Open the developer tools on this site and look at storage. It is empty. No cookies, no tracking pixels, no analytics script, no tag manager, no session recording, no consent banner — because there is nothing to consent to.

This note is about why, including the parts where it costs us.

What we actually know about you

Nothing, unless you place an order.

We do not know how many people visited today, which page they landed on, where they came from, how far they scrolled, or whether they read this sentence. If you order, we have your name, phone number and delivery address, because a courier cannot find you without them. That is the complete list.

Why, honestly

Not primarily principle. The reasoning is closer to risk management.

Data you do not hold cannot leak. We are three people. We do not have a security team, an incident response process, or anyone whose job is to notice a breach at two in the morning. Given that, the strongest guarantee we can offer about visitor data is that we never collected it. Any other promise would depend on us being good at something we are not staffed for.

Our customers are in a medical context. A dentist browsing equipment is not sensitive in itself, but building a profile of who is looking at clinical tooling is the kind of dataset that turns out to matter later, in ways nobody predicted when it was collected. The safest version is not to start.

We would not have used it. This is the part most sites will not admit. Analytics is genuinely valuable to a team with the time to act on it. We do not have that time. Installing tracking we would glance at monthly and act on never would mean taking data from people in exchange for nothing.

What it costs us

  • We are guessing. No conversion rates, no funnel, no idea which page loses people. Every judgement about the site is made without evidence, and some of them are certainly wrong.
  • We cannot run experiments. No A/B tests. Changes are made because they seem better, and we find out only through orders.
  • We do not know what works. When an order arrives we do not know whether the person came from a search, a link, or a conversation. Marketing decisions are made blind.
  • We miss problems. A page that fails on a particular browser could be broken for weeks. Analytics would surface that; we find out when someone tells us.

These are not trivial. A larger company would consider this negligent, and at a larger size they would be right.

What we use instead

Orders, and conversations. Every order comes with a phone call, and the call is where we learn what analytics would otherwise tell us: how someone heard about the device, what they expected it to do, what nearly stopped them ordering.

It is a tiny sample and badly biased — we only hear from people who ordered, never from the ones who left. But it is high quality per data point in a way a scroll-depth chart is not, and at our volume the sample size difference is smaller than it would be for a large business.

The order system is different, and that is deliberate

Orders go to a private dashboard behind real authentication — signed session cookies, per-user password hashing, an edge gate that rejects unauthenticated requests before any file is served. Customer data is protected properly because it exists and it matters.

The distinction we are drawing: collect what is required to do the job, protect it seriously, and collect nothing else. Delivering a device requires an address. Nothing about serving a web page requires knowing who is reading it.

If this changes

It might. If we reach a size where flying blind costs more than it saves, we will likely add something — and if we do, it will be privacy-respecting, self-hosted where possible, without cross-site identifiers, and we will say so here rather than quietly adding a script.

Until then: this page did not tell us you read it, and we are fine with that.