"How much will it cost?" is the first question every client asks, and the industry has spent thirty years learning to dodge it. We would rather answer it. Below are the actual ranges we quote in 2026, how those numbers are built up from day rates and hours, and the six cost lines that almost never appear in a proposal but always appear in the final invoice.
The short version
- A credible business website in Pakistan runs PKR 180,000 – 650,000; a marketing site with a CMS and custom design sits at the top of that band.
- A first release of a mobile app with accounts, payments and an admin panel is realistically PKR 1.2M – 4.5M, not the PKR 300,000 you will see advertised.
- Custom business software (ERP, CRM, LMS, booking systems) starts around PKR 2.5M and scales with the number of user roles, not the number of screens.
- Budget 15–20% of build cost per year for maintenance, hosting and third-party services. This is the line most quotes omit.
- Cheap quotes are usually not cheaper — they are smaller. The difference shows up as change requests.
How a software price is actually built
Nearly every honest quote in this industry is the same arithmetic: estimated hours × blended day rate + third-party costs + risk buffer. Everything else — "packages", "starter plans", per-page pricing — is that same arithmetic with the working hidden.
The blended day rate is the average cost of everyone who touches the project: a designer, one or two engineers, a QA tester, and a project manager who is probably splitting time across three accounts. In Pakistan in 2026, established studios bill somewhere between PKR 18,000 and PKR 45,000 per developer-day. Independent freelancers go lower; the difference is not skill so much as what surrounds the code — design review, testing, deployment, documentation, and someone who answers the phone in eighteen months when something breaks.
Where clients get surprised is the hours, not the rate. A screen a client describes in one sentence — "users can see their order history" — is a list view, an empty state, a loading state, pagination, a detail view, a filter, an error state, a permission check, an API endpoint, a database index, and a test. That is two to four developer-days, not two hours.
What websites cost
| Type of site | Typical range (PKR) | Timeline | What you get |
|---|---|---|---|
| Landing page / single-page site | 60,000 – 140,000 | 1–2 weeks | One page, contact form, analytics, mobile-responsive, basic on-page SEO |
| Business website (5–10 pages) | 180,000 – 400,000 | 3–5 weeks | Custom design, CMS so you can edit content, blog, forms, SEO structure, speed pass |
| Corporate site (15+ pages, multilingual) | 400,000 – 900,000 | 6–10 weeks | Above, plus multi-language, careers module, gated resources, structured data |
| Web application (dashboards, portals) | 1,200,000 + | 10 weeks + | Authentication, roles, database design, reporting, integrations, audit trail |
Two variables move a website price more than anything else. The first is whether the design is bespoke or template-based. A well-chosen template with brand colours, real photography and rewritten copy can be genuinely excellent, and it removes roughly 30–40% of the cost. A bespoke design system — grid, type scale, component library, custom illustration — is a different product and priced accordingly.
The second is content. We have watched more site launches slip on missing copy and photography than on any technical blocker. If you do not have final text and images, either budget for a copywriter (PKR 8,000–25,000 per page for good B2B writing) or accept that "launch" means "launch with placeholders".
What mobile apps cost
Mobile is where advertised prices and real prices diverge most sharply. The PKR 250,000 app exists, but it is a single-platform app with no backend, no payments, no admin panel and no analytics — effectively a brochure that happens to be installable.
Here is how we break an app estimate into phases, using a delivery/marketplace app as the example:
| Phase | Developer-days | Cost at PKR 28,000/day |
|---|---|---|
| Discovery, user flows, technical architecture | 6–10 | 168,000 – 280,000 |
| UI/UX design (28–40 screens, both platforms) | 12–18 | 336,000 – 504,000 |
| Backend: database, API, auth, roles | 18–30 | 504,000 – 840,000 |
| Mobile app build (Flutter, both platforms) | 28–45 | 784,000 – 1,260,000 |
| Admin web panel | 10–18 | 280,000 – 504,000 |
| Payments, maps, notifications integration | 6–12 | 168,000 – 336,000 |
| QA, device testing, bug-fix cycles | 10–16 | 280,000 – 448,000 |
| Store submission, release, handover | 3–5 | 84,000 – 140,000 |
That totals roughly PKR 2.6M to 4.3M for a real two-platform product with an admin backend. Choosing a cross-platform framework saves most of the second-platform build cost — see our comparison of Flutter and React Native for when that trade-off is and is not worth making.
The cheapest way to reduce an app budget
Cut features, not quality. Ship one user role instead of three. Ship one payment method instead of four. Ship without the referral system. A smaller product built properly can be extended; a large product built badly has to be rewritten, and rewrites cost more than the original.
What custom business software costs
ERPs, CRMs, hospital systems, school management platforms, logistics dashboards — these are priced by role complexity, and clients consistently under-count their roles. "It is just for our staff" turns into admin, branch manager, accountant, storekeeper, sales agent, auditor and a read-only owner view. Each role adds permissions, screens, reports and test cases.
- Single-role internal tool (one team, one workflow): PKR 900,000 – 1,800,000
- Departmental system (3–4 roles, reporting, exports): PKR 2,500,000 – 5,000,000
- Company-wide platform (6+ roles, integrations, audit trail, SLA): PKR 6,000,000 +
Two things reliably inflate these numbers. Integrations with systems you do not control — a legacy accounting package, a government portal, a bank's API — because you are estimating work against documentation that may be wrong. And data migration, which is never "just import the Excel file". Ten years of inconsistently entered records take real engineering time to clean, deduplicate and validate. Budget 5–12 developer-days for any serious migration.
Six costs that are missing from your quote
These are the lines clients are most often blindsided by. Ask about every one of them before signing anything.
- Hosting and infrastructure. PKR 3,000/month on shared hosting for a brochure site; PKR 12,000–60,000/month for an app with a database, file storage and backups. See our hosting comparison for where the thresholds sit.
- Third-party services. SMS gateways, email delivery, mapping quotas, payment gateway fees (typically 2.5–3.5% per transaction locally), push notification tiers, error monitoring. Individually small, collectively PKR 8,000–50,000/month.
- App store accounts. USD 99/year for Apple, USD 25 one-time for Google. Trivial money, but the developer account must be in your company's name — this is a governance issue, not a cost issue.
- Maintenance. OS updates break things. Dependencies reach end of life. Apple and Google change store requirements annually. Realistic budget: 15–20% of the build cost per year.
- Content and data entry. Someone has to write 40 product descriptions and upload 400 photos. If it is not in the scope, it is your weekend.
- Change requests. Every project has them, and that is healthy — you learn things once real users touch the product. Hold back 10–15% of the total as a change budget rather than pretending the first spec was perfect.
Fixed price or time and materials?
Fixed price transfers risk to the studio, and studios price risk. Expect a 20–30% premium, and expect a scope document that is enforced strictly, because that document is the only thing protecting the studio's margin. Fixed price works well when the scope is genuinely well understood — a marketing site, a defined integration, a redesign.
Time and materials is cheaper per unit of work and far better suited to products that will evolve, but it requires you to stay engaged. If you cannot give feedback weekly, do not choose it.
Our own default for anything ambiguous is a paid discovery phase — usually 5 to 10 days, priced separately — that produces wireframes, a technical approach and a fixed-price quote for the build. Clients occasionally resist paying for discovery. It is the cheapest money in the project: it is the only phase where changing your mind costs nothing but a conversation. If you want to run that process yourself first, our guide to writing a requirements document covers the same ground.
Why the cheapest quote usually costs the most
When two quotes for the same brief differ by 4×, the gap is almost never efficiency. It is one of these:
- Smaller scope. The cheap quote covers one platform, no admin panel, no testing, no deployment. Everything else becomes a change request at a higher effective rate.
- No QA. Testing is 15–25% of a professional build. Removing it does not remove the bugs; it moves the discovery of them to your customers.
- Junior-only staffing. Fine for well-defined work under review. Expensive when nobody senior is making architecture decisions, because those decisions are the ones you cannot cheaply reverse later.
- No handover. No documentation, no repository access, no deployment credentials. The next team's first month is spent understanding what exists. We have quoted rescue work where reading the existing system cost more than rebuilding it.
The most expensive project we ever worked on was a PKR 400,000 app. It was abandoned at 70%, and the rebuild cost PKR 3.1M — of which roughly PKR 600,000 was spent purely on understanding and untangling the first attempt.
A practical way to compare quotes
Ask all three of your shortlisted studios for the same five things. The answers will separate them faster than the price will.
- A breakdown in days per phase, not a single number. A studio that cannot break its own estimate down has not estimated it.
- Who specifically works on it, at what seniority, and how much of their week you get.
- What is explicitly out of scope. The exclusions list is more informative than the inclusions list.
- The IP and repository terms. You should own the code and hold admin access to the repository from day one. Anything else is leverage against you.
- The post-launch arrangement — warranty period, response times, monthly retainer cost — quoted before you sign, not after you depend on them.
We cover the vetting side of this in more depth in how to hire a development team without getting burned.
Getting a real number for your project
Ranges are useful for budgeting and useless for deciding. If you want an actual number, the fastest route is a 30-minute conversation about what you are building, who uses it, and what has to be true on launch day. We will tell you the range, what drives it, and — genuinely — if we think an off-the-shelf product would serve you better than a custom build. That answer has cost us work before, and it is still the right answer.
Send us the brief, or read the rest of the blog first. Both are free.