Website ROI is simple arithmetic wrapped around three awkward questions. What is an enquiry worth? What did the website really cost? And how much of the result would have happened anyway? Answer those well and the formula takes a minute. Answer them badly and any website can be made to look like a triumph or a waste.
Below you’ll find the formula, a worked example, and the step that’s easiest to skip: recording a baseline before a rebuild, so you can prove the return rather than argue about it.
The short answer
Website ROI = (value generated − total cost) ÷ total cost, shown as a percentage.
- Value generated, for a business that sells through enquiries, is qualified enquiries × close rate × average deal value × gross margin.
- Total cost is the build, content, running costs and your own team’s time, over the same period as the value.
- For a rebuild, count only the extra value above your current baseline.
- Measure over two to three years, because a website’s costs arrive up front and its value arrives over time.
Three decisions that change your website ROI
The formula is fixed; these three choices aren’t. Settle them before you plug in numbers, and write them down.
The period
A website is paid for in its first few months and earns for years. Judge it on twelve months and a good site can look like a loss; on ten years, almost anything looks brilliant. Two to three years spreads the build cost while keeping assumptions believable.
Margin, not revenue
A client who pays 100,000 doesn’t add 100,000 to the business. At a 30% gross margin, the website helped earn 30,000 of gross profit. Counting revenue overstates the value more than threefold at that margin, and it’s the quickest way to produce an ROI your finance team won’t accept.
Total or incremental
With no website today, most of what a new one produces counts. If your current site already brings in eight enquiries a month, those eight were coming anyway: only the uplift belongs in the sum.
How to value a website lead
The value of a website lead is what an average qualified enquiry is worth before anyone has replied to it:
Value of a lead = close rate × average deal value × gross margin
Here’s where each number comes from, and how it usually goes wrong:
| Input | Where to find it | The usual mistake |
|---|---|---|
| Qualified enquiries | GA4 key events, checked against your inbox or CRM | Counting spam, job applicants and suppliers |
| Close rate | Your CRM or the sales team’s log | Borrowing the rate from referrals, who behave differently |
| Average deal value | Invoices from clients who came through the website | Quoting your biggest contract, not the typical one |
| Gross margin | Your finance team | Using revenue, or subtracting overheads you’d pay anyway |
Agree what “qualified” means with your sales team: someone who could plausibly buy, from a market you serve, asking about something you sell. If clients usually buy again, use their value over a fixed period, such as two years, rather than an open-ended lifetime figure.
For an online shop, use orders × average order value × margin. If your site’s main job is something else, such as recruitment, value that outcome instead; a clear website strategy tells you which one to count.
GA4 lets you send that value with each enquiry event, so reports show value as well as counts; our guide to tracking website enquiries in GA4 shows how. Compare that value with your website cost per lead: the site’s total cost over a period divided by the qualified enquiries it produced. When cost per lead sits well below the value of a lead, the site is paying its way.
What belongs on the cost side
The build invoice is rarely the only large cost. Count everything the site needs over the same period as the value:
Internal time and ad spend are the easiest to leave out, because nobody invoices the first and the second sits in another budget. Omitting either flatters the result.
For typical price bands and the running costs quotes tend to leave out, see what a website costs. Website pricing models explained shows how fixed-price, hourly and retainer work add up over three years, the view an ROI calculation needs.
A worked example, with hypothetical numbers
The business and every figure below are invented to show the method. Amounts have no currency, because the arithmetic works the same in any.
An engineering services firm is weighing a rebuild. Its current site brings in 8 qualified enquiries a month, a figure it trusts because form, phone and chat tracking has run for a year and matches the sales team’s records.
Step 1: value one lead
With a 20% close rate, an average first contract of 50,000 and a 30% gross margin:
Value of a lead = 0.20 × 50,000 × 0.30 = 3,000
Step 2: add up three years of cost
| Cost | Amount |
|---|---|
| Design and build | 120,000 |
| Content and photography | 30,000 |
| Running costs (20,000 a year × 3) | 60,000 |
| Internal time | 40,000 |
| Total over three years | 250,000 |
Running costs are counted in full to stay conservative, though the firm would pay for hosting and care on its old site too.
Step 3: estimate the uplift
The firm expects the new site to lift qualified enquiries from 8 to 11 a month: 108 extra over three years. In practice the uplift builds gradually, especially from search, so treat it as an average.
- Extra value: 108 × 3,000 = 324,000
- ROI: (324,000 − 250,000) ÷ 250,000 = 29.6% over three years
Step 4: stress-test the assumption that matters
The uplift is the least certain number, so run it at several levels:
| Scenario | Extra enquiries a month | Extra value over three years | ROI |
|---|---|---|---|
| Cautious | 2 | 216,000 | −13.6% |
| Expected | 3 | 324,000 | 29.6% |
| Optimistic | 5 | 540,000 | 116% |
Break-even is 250,000 ÷ 3,000, roughly 83 extra enquiries over three years, or about 2.3 a month. That turns “is a new website worth it?” into a question your sales team can judge: is the current site losing more than two good enquiries a month? Evidence includes visitors who reach the contact page and leave, and service pages that attract traffic but never produce an enquiry; the conversion rate optimisation guide shows how to find both.
Step 5: check the timing
On year one alone, the same project returns about −49%: 210,000 of cost (build, content, internal time and a year of running costs) against 108,000 of extra value. On these averages it pays back early in its third year; later if the uplift builds slowly or deals take months to close.
A clearer site may also raise the close rate, as prospects arrive better informed. Model that as a separate line, so you can see which assumption the result rests on.
Value that’s real but harder to attribute
Some value never appears as a tracked form submission. List it beside the headline ROI rather than blending it in: a figure nobody can check undermines the whole calculation.
- Referrals and calls that start on the site. Add “How did you hear about us?” to forms, and ask it on calls.
- Sales and editing time saved. Prospects sent to detailed pages need fewer basic calls, and edits your team makes itself are developer fees avoided.
- Search visibility that compounds. Organic visits that replace paid clicks are worth roughly what those clicks would cost.
- Risk avoided. A maintained, secure site is less likely to go offline or be hacked, and an outage or clean-up costs enquiries as well as fees.
Put a number on these only where you can defend it, such as hours saved × hourly cost.
Record a baseline before you rebuild
Nobody can claim an ROI without a before and an after, and the before disappears the day the new site launches. If forms, calls and chat clicks aren’t tracked today, set that up first and let it run long enough to show a normal pattern, ideally a full year.
Save these before work starts:
Export them and keep them outside the analytics tool, because tracking often changes during a rebuild. With no reliable history, use the sales team’s records of where clients came from, and say so. A well-run website redesign starts from this baseline, and website KPIs: what to measure, and what to ignore covers which numbers to keep watching afterwards.
Proving the return after launch
Compare the same months year on year to allow for seasonality, and note anything else that changed, such as prices, ad budgets or sales staff, because each competes with the website for credit.
Enquiries move within weeks; revenue moves at the pace of your sales cycle. Check enquiries at three months, pipeline at six, and revenue and ROI at twelve, recording each enquiry’s source in your CRM so you can see which deals started on the website.
Enquiry counts are the start, not the finish. When we redesigned a property developer’s website, buyer enquiries rose 40% in the three months after launch, as the case study sets out. That becomes an ROI only when multiplied through the developer’s own close rate, unit prices and margins, as in the worked example. What such a site needs to win enquiries is covered in what a property developer’s website needs.
Frequently asked questions
What is a good ROI for a website?
There’s no reliable benchmark, because margins, sales cycles and each site’s job vary too widely. A practical test: does it break even under cautious assumptions, and beat the next-best use of the same budget over the same period?
How long does a new website take to pay for itself?
It depends on your sales cycle and how much of the uplift comes from search, which builds gradually. Model it month by month; with the build cost landing up front, year one nearly always looks worse than year two.
Is a new website worth it if most business comes from referrals?
It can be. Referred prospects often look you up before getting in touch, so the site’s job is to confirm the recommendation rather than win the lead from scratch. Ask new clients whether they checked it, and weigh any referrals it may be losing alongside the tracked ROI.
Should ad spend count in website ROI?
Only if the enquiries it buys are counted as value. The cleaner option is to judge organic and direct enquiries against the website’s own costs, and paid campaigns separately on their cost per lead.
What to do next
Start with the value of a lead, because the whole calculation hangs on it. Then find your break-even: the extra qualified enquiries a month a new site must produce to cover its full cost over two or three years. If that looks small next to what your current site is losing, the case for rebuilding is strong. If nothing is tracked, tracking comes first, whatever you decide about the rebuild.
Our website packages and prices show what each build includes and where it starts, so the cost side can use real figures instead of estimates. For a second opinion on your assumptions, bring them to a free 30-minute call.