Websites / Insight

What Makes a Business Website Convert?

Most business websites do not fail because they look bad. They fail because the visitor cannot tell what happens next. Here is the structure we use instead.

Devbarat Prasad Patel Founder, iSoftro Published 6 November 2026 7 min read

A business website has one job that outranks the rest: help the right visitor take the next step. Everything else — the palette, the animation, the stock photography — is in service of that. When a site does not convert, the cause is usually structural rather than cosmetic, and it is fixable without a redesign.

The visit is a decision, not a tour

People rarely read a business website from top to bottom. They arrive with a question, scan for an answer, and leave if the answer is not visible. On a phone, that scan happens in a few thumb movements. This is why the order of information matters more than the amount of it.

We structure every page around a single decision. On a service page, the decision is "is this the kind of team that can solve my problem?" On a case study, it is "did they actually do this before?" On the contact page, it is "what do I send them?" When a page tries to support three decisions at once, the visitor does the work of deciding what matters, and most will not bother.

The first screen has to answer three questions

Above the fold, a visitor is asking: what is this, who is it for, and what do I do next? A hero that answers all three in plain language does more for conversion than any illustration.

  • What is this? A specific noun, not a slogan. "Website and software development for businesses in Nepal" tells someone they are in the right place. "Digital excellence, reimagined" does not.
  • Who is it for? Naming the audience filters out the wrong enquiries, which is as valuable as attracting the right ones. A school looking for admissions software should recognise itself immediately.
  • What next? One primary action, repeated as the page continues. A second, quieter action is fine — usually WhatsApp or a phone call — but two equally loud buttons split attention.

We test this by asking someone outside the project to look at the first screen for five seconds and say what the business does. If they hesitate, the copy is doing too little work.

Proof beats persuasion

Claims are cheap. A visitor who has never heard of you is looking for evidence, and the most useful evidence is specific: a named project, a screenshot of the actual system, a sentence about what changed for the client. That is why our case studies show the starting point, the build and the outcome rather than a paragraph of adjectives.

Proof also has a shape. A school deciding on a software partner wants to see another school. A retailer wants to see a store. Placing the relevant case study next to the relevant service reduces the mental leap the visitor has to make. This is the reasoning behind the related-work blocks on our service pages: the goal is not to fill space, it is to answer "have you done this before?" before the visitor has to ask.

Testimonials work when they are attributable and specific. "Great service, highly recommended" is noise. "They moved our stock records off three spreadsheets and the dispatch team stopped calling the office every morning" is evidence. If you cannot get a full name, a role and organisation is better than an anonymous quote.

Forms and contact paths that do not punish people

The most common conversion leak we find is not the form itself but the distance to it. If the only way to reach a business is a form that asks for a budget range and a timeline before anyone has spoken, many visitors will leave and try a competitor with a phone number.

We recommend offering the path that matches the visitor's urgency:

  • A short form for people who want to write a brief at 11pm.
  • WhatsApp for people who want a quick exchange, which in Nepal is the default for many buyers.
  • A phone number for people who want to talk now.
  • An email address for people who need a record.

Then reduce the form to what you actually need to reply. A name, a way to contact them, and a sentence about the problem is enough for a first reply. Every extra required field is a small tax on the enquiry.

Confirmation matters too. After submission, the visitor should know the message arrived, who will reply, and roughly when. Our own contact form stores each enquiry and sends a notification, and the page states plainly where the details go. That sentence removes a real hesitation: people are wary of handing over a phone number to a site that never explains what happens next.

Speed is a conversion feature

Most visitors in Nepal arrive on a mid-range Android phone over a mobile connection, often while doing something else. A page that waits on a large hero video or twelve web fonts loses people before the copy is readable. We treat performance as part of the message: a fast site suggests a competent business.

In practice that means server-rendered pages where possible, compressed and correctly sized images, intrinsic width and height on every image so the layout does not jump, and a short list of third-party scripts. It also means testing on a throttled connection rather than office Wi-Fi. A page that feels instant on a laptop can feel broken on a phone in a busy shop.

Images are usually the biggest win. A photograph that is 4000 pixels wide and displayed at 600 wastes bandwidth for no visible benefit. Resizing and compressing it often removes more page weight than every script on the site combined.

Measure the step, not the visit

Traffic without a defined action is a vanity number. Before launch we agree what a successful visit looks like: a submitted form, a WhatsApp click, a phone call, a download. Then the measurement is simple — did that action happen, and from which pages?

For most small business sites, three numbers are enough: how many people reached the site, how many took the action, and which pages they were on before they did. If analytics cannot answer those three questions, it is collecting data nobody will use. We keep the setup small enough that someone can check it in a minute a week.

What we deliberately leave out

Conversion work is as much about removal as addition. A carousel that nobody clicks, a chat widget that loads a third-party script before the headline, an animated counter that shows a number the visitor cannot verify — each one adds weight and noise without helping the decision. We ask what a section proves. If the answer is "it fills the page", it goes.

The same applies to navigation. A menu with nine top-level items asks the visitor to choose before they understand the offer. Five is usually enough, and the most valuable page should be reachable in one click from the hero, not buried under a dropdown. On mobile, the menu should open instantly and close when a link is chosen; a drawer that traps the page behind an overlay is a conversion problem disguised as a design choice.

A checklist we actually use

  1. Does the first screen say what the business does, who it serves and what to do next?
  2. Is there one primary action, repeated down the page?
  3. Is the relevant proof within one scroll of the claim?
  4. Can a visitor make contact in the way they prefer, in two taps or fewer?
  5. Does the form ask only for what is needed to reply?
  6. Does the page load quickly on a mid-range phone?
  7. Is the next step obvious on a 360-pixel-wide screen?

None of these require a bigger budget. They require deciding what the page is for and removing what gets in the way. If you want a second opinion on an existing site, send us the link through the contact page and we will tell you where the leaks are before you spend anything on a rebuild.