
→ Treat patience as a strategy. Five years of nights and weekends at a few thousand a month, then quit, then 5x in two years. The curve was never a hockey stick, it just compounded.
None of this required a new idea. Excel is 40 years old. The opportunity was a better way to learn it, found by the one person annoyed enough to build it.
FAQ
What is a freemium business model?
A freemium business model gives users part of the product for free and charges for the rest. Excel Exercises offers a few free lessons, then puts every remaining lesson behind a hard paywall. The free portion exists to prove the product's value, especially when the format (hands-on Excel practice instead of videos) is unfamiliar to buyers.
How much money does Excel Exercises make?
Jake Shimota showed roughly $30,000 in revenue over the previous 4 weeks on the episode, from about 2,500 active subscribers, and says the site makes $30,000 to $35,000 a month. Expenses are under a few hundred dollars a month: GCP under $100, SendGrid at $35, plus Stripe fees.
How did Excel Exercises get its first customers?
A handful of strangers found the unfinished site and paid for it, starting with a single $50 payment. After a couple of early Reddit posts, growth came from SEO, driven mostly by a 1,500 to 2,000 word landing page explaining the method, which ranked and converted the highest-intent searchers.
Does a hard paywall work better than a free trial?
It did here. Jake's users needed to experience the hands-on format before paying, so free sample lessons beat a time-limited trial of everything. The paywall also keeps the business simple to run solo: no trial-expiry emails, no sales calls, just lessons that sell the upgrade.
I interview bootstrapped founders like this every week: real revenue numbers, real playbooks, no fluff. If Jake's story scratched an itch, the podcast goes deeper.
