Choosing a web development company in Nepal is difficult for a specific reason: the work is invisible until it is finished, and by then you have already paid for it. Portfolios can be borrowed. Testimonials can be written by anyone. The gap between an agency that will serve you well and one that will not is mostly not visible on their website.

What follows is a twelve-point checklist for vetting a Nepali web agency, the questions worth asking before you sign, and the warning signs worth walking away from. We are an agency ourselves, which means some of this is uncomfortable to publish — particularly the sections on code ownership and what “fixed price” actually means.

The 12-point vetting checklist

Score each item. An agency that clears nine or more is worth talking to seriously. Fewer than six and you are taking a risk you probably cannot see yet.

# Check What good looks like
1 Live, verifiable work URLs you can visit, not screenshots
2 Contactable references Past clients willing to take a call
3 Written scope Inclusions and exclusions, itemised
4 Code and hosting ownership Transfers to you, stated in writing
5 Named team You know who does the work
6 Discovery before quoting They ask questions before naming a price
7 Realistic timeline Includes your review cycles
8 Defined change process How scope changes are priced
9 Post-launch terms Warranty period and support, in writing
10 Technical foundations Performance, SEO and accessibility included
11 Communication rhythm Agreed cadence and single point of contact
12 Honest disagreement They push back on bad requests

1. Live, verifiable work

Ask for URLs, not images. Then visit them. Is the site still live? Does it load quickly? Does the agency’s claimed contribution match what you see? A portfolio of static screenshots is not evidence — it is a design exercise.

2. Contactable references

Ask to speak to two past clients, ideally one whose project went badly. Every agency has one. How they describe it tells you more than any case study. An agency that cannot produce a single reference after several years of trading is telling you something.

3. A written scope with exclusions

Any proposal can list what is included. The useful document lists what is not: content writing, photography, third-party licence fees, data migration, training, post-launch changes. Disputes almost always happen in the space between what you assumed and what they quoted.

4. Code and hosting ownership

This is the clause most clients never think to check, and the one that hurts most. Ask directly: on final payment, do I own the source code, the design files and the hosting account?

Some agencies retain the hosting account so you cannot leave. Some build on proprietary systems you cannot export. Some simply never hand over repository access. None of that is necessarily malicious — but you will discover it at the worst possible moment, which is the day you decide to work with someone else. Get the answer in writing. The same applies to your domain and hosting accounts, which should be registered in your name, not your developer’s.

5. A named team

Who will actually build this? Agencies that win work with senior staff and deliver it with juniors are common everywhere, Nepal included. You do not need CVs. You do need to know whether the person in the meeting is the person doing the work.

6. Discovery before quoting

An agency that quotes a fixed price for a website before understanding your content, integrations and approval process is guessing. The guess is usually padded to cover the risk, which means you pay for their uncertainty — or it is not padded, and the shortfall gets recovered by cutting the parts you cannot see.

Good agencies ask irritating questions first. That is the process working.

7. A timeline that includes you

Most projects are delayed by the client, not the developer — waiting on content, feedback or a decision. A timeline that assumes instant approval at every stage is fiction. Ask what happens to the schedule if you take a week to review, because you will.

8. A defined change process

You will change your mind. The question is what happens when you do. Good agencies have a stated process: change requests are scoped, priced and approved before work starts. Bad ones either absorb changes silently until the relationship sours, or bill for them without warning.

9. Post-launch terms in writing

What happens if something breaks two weeks after launch? Is there a warranty period covering defects? What is the response time for a site that is down versus a typo? Ambiguity here reliably becomes an argument.

10. Technical foundations included, not upsold

Mobile responsiveness, reasonable load speed, semantic HTML, proper metadata, a working sitemap and basic accessibility are not premium features. They are what building a website means in 2026. An agency selling them as add-ons is telling you their baseline is below standard.

11. A communication rhythm

Agree the cadence before you start: a weekly update, a single point of contact, one channel of record. Projects rarely fail loudly. They fail through three weeks of silence, during which assumptions drift apart.

12. Willingness to disagree with you

The most underrated signal. If you ask for something that will hurt your project — a carousel on the homepage, six typefaces, a feature nobody will use — does the agency say so, or take the money?

An agency that agrees to everything is not being accommodating. They are treating your project as a transaction. You want the one that says “we can build that, but here is why it will not do what you are hoping for.”

Questions to ask before you sign

  • Who owns the code, designs and hosting when we are done?
  • What is explicitly excluded from this quote?
  • What has to happen on our side to hit this timeline?
  • How are changes priced once we start?
  • What is the warranty period after launch?
  • Who is actually building this, and are they available for the whole project?
  • Can I speak to a client whose project did not go smoothly?
  • What would you push back on in what I have described?

The last one is the most revealing. If a prospective agency has no reservations at all about your brief, they have not read it closely enough.

Fixed price, retainer, or time and materials?

Model Works when Risk you carry
Fixed price Scope is genuinely well defined Changes become expensive; corners get cut invisibly
Time and materials Scope will evolve, you can steer weekly Budget drifts without discipline
Monthly retainer Ongoing work after launch Paying for capacity you may not use

Fixed price feels safest and often is not. It transfers all uncertainty to the agency, who must price for the worst case. If the project turns out simpler, you overpaid. If it turns out harder, the shortfall gets recovered somewhere — usually in testing, accessibility, or the parts of quality you cannot inspect.

For a well-defined brochure site, fixed price is sensible. For anything where the requirements will genuinely be discovered along the way, a staged approach — fixed price for discovery and design, then a properly informed estimate for the build — protects both sides.

How to check a portfolio is real

Three checks take about five minutes each.

Visit the live site. If the URL is missing, dead, or redirects somewhere unrelated, ask why. Sites do come down legitimately. The explanation should be immediate and specific.

Check who built it. Many sites credit their developer in the footer or in the page source. A quick look at the HTML often reveals the theme or framework used, which tells you whether the work was custom or a customised purchase.

Test it like a user. Load it on your phone on mobile data. Submit the contact form. Does it work? Is it fast? An agency showcasing a site that takes eight seconds to load on mobile either did not notice or did not care, and both answers matter.

Warning signs

  • A price before any questions. They are quoting a template, not your project.
  • Guaranteed Google rankings. Nobody controls Google’s systems. This claim is a reliable marker for other exaggerations.
  • Pressure to decide immediately. Real agencies have pipelines, not closing deadlines.
  • No written contract. Non-negotiable, regardless of how well the conversation went.
  • Full payment upfront. Staged payments against milestones protect both parties.
  • Vagueness about ownership. The single strongest signal to walk away.
  • A portfolio you cannot verify. Screenshots without URLs.
  • Quotes far below everyone else. Something is being left out. Ask what.

On that last point: if four agencies quote around NPR 100,000 and one quotes NPR 25,000, the cheap quote is not a bargain, it is a different scope. Ask the cheap agency what they have excluded, and ask the others what the cheap one has probably missed.

Frequently asked questions

How much should I expect to pay a web development company in Nepal?

A business website generally costs NPR 60,000 to 150,000, e-commerce NPR 150,000 to 500,000, and custom applications NPR 200,000 upward. We break the drivers down in detail in how much a website costs in Nepal.

Should I hire a freelancer or an agency?

A freelancer is cheaper and often excellent, but carries single-point risk: if they become unavailable mid-project, you have no continuity. An agency costs more and buys redundancy, defined process and someone to escalate to. For a simple site a good freelancer is fine. For anything your business depends on, the redundancy matters.

Is it safe to hire a Nepali agency if I am overseas?

Yes, and many do. Insist on a written contract specifying IP assignment, staged payments, and an agreed communication cadence that accounts for the timezone difference. Nepal runs at UTC+5:45, which gives good overlap with Australia, the Gulf and most of Asia, and a workable morning window with Europe.

What if I already have a website and just need it improved?

Ask for an audit before anyone proposes a rebuild. Many sites have specific fixable problems — speed, structure, a broken conversion path — that do not warrant starting over. An agency that recommends a full rebuild without auditing what exists is selling, not advising.

How long should the selection process take?

Two to four weeks is reasonable: brief three or four agencies with the same document, compare like-for-like, take reference calls. Rushing this is how people end up with the cheapest quote by default.

What if the relationship goes wrong mid-project?

This is exactly why ownership and staged payments matter. If you own the code and have paid only for completed milestones, you can take the work elsewhere. If you have paid everything upfront and the agency holds the hosting account, you have very little leverage. Structure for this at the start, when nobody expects to need it.

Where to go from here

Run the twelve points against any agency you are considering, including us. If the answers are specific and written down, you are probably in good hands. If they are vague and verbal, that vagueness is the product.

If you want to see how we scope work, our services pages set out what is included, and our case studies are live platforms you can look at. When you are ready, tell us what you are building — and ask us all twelve questions.

Share this article