I built four boring finance calculators 鈥?and the numbers explain why tool sites make money
Most "make money online" advice points at SaaS. But a lot of quiet revenue on the web comes from single-purpose calculators with no login, no signup, and no backend.
I built four to test the idea:
- mortgage calculator (principal, interest, taxes, insurance, PMI, extra payments)
- auto loan calculator (price, down payment, trade-in, sales tax, term)
- refinance calculator (current vs new payment, break-even)
- credit card payoff calculator (months, interest, extra payment)
They all run in the browser. No upload, no account.
Why calculators, not another AI app
A calculator has three useful properties:
- The search intent is explicit. Someone typing "mortgage calculator with PMI" is already in the problem.
- The page is cheap to run. Static HTML and JavaScript. No GPU, no per-request cost.
- The commercial value is in the keyword, not the code. Finance and insurance keywords have far higher RPM than generic utility keywords.
The last point is the uncomfortable one. A JSON formatter might get traffic, but the advertiser value is low. A mortgage calculator gets less traffic and far better monetization per visit.
What I actually shipped
The finance hub is here: https://www.mengstone.com/finance/
Each calculator has a long-tail page next to it:
- mortgage with taxes and insurance
- mortgage with PMI
- mortgage extra payments
- 30-year vs 15-year
- auto loan with bad credit
- auto loan with trade-in
- auto loan with taxes
- refinance break-even
- refinance closing costs
The long-tail pages are not filler. Each one answers a narrower question and links back to the calculator. That is the same structure the big calculator sites use.
What is still missing
Distribution and an ad layer.
A calculator with no traffic makes zero. A calculator with traffic and no ad account also makes zero. The code is the easy part; the account, the approval, and the search rankings are the hard part.
I am not going to pretend the pages are ranked or that the calculators earn anything yet. They do not.
The takeaway
If you want to test a tool-site idea, pick a boring query with commercial intent, ship one clean page, add the long-tail variants, and only then think about ads. The interesting engineering is mostly in the calculator math and the page speed 鈥?not in the stack.
