Why we build and run our own products before selling the method

The tagline on this site says that our methods are proven on our own products. That is a claim, so here is what it means in practice: Tovanin builds and operates eleven products of its own, across Israel, Latin America and international markets, and every technique we offer a client was used on one of them first, where it either worked or cost us money. This article is about what that arrangement teaches, and where it has limits.

Client work ends at launch; a product does not

A studio that only builds for clients rarely sees month four. The site is delivered, the invoice is paid, and whatever happens in Search Console afterwards happens to someone else. Owning a product removes that blind spot. When EntrenAU, our exam practice product for Latin America, reached 6,500 visitors in its first 90 days, we saw which pages brought them, which payment step lost them, and which emails they ignored. None of that is visible from the outside, and all of it changes how we build the next site.

What the products taught us, specifically

Each row is a mistake we made or nearly made. The PsychometryBest one is instructive. The site originally sent quiz answers to the browser along with the questions, which is the obvious way to build a quiz. It is also a way to let anyone with a script download the whole question bank. Moving the checking to the server was a real piece of work, and it is now the default on every practice product we build. A client would never have paid for that lesson; we paid for it ourselves.

Why this keeps our advice honest

When we tell a client that a keyword is out of reach for a new domain, or that a second language will double the routing work, or that an AI feature needs a spending cap before it needs a prompt, we are not quoting a best practice. We are describing a bill we have paid. Google's own guidance on helpful content asks whether a page demonstrates first-hand experience; the same test applies to a studio's advice.

It also changes what we refuse. We do not sell link building, because we have never seen it do anything durable for our own domains. We do not build features that a client cannot maintain alone, because eleven products maintained by one person taught us what that takes. And we do not promise a ranking, because Search Console has shown us too often how little of that is in anyone's control. Owners sometimes find that frustrating in the first call. It is usually the reason they come back.

The limits of the arrangement

• Our products are in education, immigration, real estate data, insurance comparison and AI tools. If your business is in a very different field, some lessons transfer and some do not, and we will say which.
• One engineer runs all eleven. That is possible because each is built to be maintained by one person, which is also how we build for clients. It also means we do not take on projects that need a team.
• Our numbers are our numbers. We publish only what we can verify on our own products, and we do not dress them up as industry averages.

How the products shape a client project

Every client site starts from the same foundation as our own products: Next.js on Firebase App Hosting, multilingual routing when the market needs it, analytics and Search Console configured on launch day, and an alert when something breaks. The services page lists the deliverables. The portfolio has the products themselves, each with what was built and one concrete fact. If you would like to know which of those lessons apply to your project, write to us and we will tell you plainly, including the ones that do not.