How to know whether your website pays for itself

Most business owners cannot say what their website earns, only what it cost. This guide shows how to work it out honestly, using information you already have, without buying anything and without pretending to a precision that nobody really has.

Talk to us about your site

To measure website return on investment, add every cost the site incurs in a year, including hosting, maintenance, content, advertising, and the time your team spends on it, then count the revenue from customers whose first contact came through the site. Ask every new customer how they found you and record the answer, because analytics alone cannot see phone calls, referrals, or the long gap between first visit and purchase.

The question almost nobody can answer

Ask a business owner what their website cost, and the answer arrives immediately. Ask what it earned last year, and the room goes quiet. This asymmetry is at the heart of a great deal of bad decision-making about websites, because a cost with no visible return always looks like an expense to be reduced.

The consequence is predictable. The site is neglected, then rebuilt cheaply, then neglected again, and the cycle repeats without anybody ever discovering whether the thing works. Meanwhile the competitor who measures, and therefore improves, quietly takes the customers.

The good news is that a rough but honest measurement is entirely achievable. It does not require expensive software. It requires a decision to ask every customer one question and write the answer down.

First, count all of the costs, including the invisible ones

Most people count only what they paid the designer. That is the smallest part of the true figure over a year, and using it makes the return look better than it is.

  • The build itself, spread over the years you expect to keep the site rather than counted once.
  • Hosting, the domain, security certificates, and any paid plugins or services.
  • Maintenance and fixes, whether paid to somebody outside or absorbed by your own staff.
  • Content: writing, photography, video, and the time spent approving all of it.
  • Any advertising that sends people to the site, which belongs here because it exists to feed the site.
  • Your own time and your team's time, which is real money even when no invoice is issued.

Second, decide honestly what revenue belongs to the site

This is where the argument usually starts, so it is worth being fair rather than flattering. A customer who found you through a friend and then looked at your website before calling was not created by the website. But the website may well have kept them, and a poor one would have lost them.

A practical approach is to count three groups separately. First, customers whose first contact with your business was through the site, which is clean and defensible revenue. Second, customers who came from elsewhere but used the site to check you before deciding, where the site played a supporting role. Third, everybody else.

Use the first group for your main figure, and mention the second as a note. This keeps the number credible when you present it to a partner, an accountant, or yourself in six months.

Third, accept that attribution will never be perfect

Analytics tools see part of the picture and confidently present it as the whole. They cannot see the person who searched on a phone, wrote down your number, and called from the office. They struggle with the customer who first visited months ago. They increasingly lose the source of visitors who arrived through an AI assistant, which reports itself as a direct visit or as nothing at all.

This is not a reason to give up. It is a reason to combine two sources. Use the analytics to see behaviour, which pages people read and where they stop. Use a direct question to your customers to see origin, which no tool can capture reliably.

The direct question is the more valuable of the two, and it is almost free. Add one line to your intake process: how did you first hear about us. Record the answer for every single customer. After a few months you will know more about where your business comes from than most companies with expensive dashboards.

Fourth, work out the payback, not just the ratio

Return on investment is usually expressed as a ratio, which is fine, but for a website the more useful question is how long it takes to pay for itself. A site that costs a certain amount and produces a steady flow of enquiries has a payback period, and that period is the number to compare against other things you might do with the same money.

Be careful with the time frame. A new site rarely performs at its best in the first months, because search engines need time and because the early enquiries have not yet turned into finished work. Judging a website after eight weeks is like judging a garden in February.

Equally, do not spread the judgement over so long that nothing is ever concluded. A year is usually enough to see the shape of the thing, particularly if you have been recording where customers come from since the first day.

Why the cheapest website is frequently the most expensive

The cheapest site wins on the only number anybody measured, which was the invoice. Then it earns nothing, and the cost of earning nothing is invisible and enormous. A year of missed enquiries does not appear on any statement, and so it is never counted.

Consider what one lost customer is worth in your business, then consider how many the site might be losing every month through the leaks described elsewhere in this guide. In most businesses, particularly those where a single customer is worth a meaningful sum, the arithmetic stops being close very quickly.

This is not an argument for spending more for the sake of it. It is an argument for measuring the right thing. The price of the website is a rounding error compared with the revenue it either produces or fails to produce.

Setting up honest measurement in your first month

If you do nothing else after reading this, do these five things. They take a short time to arrange and they will still be useful in five years.

  • Track form submissions and taps on your phone number as recorded events, not as a guess.
  • Use a separate phone number or a call tracking service if calls are a large part of your business.
  • Ask every new customer how they first heard about you, and record it in the same place every time.
  • Keep a simple monthly record: enquiries received, enquiries answered, work won, and value of that work.
  • Note the date of every significant change to the site, so later movements can be explained rather than guessed at.

When a rebuild is justified, and when it is not

A rebuild is justified when the structure itself is the obstacle: when the site cannot be made fast, cannot be edited without a developer, cannot be made to work properly on a phone, or was built in a way that search engines and AI assistants cannot read. In these cases repair work is money poured into a foundation that will not hold.

A rebuild is not justified simply because the design feels dated, or because a new person has arrived with different taste. Redesigns of that kind routinely lose revenue, because they change the pages that were working alongside the ones that were not, and nobody measured which was which beforehand.

The safe approach is to measure first, keep whatever converts, and treat the rebuild as a way to fix specific known problems rather than as a fresh start that discards everything, including the parts that were quietly earning.

How Licheo treats the numbers

We do not publish estimated results, and we do not put invented figures in proposals. Numbers we show come from real sources, from the tools connected to your own site, and where a figure is not yet available we say so plainly rather than filling the gap with something that sounds good.

The same discipline applies to pricing. We do not publish a figure for our own work, because the right scope depends on the business, and we would rather have an honest conversation than anchor you to a number on a page. The measurement we set up for you, on the other hand, is meant to be permanent, simple, and yours.

Every part of this, in detail

Each section below is a full guide of its own, on one specific piece of the problem.

  • The real running cost: Ask a business owner what their website cost and you get one number, usually the build. That figure is a fraction of the real total, and using it makes every calculation about whether the site pays for itself flattering and wrong.
  • Tracking leads: Most businesses can tell you their visitor count to the nearest hundred and cannot tell you how many enquiries the site produced last month. This page fixes that, in an afternoon, with things you already have.
  • Call tracking: In many businesses the majority of good enquiries arrive by telephone, and the telephone is the one channel your analytics cannot see. Call tracking closes that gap, at a price worth understanding first.
  • Ask your customers: There is a piece of research available to every business, costing nothing, that most never carry out. Ask each new customer how they first heard of you, write the answer down, and keep doing it.
  • Payback period: Return on investment is usually expressed as a percentage, which sounds precise and is easy to manipulate. For a website, a plainer question is more useful: how many months of results does it take to cover what it cost?
  • The cheap website: The cheapest site wins on the only number anybody was measuring, which was the invoice. The other costs arrive later, in columns nobody connected to that decision.
  • Budget planning: Most website budgets are set by the first quote that arrives, which means the supplier decides your budget. A better approach is to work out what the site is worth to you first, then see who can do it.
  • Attribution explained: A customer reads an article, forgets you, sees a recommendation, searches your name, and calls. Which of those deserves the credit? Every attribution model answers differently, and none of them are right.
  • Reading analytics: Analytics tools present hundreds of numbers, most of which exist for large organisations with staff to interpret them. For a business owner, five views answer nearly every question worth asking.
  • The monthly check: Websites do not fail suddenly. They drift: a form stops delivering, a price goes stale, a competitor publishes something better. A short monthly habit catches all of it while it is still small.
  • When to spend more: Spending more is sometimes exactly right and sometimes the most expensive way to avoid a harder conversation. The difference is whether you can name the problem the money is solving.
  • Site vs retainer: Both are legitimate uses of the same money, and businesses choose between them mostly on who called them last. There is a better way to decide, using figures you already have.

Questions people ask

How do I calculate website ROI?
Add all yearly costs of the site, including hosting, maintenance, content, advertising that feeds it, and staff time. Then total the revenue from customers whose first contact came through the site. Compare the two, and also work out how many months of that revenue it takes to cover the cost, since payback period is usually more useful than a ratio.
How do I know which customers came from my website?
Ask them. Add one question to your intake process about how they first heard of you, and record the answer for every customer without exception. Analytics cannot see phone calls made from memory, referrals that were checked online first, or visits that began months earlier, so the direct question is the more reliable source.
What if my analytics show most traffic as direct?
That bucket now hides a great deal, including visitors sent by AI assistants and people who typed your name after seeing you elsewhere. Treat a large direct share as a measurement problem rather than a result, and lean more heavily on asking customers directly where they came from.
How long should I wait before judging a new website?
Give conversion changes a few weeks, since they act on visitors who are already arriving. Give search visibility longer, because engines must crawl, index, and build confidence in new pages. A year gives a fair picture, provided you have been recording customer sources from the beginning.
Is a cheap website ever the right choice?
It can be, for a business that genuinely does not sell through its site. For everybody else the invoice is the smallest part of the cost. A year of enquiries that never happened does not appear on any statement, which is exactly why it is so easy to overlook.
Should I count advertising costs against the website?
Yes, when the advertising sends people to the site, because the two work as one system. Separating them lets a poor site hide behind advertising results, and lets advertising take the blame for a page that could not convert the visitors it was sent.
Does Licheo report on this?
Yes, using real data from the tools connected to your site rather than estimates. Where a figure is not yet available we say so rather than presenting a guess, because a number that sounds convincing and turns out to be invented is worse than no number at all.

Find out what your website is really worth

We set up honest measurement, show you where the enquiries genuinely come from, and build the pages that produce more of them.

Talk to us about your site See real client results