Measuring Website ROI: What a Business Owner Should Actually Track

"The website looks nice, but is it earning?" Fair question. It comes up in almost every renewal conversation we have, and too many web companies answer it with traffic charts and jargon. Measuring website ROI does not need a marketing degree. It needs five numbers, a notebook or spreadsheet, and the honesty to count phone calls and Viber messages as seriously as online orders. Let us set it up the way we do for our own clients.

First, define what a "result" means for your business

ROI is return on investment, and the return depends on what the site is for. An ecommerce store counts orders. A hospital counts appointment bookings. A trekking agency counts itinerary enquiries that become bookings. A restaurant might count table reservations and Maps direction taps. Pick one primary result and, at most, one secondary. If everything is a goal, nothing is measurable.

The five numbers that tell the whole story

• Visitors per month. From Google Analytics. Just the trend, not the vanity.
• Enquiries per month. Forms, WhatsApp and Viber chats started from the site, and calls from tapped tel: links. Yes, count the chats. In Nepal they are usually the majority.
• Customers from those enquiries. Ask every new customer "where did you find us?" for 90 days. Imperfect, but plenty.
• Average customer value. What a typical customer spends with you, including repeat visits if they tend to return.
• Total website cost. Build cost spread over its useful life (three to five years is realistic), plus yearly domain, hosting and maintenance.

A worked example in real NPR

Say a Pokhara guesthouse pays NPR 60,000 for a site, plus NPR 15,000 a year in domain, hosting and upkeep. Over four years that is NPR 120,000 total, or NPR 2,500 a month. The site brings 600 visitors a month and produces 20 enquiries, of which 5 become direct bookings averaging NPR 8,000. That is NPR 40,000 a month of revenue against NPR 2,500 of cost, a sixteen-fold monthly return, and every direct booking also dodges a 15 to 20 percent OTA commission. Run your own numbers this way. Sometimes the answer is embarrassing in a good direction; sometimes it tells you the site needs work, which is also worth knowing.

Count what does not show up in analytics

Some returns resist the spreadsheet but are real. The corporate client who checked your site before taking your call seriously. The job applicant who judged you employable because the site looked alive. The hours saved because the site answers "what is your price" fifty times a day so your staff do not. We suggest logging these as notes rather than forcing fake numbers onto them, but do not pretend they are zero either. A website is partly infrastructure, like a signboard: nobody calculates signboard ROI, yet nobody serious operates without one.

When the ROI is bad, diagnose in this order

Low visitors? That is an SEO and promotion problem: your Facebook posts should link to the site, and your pages need to target what people actually search. Visitors but no enquiries? A conversion problem: weak offers, hidden phone numbers, no chat buttons, slow pages on mobile data. Enquiries but no customers? A sales problem: response time and follow-up, not the website. Most "the website did nothing" stories we investigate turn out to be one of these three, and each has a specific, fixable cause.

Set the measurement up on day one, not year three

The saddest audits are sites that worked for years with nobody counting. Every website WebsNP delivers goes out with analytics installed, enquiry forms wired for tracking, and click-tracked call and chat buttons, so the five numbers above are simply waiting in a dashboard rather than needing archaeology. With build prices fixed in a 24-hour proposal, a free domain and first-year hosting included, the investment side of your ROI equation is known before you spend a rupee. The return side is yours to grow. Now you know how to keep score.