Offshore software development means hiring a team in another country to build or maintain your software. It is not new, and most of what is written about it is written by agencies selling it — including, to be fair, us. So this is the version we would want to read if we were on the other side of the table: what the models actually are, what the money actually buys, and the specific ways these arrangements fail.
We are a studio in Kathmandu. We build for clients in Nepal and abroad. Take the sales pitch out and the useful parts should still stand on their own.
Three models, and they are not interchangeable
Almost every arrangement is one of three. Choosing the wrong one is the most expensive mistake in this whole process, and it is usually made in the first conversation.
- Project outsourcing — you describe an outcome, agree a price, and receive software. Best when the scope is genuinely knowable up front: a marketing site, a defined integration, a rebuild of something that already exists.
- Dedicated team — you rent engineers full time and direct them yourself. Best when the work is ongoing and the scope will change, which is most product work.
- Staff augmentation — one or two engineers join your existing team and your existing process. Best when you have a working team and a specific gap, not when you have no technical leadership at all.
The common failure is picking project outsourcing for work that is really product development. You write a specification, sign a fixed price, and then reality arrives: the requirement was wrong, the integration is harder than the docs implied, a user tests it and hates it. Now every change is a variation order, and both sides start arguing instead of building.
If you cannot write down what done looks like, you do not want a fixed price. You want a team.
What the rate actually buys
Offshore rates vary enormously, and the number on its own tells you very little. What matters is what is inside it.
- Is the person quoted the person who writes the code, or a layer above them?
- Does the rate include QA, or is testing your job?
- Does it include a project manager, and is that manager shared across six clients?
- Who pays when something ships broken — and is that written down anywhere?
A higher rate that includes testing and a named technical lead is routinely cheaper than a low rate that includes neither. The low rate is not a discount; it is a smaller scope wearing the same word.
Be equally wary of the other direction. If a quote is close to what you would pay at home, you are paying for a sales operation, not for engineering.
Where offshore work actually goes wrong
In our experience the failures are rarely about talent. They are about these four things, in roughly this order.
- No one owns the decisions. The client assumes the agency will make the product calls; the agency assumes the client will. Nobody does, and the work stalls politely for weeks.
- Time zones used as an excuse rather than a schedule. Any gap is workable with a fixed overlap window and written handovers. It becomes fatal when every question costs a day.
- Requirements that live only in conversations. If a decision is not written down, it will be relitigated, and the version that survives will be whoever argued most recently.
- No access to the real thing. Teams kept away from the production data, the users and the support inbox build what was described rather than what is needed.
Notice that none of these are solved by choosing a different country. They are solved by how the engagement is run.
Questions worth asking before you sign
These separate serious partners from order-takers faster than any portfolio does.
- Who exactly is on my team, and are they on other projects at the same time?
- Show me a project that went badly and tell me what you changed afterwards.
- What hours do we overlap, and who is reachable inside them?
- Who owns the code, the repositories and the hosting accounts — on what date?
- What happens if I want to stop in three months?
That last one matters most. A partner confident in the work will answer it plainly. The answer should be that you keep everything and can hand it to anyone, because it was built to be handed over.
On choosing a country
Most comparisons rank destinations by average hourly rate, which is close to useless — the range inside any country is wider than the gap between countries. Three things are worth comparing instead.
- Overlap hours with your working day, because that sets how fast decisions get made.
- Whether you can talk directly to the engineers, or only to an account manager.
- Whether the work you are buying is what they actually do, rather than something they are willing to try.
Nepal, where we are, overlaps comfortably with Australia, Singapore and the Gulf, and gives a few usable hours with the UK. It overlaps badly with the US west coast, and we would say so before taking that work.
Is offshore right for you at all?
Sometimes not, and it is worth saying so.
If the work is small, genuinely one-off and needs constant in-person input, hiring locally or using a freelancer will be simpler. If you have no technical person at all and no appetite to become one, a dedicated team will drift unless you buy technical leadership with it. And if the project is already late and already troubled, adding an offshore team will not rescue it — the problems travel.
Offshore works best when the work is ongoing, the scope will keep moving, and somebody on your side can make decisions and is prepared to make them in writing.
Written by the team at Nandi Solutions, a software studio in Kathmandu, Nepal. We build web and mobile applications, management systems and e-commerce for clients in Nepal and abroad.
All software articles




