Website ROI: Does a New Site Actually Pay for Itself?

Most arguments for investing in a website are qualitative. Better user experience, stronger credibility, improved brand perception. All true, none of them answerable when a finance director asks what the return is.

The return is calculable. Not precisely, but closely enough to make a decision with. Here’s the arithmetic, using Dubai numbers.

The basic model

A website generates value through one primary action: enquiry, purchase, booking, or call. The value of the site is that action’s volume multiplied by what each one is worth.

Annual value = visitors × conversion rate × close rate × average deal value

Four inputs. Most businesses know two of them and guess the rest, which is why the conversation stays qualitative.

Visitors comes from analytics. You have this.

Conversion rate is the share of visitors who take the primary action. Critically, this must include phone and WhatsApp, not just form fills. Ruler Analytics found 56% of legal conversions and 53% of professional services conversions happen by phone. If you’re only counting forms in a market where WhatsApp is the default business channel, you’re understating performance by roughly half.

Close rate is the share of enquiries that become customers. Your sales team knows this even if nobody has written it down.

Average deal value should be lifetime value where you have repeat business, not first transaction.

What a normal conversion rate looks like

You cannot judge your own number without a benchmark, and cross-industry averages will mislead you badly.

Ruler Analytics puts median B2B website conversion at 2.9% across fourteen industries and over 100 million tracked data points. By sector, that same metric ranges from 1.1% for B2B SaaS to 7.4% for legal services. Ecommerce typically lands between 1.4% and 3%.

A 2% conversion rate is a crisis for a law firm and a good result for enterprise software. Compare against your category, not the average.

A worked example

Take a Dubai professional services firm. Numbers are illustrative but the ranges are realistic.

Current position

Input	Value
Monthly visitors	2,000
Conversion rate	1.5%
Monthly enquiries	30
Close rate	20%
New clients per month	6
Average client value	AED 25,000
Annual revenue from site	AED 1.8m

After a rebuild that moves conversion to 3%

Still below the 7.4% ceiling for the category, so not an ambitious target.

Input	Value
Monthly enquiries	60
New clients per month	12
Annual revenue from site	AED 3.6m

Incremental annual revenue: AED 1.8m.

Against a build cost somewhere in the tens of thousands, the payback period is measured in weeks rather than years. Even at a fifth of that improvement, the site pays for itself inside a quarter.

That’s the case, stated properly. It also shows exactly where it falls apart.

Where the model breaks

If you can’t measure conversion, none of this works. Most Dubai sites we look at aren’t tracking phone clicks, WhatsApp clicks, or form submissions as distinct events. Without that, every figure above is invented. Instrumenting measurement costs almost nothing and should happen before any rebuild decision, not after.

Traffic and conversion are different problems. A rebuild improves conversion. It does relatively little for traffic in the short term, and can hurt it badly if URLs change without a redirect map — documented losses of around 40% follow structural changes launched without one.

Some sites aren’t conversion instruments. A holding company’s site succeeds if a prospective partner comes away with the right impression. Judging it on form fills pushes you toward decisions that damage it. Not every site should be run through this model, which we’ve covered in effective web design.

The improvement isn’t guaranteed. Doubling conversion is achievable when the current site is genuinely poor. It’s not achievable when the site is already competent and the problem is upstream in positioning or pricing.

The four things that actually move the number

Ranked by evidence, not by how interesting they are.

Speed. Google and Deloitte studied 37 brand websites across over 30 million sessions and found a 0.1-second improvement in mobile load time lifted retail conversions by 8.4%. On Google’s published case studies, Rakuten 24 isolated Core Web Vitals optimisation in an A/B test and recorded a 53.37% increase in revenue per visitor. Around 49% of the mobile web currently fails Core Web Vitals, so this is available to most businesses.

Mobile. Mobile drives 70 to 76% of ecommerce traffic but converts around 42% below desktop. In the UAE, where mobile carries roughly 75.3% of web traffic, that gap is where most of the lost revenue sits. A site that performs well on desktop and adequately on mobile is failing at its main job.

Clarity of the primary action. One action, prominently placed, repeated at the bottom of the page. Around 12% of visitors reach the bottom, and they’re the highest-intent segment. Standard advice to keep the call to action above the fold leaves them with nowhere to go.

Testing. Only 17% of marketers actively A/B test landing pages, and teams that test consistently report average conversion gains between 37 and 49%. Doing this twice a year puts you ahead of nearly everyone in your category.

Each of these is a build decision rather than a design preference, which is why we set performance budgets during design on custom builds rather than optimising afterwards. The reasoning is in web design vs development.

Costs the model usually misses

Content production. Copy, photography, and product data are rarely in the build quote and always on the critical path. Content delays are the primary cause in roughly 60% of projects that overrun.

Internal time. Budget 75 to 150 hours across three to five people for a mid-sized project: briefing, feedback rounds, user acceptance testing, training.

Maintenance. Updates, security, backups, and performance monitoring. Ongoing maintenance is not optional on a WordPress site, and leaving it out of the model understates total cost.

Opportunity cost of delay. If the current site converts at 1.5% and the rebuild would take it to 3%, every month of deliberation costs roughly what the improvement is worth per month. That cuts both ways: it’s an argument for deciding quickly, not for deciding carelessly.

Before you commission anything

Five things to establish. If you can’t answer them, the problem is upstream of design.

1. What is the single primary action, and can you name it without hesitating?
2. What is your current conversion rate, measured including phone and WhatsApp?
3. What is the benchmark for your sector?
4. What is your close rate and average client value?
5. Is the gap between your rate and the benchmark a website problem, or a positioning, pricing, or traffic problem?

If your rate is already at benchmark, a rebuild will produce a better-looking site and roughly the same revenue. That’s a legitimate reason to build one. It just isn’t an ROI argument.

For the ongoing version of this, once a site is live, see how to measure if your website is actually performing.

Frequently asked questions

Want help working out whether your numbers justify a rebuild? Talk to us. If they don’t, we’ll say so.