Skip to content

How to plan your first web application without wasting time or budget

The questions to answer before a single line of code — the problem, the people, the core flows, what to leave out, and what happens after launch.

WebJun 4, 20264 min read

A hand sketching app screen wireframes on paper

Most web application projects that run over budget do not fail in the code. They fail in the first two weeks, when nobody wrote down what the system is for, who uses it, and what it does not need to do yet. Everything after that is expensive guessing.

You do not need a hundred-page specification. You need clear answers to a handful of questions — and they are worth answering before you talk to any developer, including us.

1. Write down the one problem

Finish this sentence in plain language: "Today, ___ takes too long / goes wrong / costs too much, because ___." Perhaps orders arrive through five WhatsApp numbers and get lost. Perhaps stock is counted on paper and nobody trusts the totals. Perhaps customers phone the office to ask where their parcel is.

If you cannot finish the sentence, you are not ready to build. If you can, you already have the test for every later decision: does this feature help solve that problem?

2. List the people, not the features

Name every kind of person who will touch the system, and what each of them needs to get done. A typical business application has more of these than people expect:

  • Customers — place an order, check its status, pay.
  • Staff — process orders, update stock, answer questions.
  • Managers — see what is happening today and this month.
  • An administrator — add users, set prices, change settings.

Each role becomes a set of screens and permissions. Discovering a fifth role halfway through the build is one of the most common causes of delay.

3. Map the three flows that matter most

For each role, describe the two or three journeys they will repeat every day, step by step, as they happen now. "Customer finds a product, adds it to the cart, pays with eSewa, gets an SMS." These core flows are what has to be excellent. Everything else can be good enough.

4. Decide what is not in version one

This is where budgets are actually saved. A first version should do the core flows properly and leave out anything that can wait. Good candidates to postpone:

  • A mobile app, when a fast mobile website will do for the first months.
  • Reports beyond the three numbers the owner checks every day.
  • Loyalty points, referral schemes and other growth features.
  • Automating a process that happens once a month.
  • Integrations with systems you are not sure you will keep.

Ship the smallest thing that solves the problem, use it for a month, then decide what comes next. Real usage will change your priorities more than any planning session.

5. Check whether you should build at all

Not every problem needs custom software. An online shop with standard needs may be better served by an existing platform. Accounting, payroll and inventory are well covered by off-the-shelf systems and open-source ERPs. Custom development earns its cost when your process is genuinely different from the template — when the way you work is part of your advantage. A good partner will tell you when you do not need them.

6. Plan for how Nepal actually works

Software built here has a few requirements that are easy to forget until late in the project:

  • Payments through the gateways your customers use — eSewa, Khalti, Fonepay — and often cash on delivery alongside them.
  • Dates in Bikram Sambat as well as AD, especially for invoices, reports and anything your accountant will read.
  • Nepali (Devanagari) text that displays and searches correctly, if any of your users write in it.
  • SMS or WhatsApp notifications, since email is not where most customers look.
  • Pages that stay usable on a modest Android phone with a weak connection.

7. Budget for after launch

Launch is the start of the system's life, not the end of the project. Hosting, domain and SSL renewals, security updates, backups, bug fixes, and the small changes you will want once real people use it all carry ongoing cost. A common rule of thumb is to set aside a meaningful share of the build cost each year for maintenance. Agree in advance who does this work and how it is charged, so it is not a surprise in month three.

A one-page brief to take to any developer

  • The problem, in one sentence.
  • Every role, and what each needs to do.
  • The core flows, step by step.
  • What is deliberately left out of version one.
  • Payments, languages, dates and notifications you need.
  • Your budget range and the date it needs to be live.

With that page in hand, you will get quotes you can compare, a build that stays on track, and a first version that does its job. If you would like help writing it, we do this with clients before every project — and it costs far less than discovering the answers halfway through.

All software articles

Build your next digital product with Nandi today

Let our experienced team design, build and maintain software that saves time, delights users and grows your business.

Get Started