Ask it of your own business before you read on. Not how many enquiries you received — how many people who became clients arrived by way of the website.
There are three kinds of answer. A number with a source behind it. A guess. And I don't know. The one worth paying attention to is the answer that feels like the first and is actually the second.
Two things get conflated here, and separating them is most of the work.
The first is that enquiries received is not the same as enquiries the website produced. An inbox tells you that a message arrived. It does not tell you where the person had been, what they read, or which page they were on when they decided to write. A referral who checked the site before calling and a stranger who found it by searching arrive in the same inbox looking identical.
The second is that having analytics installed is not the same as counting an enquiry. Analytics arrives counting page views, because page views are the same on every website in the world. A submitted enquiry is specific to your site, so nothing counts it unless somebody sat down and defined it as a thing to count. That definition is a discrete piece of work. It either happened or it did not, and it is checkable in an afternoon.
Here is how to check. Submit your own form — it is your website, you are allowed. Then find out whether anything was recorded at the moment the server accepted the submission. Not when the button was clicked, which is a different event and a much weaker one, because a click that fails validation still counts as a click. If the only record of a successful enquiry is an email landing in an inbox, then the count lives in a human being's memory of a busy month.
This is worth being precise about, because it is not a bad number. A bad number is informative. Twelve enquiries last month and four the month before tells you something happened and gives you somewhere to start looking.
No number is different in kind. It means every change made to the website is unfalsifiable. You rewrote the homepage. Did it help? There is no before to compare against, so the answer is whatever the person who rewrote it says it is, and the person who rewrote it is the least reliable available witness. The same applies to money spent on advertising that lands on those pages, to a new service page, to a rebuild. All of it becomes a matter of opinion held confidently.
So the sentence to hold on to is: you cannot improve what you cannot count.
Now the limits, because they matter and are usually left out.
Counting does not make anyone call. It is instrumentation, not improvement, and anyone presenting it as a growth measure is selling you something. What it produces is a before, so that an after means something.
And even then, a count before a change and a count after it is a comparison across time rather than a controlled test. The seasons are not controlled. Your referral pattern is not controlled. A quiet August is a quiet August. A comparison across time is worth having and it is not proof, and I would rather say so here than have you discover it later.
What I would not do is start with a redesign. A redesign of an uncounted website produces a new uncounted website, and a strong feeling that it is better. The first thing worth building is the event — the definition of what an enquiry is, fired reliably at the moment one actually happens, end to end. It is cheap relative to everything else on the list, and until it exists, everything else on the list is unmeasurable.
The Search-to-Consultation Diagnostic is where this gets decided for one specific business — the site read as a stranger, service line by service line, and the path written down. What it covers and what it costs is on the services page.