The most common question we hear from a business owner sounds like this: “The site is up, people visit it, but orders are few. Does it even work?” And almost always there’s nothing to answer with, because no one measured anything. Learning how to calculate your website conversion rate is the cheapest way to stop guessing and start seeing whether a page brings clients or just looks nice. This article is jargon-free and to the point: what to turn on, which numbers to watch, and what to do with them.
Why “lots of visitors” isn’t a result
Picture a shop on a busy street. A thousand people walked past the door today — but how many came in? How many reached the counter? How many bought anything? “A thousand walked past” sounds pleasant, but the rent isn’t paid by foot traffic on the street; it’s paid by whoever reaches the till.
A website works the same way. Visitor count is your “people on the street.” It’s the loudest number and, on its own, the least useful. Twenty people, three of whom left an inquiry, are worth more to most businesses than a thousand random views with no contact at all.
So the first habit to drop is judging a site by feel and by traffic. Inaction here costs real money: you invest in ads, in content, in the build itself — and you don’t know which step works and which one drains the budget. Without measurement you don’t optimize a site, you just decorate it.
What conversion is and how to calculate it
Conversion is the share of visitors who took the action you want. The formula is so simple you don’t need to memorize it:
conversion = (target actions ÷ visitors) × 100%
An example. Over a week 500 people visited the site and 15 left an inquiry. Conversion to inquiry = 15 ÷ 500 × 100% = 3%. That’s it. This is the number worth watching regularly — it doesn’t depend on whether traffic is “a lot” or “a little,” and it lets you compare week to week honestly.
The thing people confuse: the target action isn’t necessarily a purchase. For an online store, yes — an order. For a service, it’s an inquiry, a call, a message in a messenger, a completed brief form. For a business-card site the goal might even be a tap on the phone number. First decide what “the client made a move toward you” means for your business — and count exactly that.
Which numbers actually matter to an owner
Analytics can show hundreds of metrics. An owner needs a handful of them; the rest is distraction. Here’s the short set that genuinely answers “does the site work or not.”
Conversion to inquiry/order. The headline number. It dropped — something broke or is scaring people off; it rose — you did something right and should repeat it.
Traffic sources. Where people come from: search, ads, social, direct links. This shows which channel actually brings people who leave inquiries, not just people who glance and leave.
Behavior on key pages. How far people get and where they close the tab. If people bail on the checkout or the form en masse, the problem isn’t traffic — it’s that specific step.
Actions, not minutes. “Average time on site” sounds impressive but says almost nothing: a person could leave a tab open and go make coffee. Watch actions — a click, a form, a call — not seconds.
Leave the rest — scroll depth, click maps, dozens of secondary metrics — for the stage when those four basics are already under control. Starting with them means drowning in data instead of making decisions.
How to set up analytics without corrupting the picture
Technically, connecting analytics is a matter of one careful setup, and the tool matters less than making sure the numbers are honest. Three things most often spoil the data.
Don’t count yourself. Your own visits, your staff’s, and those of whoever built the site should be excluded. Otherwise a dozen daily checks of your own inflate traffic and deflate conversion — a false picture from day one.
Set up the actual target actions. Analytics out of the box counts views but doesn’t know that a submitted form or a tapped number matters to you. These events have to be defined separately — otherwise you’ll see visitors but not results.
One reliable source of truth. When numbers in three different places disagree, trust in them evaporates and you go right back to gut-feel decisions. One system you trust beats three that argue with each other.
We usually recommend putting analytics in place before launch, together with the site, not “sometime later.” Then you have a baseline from day one, and every later change is judged honestly: did it get better or worse?
What to do with the numbers next
Measuring is half the job. The other half is acting. The routine is simple and repeatable.
Look at conversion regularly — once a week is enough. Spot a drop at a specific step (say, many people open the form but few submit it) — that’s a hint about where the friction is: a form that’s too long, an unclear button, no trust on that screen. Change one thing, wait, watch the same number. Better — keep it; worse — roll it back.
That’s how a site goes from “it’s kind of there” to a tool you can steer. Not through guarantees and nice words, but through small verified steps, each one visible in the numbers. And then “does it even work?” turns into “what do we improve next?”
At LPF we’ve worked with Ukrainian small and medium businesses since 2018, running our own infrastructure and ongoing support — and we treat analytics as part of the job, not an add-on. If you want to finally see your site’s real numbers, start with a short brief conversation: we’ll work out what conversion means for your business and what to measure first.