Ask five business owners whether their website makes them money. Four will say probably, or that it is hard to tell. The fifth will answer with a number, and it is usually the one who put the most money into it, so they had to work it out.
It is not complicated and it needs no expensive tool. Four numbers and a calculator will do. Most companies do not do it not because it is hard, but because they are afraid of the answer.
A website cannot be run on a feeling, because the feeling follows the most recent job. A big enquiry on Tuesday and the site is suddenly wonderful. A fortnight of nothing and it is useless. Neither has to be true.
Short answer
- Four numbers are enough: visits, enquiries, the ratio between them, and the value of a job. Everything else follows.
- Payback is a simple division: the price of the site divided by the profit from one job tells you how many jobs pay for it.
- Average conversion rates from articles on the internet are useless. The only useful comparison is your site against itself three months ago.
- In trades and services most enquiries arrive by phone, not through a form. Anyone not measuring calls is measuring a minority.
- When the numbers do not add up, the fault is usually in the ratio, not in the traffic. Bringing more people to a site that does not convert is the most expensive way to change nothing.
The four numbers it all rests on
Everything else your tools display is either derived from these or decoration.
| Number | Where you get it | How often to look |
|---|---|---|
| Visits per month | Your analytics | Monthly |
| Enquiries per month | Form, e-mail and phone combined | Monthly |
| Ratio of enquiries to visits | The first two divided | Monthly, but judged quarterly |
| Profit from one job | Your margin, not your revenue | Once a year |
The last row is where this most often falls over. Companies know the revenue of a job but not the profit in it, so they end up calculating payback from a figure three times higher than it should be. If you do not know exactly, estimate low. A payback that still works with a cautious estimate is one you can rely on.
The calculation takes five minutes
A worked example to put your own numbers into. Treat it as a method, not as a claim about your company.
The site cost 60,000 CZK and running it is 5,000 CZK a year. It gets 400 visitors a month and produces 8 enquiries, a ratio of 2 per cent. One in four enquiries becomes a job, so 2 jobs a month. The profit on a job is 12,000 CZK.
So the site brings in 24,000 CZK of profit a month. The first year costs 65,000 CZK, so it pays for itself in under three months and earns for the rest of the year. That is why a website is judged on payback rather than on price.
This is the most important sentence in the article, so here it is another way: improving the ratio is almost always cheaper than bringing in more people. Doubling traffic takes months of work or money spent on advertising. Lifting the ratio from two to three per cent often means rewriting the homepage and shortening a form.
Why average conversion rates are useless
There are dozens of articles online with tables of average conversion rates by industry. It looks like a yardstick for telling whether you are any good. It is not.
The reason is simple: a conversion rate depends on what you count as a conversion and where your people come from. A site selling something for a hundred pounds and a site taking enquiries for building a warehouse have rates that cannot be compared. And two companies in the same trade will differ depending on whether their visitors arrive from search or from advertising.
The only comparison that means anything is your site against itself. This month's ratio against the same month last year, and against the state before your last change. Anyone who has measured for three months knows more about their site than somebody who has read ten tables of averages.
Our conversion rate is 1.4 %, the industry average is supposedly 2.3 %, so we are below average. Nobody knows what counts as a conversion in that average or where its visitors came from.
We are at 1.4 %, three months ago it was 0.9 %. We rewrote the homepage and cut the form from thirteen fields to five. We left everything else alone, so we know what did it.
Phone calls: the enquiries you do not know about
In trades, home services, healthcare and technical supply, most enquiries arrive by phone. The customer finds a number, calls, and it leaves no trace in any measurement. The company then sees three enquiries in its statistics when it actually received fifteen.
The fix does not have to be technical. Three options, in order of effort.
- 1Ask. One sentence at the start of the call, „did you find us online?“, and a tally on a sheet of paper. In a month you have a number you could not get otherwise and it cost nothing.
- 2Measure clicks on the phone number. On mobile people tap the number, and that can be recorded as an event. It is not a count of calls, but it is a good indicator of the trend.
- 3Put a different phone number on the website from the one on your cards and in directories. Forward it to your own and you have exact figures without asking. Worth it where calls are the main channel.
Anyone not doing this is measuring a minority and deciding on it. The most common consequence is a company cutting back or scrapping a website that in reality brings in most of its work, without knowing.
Four traps in the evaluation
All of them lead to a decision that looks sound and is wrong.
Too short a period. A fortnight says nothing, because enquiries do not arrive evenly. Judge by the quarter even if you look monthly.
Last click. Most tools credit a job to the last source somebody arrived from. If a person finds you in search, comes back directly three weeks later and sends an enquiry, it counts as direct traffic and search looks useless.
Seasonality. Comparing November with June at a landscaping company makes no sense. Compare with the same period last year.
Your own visits. If you and three colleagues open the site daily, on low traffic that visibly inflates the figures and depresses the ratio. It is worth excluding them from the measurement.
A rhythm you can keep
The dashboard you open beats the one that looks good. A practical split looks like this.
- Monthly, five minutes: visits, enquiries, ratio. Write them into a spreadsheet, nothing more. The point is to build a run of figures over a year, not to evaluate them on the spot.
- Quarterly, half an hour: compare with the previous quarter and with the same period last year. This is where conclusions get drawn.
- Yearly, an hour: work out the payback and decide whether to keep investing in the site, and in what.
That spreadsheet matters more than the tool you collect the numbers with. A company with twelve rows for the year spots a change immediately. A company that glances at a dashboard occasionally and writes nothing down spots nothing, because it cannot remember how things were before.
When measuring is not worth it
Two cases where it is more honest to leave it than to manufacture numbers.
At fifty visits a month a conversion rate tells you nothing. One extra enquiry moves it by two percentage points and you draw a conclusion that is pure chance. There you count enquiries in absolute numbers and leave the ratio alone.
And the second case: a website whose job is to prove the company exists, so a customer can check you out before calling on a recommendation. Such a site measures no enquiries, because none come through it. Its success shows differently: people did not stop taking you seriously.
How we handle it
Visit and enquiry measurement is part of the site from the company website package onwards, not an extra and not an outside service.
- It needs no cookie banner, because nothing is stored on the visitor's device. The visitor sees the content straight away.
- It sees every visit, not only those that accepted a banner and run no blocker. With third party tools that difference runs into tens of per cent.
- The dashboard links visits to enquiries, so you do not calculate the ratio by hand. You can also see which pages the enquiries come from.
- Taps on the phone number are recorded as an event, so even companies whose customers call have at least an indicator of the trend.
- The dashboard is yours, not ours. You log into the admin next to the enquiries.
And the honest other side: cookieless measurement cannot tell that the same person came back three weeks later, so it will not tell you how many jobs began with an advert a month ago. For most company websites that is a good trade, but anyone building decisions on multi-channel attribution needs something else, and it is fair to say so upfront.
Frequently asked questions
How do I tell whether my website makes money?
You need four numbers: visits a month, enquiries a month, the ratio between them, and the profit on one job. From those you get the monthly contribution and divide the price of the site by it. The result is how many months the site takes to pay for itself. Take profit from your margin, not your revenue, or the payback comes out several times better than it is.
What is a good conversion rate for a company website?
There is no useful answer to that. The rate depends on what you count as a conversion and where your visitors come from, so two companies in the same trade differ by multiples. The only comparison that means anything is your site against itself three months ago, or against the same period last year.
How do I measure enquiries that arrive by phone?
The simplest way is to ask: one sentence at the start of the call and a tally on paper. More precise is measuring taps on the phone number on mobile, which indicates the trend rather than counting calls. Most precise is putting a different number on the website and forwarding it. In trades and services this matters: without it you are measuring a minority of your enquiries.
Should I invest in traffic or in conversion?
Almost always in conversion. Doubling traffic takes months of work or money on advertising, while lifting the ratio from two to three per cent often means rewriting the homepage and shortening a form. Bringing more people to a site that does not convert is the most expensive way to change nothing.
How long should I measure before drawing conclusions?
Look monthly but conclude quarterly. A fortnight says nothing, because enquiries do not arrive evenly. On a site with fewer than a hundred visits a month the conversion rate has no meaning at all, and it is better to track enquiries in absolute numbers.
Do I need Google Analytics to measure this?
No. The four numbers in question can be produced by measurement on your own server, which additionally sees every visit rather than only those with consent and no blocker. You need a third party tool once you build decisions on advertising audiences and multi-channel attribution.
Summary
- Four numbers are enough: visits, enquiries, the ratio between them, and the profit on one job.
- Take profit from your margin, not your revenue, or the payback comes out several times better than it is.
- Improving the ratio is almost always cheaper than bringing in more people.
- Average conversion rates from the internet say nothing. Compare the site with itself.
- In trades and services, not measuring phone calls means measuring a minority of enquiries.
- Twelve rows in a spreadsheet over a year are worth more than a dashboard nobody opens.
Not sure what your site brings in
Tell us what you measure today and what you do not. We will go through it with you and say which numbers are missing and how to get them, even if that means a tally on paper next to the phone. Half an hour, no commitment.