How to Measure Website ROI: Metrics That Actually Matter
A client once asked us a simple question: "Is our website actually worth what we're paying for it?" They had a decent amount of traffic, a redesign from two years back that everyone still liked the look of, and a marketing budget that included SEO, some paid ads, and hosting. What they didn't have was an answer. Traffic was "fine." Bounce rate was "fine." Nobody could say whether the site was making or losing them money, because nobody had ever set up a way to find out.
This is more common than most business owners would like to admit. Websites get built, redesigned, and maintained based on gut feeling (it looks more professional now, competitors have nicer sites, the old one felt dated) without a clear mechanism for knowing if any of it pays off. Measuring website ROI isn't complicated in principle, but it requires deciding, in advance, what "working" actually means for your business. Most sites never get that definition, which is why "is our website working" turns out to be a much harder question than it sounds.
Why vanity metrics don't answer the question
Traffic, pageviews, and average time on site are the numbers most dashboards put front and center, and they're the easiest to feel good or bad about. More visitors feels like progress. A longer average session feels like engagement. Neither one tells you whether the website is generating revenue.
A site can pull in thousands of monthly visitors from blog content that ranks well for topics loosely related to the business, with almost none of those visitors ever becoming a lead or a sale. A site can also have modest traffic and be enormously profitable, because the people who land on it are exactly the right people and the path to contacting the business is short and obvious. Traffic is an input, not an outcome. Treating it as the scoreboard is how businesses end up "winning" at metrics that have nothing to do with revenue.
Time on site has the same problem in reverse. A long average session can mean genuine interest, or it can mean visitors are confused, hunting for information that should have been on the page in three seconds, or stuck trying to find a phone number. Without connecting these numbers to what happens next (a form fill, a purchase, a call), they're just activity, not evidence.
None of this means traffic and engagement are useless. They're diagnostic inputs, useful for understanding where visitors come from and how they behave. But they answer a different question than "is this website worth what we're spending on it." That question needs an actual ROI calculation.
How to actually calculate website ROI
The formula is not complicated: revenue or value generated by the website, minus the total cost of running it, divided by that cost. What's hard is being honest about both sides of that equation.
Total cost usually includes more than people initially count:
- The build or redesign cost, amortized over its realistic lifespan (a site that will run for three years shouldn't have its full build cost charged against a single year of ROI)
- Hosting and any recurring platform or plugin fees
- Ongoing maintenance and update costs
- Marketing spend that exists specifically to drive traffic to the site (SEO retainers, paid ads, content production), since this is money spent to make the website valuable, not a separate line item
Revenue or value generated depends on the business, but it should be tied to actions that are directly attributable to the website, not the business's revenue overall. For an ecommerce store this is fairly direct: sales that went through the site. For a business that sells through calls or in-person meetings, it means putting a value on each qualified lead the site produces, based on your actual close rate and average deal size, and counting the leads the site is responsible for.
A basic version looks like this: if a service business closes 20% of qualified leads at an average value of 15,000 SEK, and the website generated 30 qualified leads last quarter at a total cost (build amortization, hosting, maintenance, and the marketing spend driving traffic) of 25,000 SEK, that's roughly 90,000 SEK in expected revenue against 25,000 SEK in cost: a strong return, and a number worth tracking quarter over quarter rather than calculating once and forgetting about it.
The point of doing this calculation isn't precision to the decimal point. It's having a number you can watch move over time, and a way to tell whether a redesign, a new page, or a CRO project actually paid for itself.
The metrics that matter depend on your business type
"Website ROI" isn't one metric. It's a calculation built from the right inputs for your specific business model. Chasing the wrong metric is almost as unhelpful as chasing none at all.
Lead generation businesses
If the website's job is to produce inquiries (for a B2B service, an agency, a consultancy), the two numbers that matter most are cost per lead (total spend divided by number of qualified leads generated) and lead-to-customer conversion rate (what percentage of those leads actually become paying customers). A website that generates a lot of cheap leads that never close isn't performing well, even if cost per lead looks great in isolation. Both numbers need to move together.
Ecommerce
For online stores, the more useful metrics are revenue per visitor (total revenue divided by total sessions, which normalizes for traffic fluctuations), customer acquisition cost (what it costs, in total marketing and platform spend, to acquire one paying customer), and average order value. Watching these three together tells a clearer story than conversion rate alone: a store can have a low conversion rate but a high average order value and still be very profitable, or a high conversion rate on low-margin products that barely covers acquisition cost.
Service businesses
For businesses that sell services booked through a call or consultation (clinics, trades, local professional services), the metric that matters is the number of booked consultations or quote requests the site generates, tracked against how many of those actually convert into paying work. Form fills alone can be misleading here, since plenty of form submissions never turn into a real booking; where possible, track the booking itself, not just the intent to book.
| Business type | Primary metrics to track |
|---|---|
| Lead generation | Cost per lead, lead-to-customer conversion rate |
| Ecommerce | Revenue per visitor, customer acquisition cost, average order value |
| Service business | Booked consultations, quote requests, request-to-booking rate |
Attribution: the part nobody gets perfectly right
Even once you've picked the right metrics, there's a harder problem underneath: a customer rarely converts on their first visit. They might find the site through a Google search, leave, see a social post two weeks later, come back, get a retargeting ad, and finally submit a form a month after their first visit, across two or three different devices. Which of those touchpoints gets credit for the conversion?
First-touch attribution credits whatever brought the visitor in the very first time (useful for understanding what generates initial awareness), but it undervalues everything that happened afterward to actually close the deal. Last-touch attribution credits whatever happened right before conversion (easy to measure), but it can overvalue a retargeting ad or a branded search that only worked because of awareness built earlier. Multi-touch attribution tries to split credit across the whole journey, which is more accurate in theory but requires tracking infrastructure most small and mid-sized businesses don't have, and even then involves judgment calls about how to weight each touchpoint.
The honest answer is that perfect attribution isn't realistic for most small businesses, and chasing it is a good way to spend a lot of time and money without a proportional payoff. A reasonable approximation (last-touch attribution as a baseline, with an awareness that some leads were influenced earlier in the journey) is good enough to make real decisions with. The goal of attribution isn't certainty; it's directional confidence that lets you decide where to invest more and where to pull back.
This is also where ROI measurement connects to a bigger diagnostic question: if the numbers show visitors are arriving but not converting, the ROI calculation tells you there's a problem, but not where it is. That's a separate piece of work: figuring out at which point in the journey value is actually being lost, whether that's unclear messaging, a confusing path to contact, or a slow, frustrating checkout. We've written more on how to diagnose that gap in Why Your Website Gets Traffic But Not Customers, which walks through the common reasons traffic doesn't turn into paying customers.
Set a baseline before you change anything
One of the most common mistakes is redesigning or optimizing a website without ever recording what the "before" numbers actually were. Six months later, there's a vague sense that things feel better, but no way to prove it, because nobody wrote down the cost per lead, conversion rate, or revenue per visitor before the changes went live.
Before any redesign, CRO project, or significant marketing push, record your current numbers for at least a full business cycle (a month at minimum, a full quarter if your sales cycle is longer or your traffic is seasonal). Note the average cost per lead or customer acquisition cost, the conversion rate at whatever stage matters most for your business, and the total cost of running the site during that period. This baseline is what makes "improvement" a measurable fact rather than a feeling. Without it, every future conversation about whether a change was worth it turns into guesswork and opinion.
A practical approach without a dedicated analytics team
Most small businesses don't have, and don't need, a full analytics team to do this well. What's needed is a small number of numbers, tracked consistently, reviewed on a fixed schedule.
- Pick three to five metrics that map to your business type from the table above, and ignore the rest of the dashboard. More metrics doesn't mean more insight; it usually means more noise and less follow-through.
- Set up basic conversion tracking (form submissions, phone clicks, checkout completions, booking confirmations) inside whatever analytics tool you're already using. This is usually a one-time setup, not an ongoing burden.
- Review the numbers on a fixed cadence, monthly for most businesses, quarterly if traffic and lead volume are low. A recurring 30-minute review is far more useful than an occasional deep dive that only happens when something feels wrong.
- Keep a simple running log (a spreadsheet is enough) of cost, leads or sales, and conversion rate over time, along with notes on what changed (a new page, a pricing update, a redesign). This turns scattered numbers into a timeline you can actually learn from.
- Recalculate ROI after any significant change, comparing against the baseline you set beforehand, rather than assuming a redesign or new campaign worked because it feels more polished.
None of this requires expensive tooling or a specialist hire. It requires deciding, once, which numbers actually matter for your business, and then being disciplined enough to look at them regularly instead of only when something feels off.
If you have traffic and engagement but aren't sure whether your website is actually paying for itself, or you know it isn't and can't pinpoint why, that's exactly the kind of question a structured conversion audit is built to answer. Our conversion rate optimization work starts with getting your baseline metrics right before touching anything on the site. Get in touch through our contact page if you'd like a second set of eyes on your numbers.