Everything that goes into a website that actually earns its keep — what the four tiers of "website" really cost, how to choose between WordPress, no-code and custom, why speed is a revenue problem, and how to be found by search engines and AI assistants alike. Plus exactly how we build: senior engineers, a fixed written quote, and 100% ownership at the end.
Web development is the work of building and running anything that lives at a web address — from a single marketing page through to a full application with accounts, dashboards, and a database behind it. It covers the visible interface, the server-side logic and data, and the performance, security, and findability that decide whether the thing actually works for the business.
AI-assisted coding and automation reduce development time by up to 70%, getting your product to market faster.
Semantic markup, structured data, sitemaps, and blazing load times — built to rank from launch day.
Performance-first architecture with Core Web Vitals optimization, lazy loading, and efficient caching strategies.
Built-in security best practices including OWASP compliance, encryption, and secure authentication from day one.
AI-reviewed code following industry best practices ensures long-term maintainability and easy scaling.
Cloud-native design that grows with your business, from startup MVP to enterprise-scale applications.
"Website" describes at least four different products, and most budget surprises in this industry trace back to a buyer and a builder using the word to mean different things. A landing page and a customer portal are both websites in roughly the way a bicycle and a lorry are both vehicles — the same word, wildly different engineering.
Getting specific about which tier you need is the highest-leverage decision in the whole project, because it determines the cost, the timeline, the technology, and whether what you build will still fit in three years. The tell is simple: the moment a requirement involves someone logging in, something being calculated, or data being stored and retrieved per user, you have left "website" and entered "web application."
One-page site — a single scrolling page that says who you are and asks for one action. The right first presence for a new business; from about $500 (₹25,000), live in roughly a week.
Multi-page marketing site — several pages, a CMS or blog your team can edit, enquiry flows, SEO foundations, analytics. Scoped per project, typically live in 2–4 weeks.
Custom web application — accounts, dashboards, business logic, a real database, integrations. This is software, not a brochure; roughly $3,000–$6,000 (₹1.5–2.5 lakh) for a focused first version.
Platform / multi-tenant SaaS — many isolated customer organisations, billing, admin tooling, compliance. A different discipline again, covered in our custom SaaS guide; from about $8,000+.
E-commerce — cuts across the tiers: catalogues, variants, tax and shipping rules, order management, refunds. Always an order of magnitude beyond a brochure site.
Price tracks engineering complexity, not page count — which is why "how many pages?" is such a poor opening question. A five-page site with a custom booking flow costs more than a twenty-page site that is one template repeated with different words. The genuine drivers are custom design versus template, dynamic functionality, integrations, who writes the content, e-commerce, performance and accessibility done properly, and multilingual support.
Timelines follow the same logic, with one addition that catches almost everyone: content. A build waiting on copy and photography is not being built. In our experience it is the single most common cause of a slipped launch date, and it is entirely preventable by agreeing who writes what before the project starts rather than in week three.
Budget the running costs from day one too, because they rarely appear in a quote: a domain at roughly $10–15 a year, hosting from a few dollars a month upward, and real maintenance. SSL should be free — Let's Encrypt is the standard and competent hosts automate it, so a line item charging for a certificate deserves a question. The full breakdown, including why the same brief honestly draws quotes from $800 to $25,000, is in how much a website costs in 2026.
The decision comes down to one question: in three years, will this site be describing your business or running part of it? Describing means content, and for content WordPress is frequently the correct, cheapest, fastest answer — we say so to clients regularly, against our own margin. Running means behaviour, and behaviour is where custom earns its cost.
It is worth being precise about the security argument, because it is almost universally misstated. Patchstack's State of WordPress Security in 2026 counted 11,334 new vulnerabilities across the WordPress ecosystem during 2025 — but only six of them were in WordPress core, with 91% in plugins. So the honest conclusion is not "WordPress is insecure"; core is remarkably well maintained. It is that on WordPress your plugin list is your attack surface, and it deserves the same scrutiny as any other dependency. The full comparison, including page builders and the three-year cost view, is in WordPress vs custom-coded.
No-code tools occupy a third position worth taking seriously: excellent for validating an idea, running an internal tool, or getting something live this week, and limited exactly where your requirements become specific. The honest test is ownership and ceiling — no-code is faster and cheaper until you hit its limits, at which point the migration cost arrives all at once. We work through that trade-off in custom development vs no-code.
Choose WordPress — content-led site, modest budget, non-technical team publishing regularly, standard requirements a mature plugin already solves. Keep the plugin list lean and patch promptly.
Choose no-code — validating an idea, an internal tool, or a simple workflow that needs to exist this week. Accept the ceiling knowingly.
Choose custom — application behaviour, performance as a competitive requirement, integrations no plugin covers, a design that must be unmistakably yours, or a data posture where every dependency needs justifying.
Because the visitor never sees the thing they left. A site that feels instant on an office connection can feel broken on a mid-range phone on 4G, and that is where a large share of your audience actually is. Google's own research puts abandonment at 53% of mobile visits when a page takes longer than three seconds to load — which means performance work is not polish, it is the difference between a visit and a bounce.
The causes are consistent and fixable: spinner-first loading where the page renders nothing until JavaScript arrives, render-blocking CSS, images shipped at several times their display size, JavaScript bundles far larger than the page needs, and layout that shifts under the reader's thumb as things load in. We broke down each cause and its fix in why your website feels slow on mobile.
There is a second, newer reason to care that most agencies have not caught up with. Google renders JavaScript, but the crawlers behind most AI assistants do not — they fetch your HTML and move on. A client-rendered site can therefore look perfectly fine to Google while being effectively blank to the systems answering your buyers' questions. Server-rendering or prerendering stops being a performance nicety and becomes the entry ticket to AI visibility. It is why every route on this site ships as static HTML, verified with JavaScript disabled, as a release gate rather than an aspiration.
A website nobody finds is an expensive business card. Findability is not a phase bolted on after launch; it is a set of build decisions, and retrofitting them costs multiples of designing them in. The foundations are unglamorous: pages that are genuinely crawlable and indexable, a sane URL structure, real HTML headings, fast loads, structured data where it applies, and a page per question your buyers actually ask.
What has changed since 2024 is that ranking is no longer the only goal. A growing share of buyers ask an assistant and act on the single answer they get back, so the objective becomes being the source that answer cites. Google published its first official guidance on this in May 2026, and it is refreshingly plain: the only technical bar it names is that a page must be indexed and eligible to be shown with a snippet, and it explicitly says you do not need llms.txt files, special AI schema, or content chunking. Content that could only have come from you is what wins.
One trap worth knowing because it is invisible: a `nosnippet` or `max-snippet:0` directive — often added years ago to deter scrapers — removes you from AI Overviews while leaving classic results looking fine. Audit those before spending anything on content. The whole discipline, from eligibility through measurement, is laid out in our AI search guide.
Renderable without JavaScript — the single biggest AI-visibility gate, and the one most React and Vue sites silently fail.
Answer-first pages — state the answer plainly, early, under a question-shaped heading. Engines can only lift what your page actually says.
Original, first-hand material — your data, prices, and measured results. If a language model could have written the page, it has no reason to cite it.
Clean technical foundations — indexable pages, honest headings, structured data, fast loads, no snippet suppression.
Security on the web is mostly discipline rather than heroics, and the same short list covers the overwhelming majority of real incidents: keep dependencies patched, validate and escape everything a user can send you, put authentication and authorisation checks on the server rather than trusting the browser, serve everything over HTTPS, and never let secrets reach client-side code. The OWASP Top 10 has been the working checklist for years for the good reason that it keeps being what actually goes wrong.
If your site collects personal data — and a contact form does — you have legal obligations as well as engineering ones. For an Indian business that means the DPDP Act: consent at the point of collection, using data only for the stated purpose, and honouring deletion requests. Serve European visitors and GDPR applies on top. The good news is that compliance is overwhelmingly an architecture property: designed in it is nearly free, retrofitted it is expensive.
Maintenance is the line most owners try to delete and the one that quietly decides whether the site is an asset or a liability in year three. Dependencies get security patches, browsers change, platforms deprecate. A site nobody has touched in two years has not saved money — it has accumulated vulnerabilities. We scope maintenance honestly up front rather than leaving it off the quote and surprising you later.
Our position is deliberately unfashionable: a business website should be built on boring, proven, widely-adopted technology, because every exotic choice is a tax you pay later in hiring, debugging, and the size of the community that has already solved your problem. The differentiator is your product and your content, never the framework underneath them.
In practice that means React and Next.js with strict TypeScript, Tailwind for styling, PostgreSQL when there is real data, and managed platforms like Supabase and Vercel for the undifferentiated plumbing — authentication, storage, hosting — that you should not be paying senior engineers to reinvent. Marketing sites ship as prerendered static HTML, which is simultaneously the fastest thing to serve, the cheapest thing to host, and the most legible thing to a crawler that does not run JavaScript.
The consequence that actually matters to you is that a stack this mainstream can be taken over by any competent team in the world. That is not incidental — it is the technical half of the no-lock-in promise. Code ownership means very little if the code is written in something only its authors understand.
It starts with a free scoping call: you describe the business problem, we ask the questions that surface the real requirements, and you get a fixed written quote against a written scope. No hourly meter, no estimate that doubles. If the honest answer is that you need a one-page site rather than the platform you came in asking about, we will say so — right-sizing the build is the cheapest advice we give and it is free.
From there you see working software early and often, deployed to a real environment you can click through rather than screenshots in a deck. Questions come to you in batches with a recommendation attached, so decisions cost you minutes rather than meetings. Marketing sites typically go live in 2–4 weeks; a custom application is about a month for a focused first version.
Everything runs on accounts you own from day one — the repository, the hosting, the domain, the analytics — so handover is a standing state rather than an event. At launch you have the source code, the deployment pipeline, and documentation of the decisions that matter. Most clients keep us for iterations on the same fixed-quote terms; some take it in-house, and the codebase is written so a competent team can. Both outcomes are fine by design.
Free scoping call — no obligation, and an honest answer about what you actually need.
Fixed written quote — agreed against a written scope, so overruns are our problem.
Live from week one — a real deployed environment, not screenshots.
You own everything — code, accounts, domain, design files. From day one, not at handover.
The quality of the quotes you receive is almost entirely determined by the quality of the brief you give. You do not need a technical specification — you need one page of plain language: what the site must accomplish, who it is for, what a visitor should do, what it must connect to, who provides the content, and when you need it live. That page is the highest-leverage hour of the entire project, and our guide to writing a brief walks through it.
Then compare like with like. Ask every bidder for a fixed written quote against that scope with exclusions stated, ask whether the design is a template or built for you, and confirm in writing that you own the source code and the domain. Those three questions explain most of the price gap between any two quotes you will receive. If someone will only quote hourly, you are being asked to carry their estimating risk — the fixed-price vs hourly trade-off covers when that is reasonable.
Finally, look at what they have actually shipped and whether it is fast. Open a prospective agency's own portfolio sites on a phone on mobile data. A web developer whose own work is slow is telling you what they will build for you, and it is the cheapest due diligence available. Our checklist for choosing an agency is written to be applied to everyone on your list — including us.
“He understood our vision and translated it into a professional, modern, well-structured design that aligned perfectly with our brand — clear communication, real attention to detail, and high-quality work delivered on time.”
products we've shipped ourselves
We don't only build SaaS for clients — we design, ship, and run our own products, on the App Store, Google Play, and the web.
See our products“They wrote the content and built the whole site — it came out more premium than I imagined, and the entire process was effortless on my side.”
“We needed a Shopify store live fast — they had us selling in about a week. Storefront, checkout, inventory: all clean, all done.”
“They delivered a website that looks stunning and perfectly captures the feel of my brand — every detail executed with precision, on schedule and within budget, and stress-free the whole way.”
In-depth, founder-written guides that go deeper on every decision above — real numbers, sourced facts, and free. Start with the question you are weighing right now.
A one-page site starts from about $500 (₹25,000) and goes live in roughly a week. A multi-page marketing website with a CMS, forms, and SEO foundations is scoped per project and typically ships in 2–4 weeks. A custom web application with accounts and dashboards is about $3,000–$6,000 (₹1.5–2.5 lakh). Every figure is a fixed written quote agreed after a free scoping call, and you own 100% of the code.
About a week for a one-page site, 2–4 weeks for a multi-page marketing site, and roughly a month for a focused first version of a custom web application. The variable that stretches these most is content: a build waiting on copy and images is not being built, so agreeing who writes what before the project starts protects the date more than anything else.
Ask whether, in three years, the site will describe your business or run part of it. Describing means content, and WordPress is often the right, cheapest answer — keep the plugin list lean, since Patchstack found 91% of WordPress-ecosystem vulnerabilities are in plugins while core had only six in 2025. Running means behaviour — logins, dashboards, custom workflows — and that is where a custom build stops being the expensive option.
With us, always — the repository, the hosting and domain accounts, the analytics, and the design files are yours from day one, so handover is a state rather than an event. Confirm this explicitly with any provider: some cheap builds retain the code or register the domain to themselves, which turns leaving into a rebuild.
It has to, and that means more than a layout that resizes. Google's research puts abandonment at 53% of mobile visits when a page takes over three seconds, and phones on mobile data are the realistic test — not a desktop on office WiFi. We build against that budget: prerendered HTML, right-sized images, restrained JavaScript, and no layout shifting under the reader's thumb.
Both are build decisions, not an afterthought. For Google, the foundations are crawlable indexable pages, sane structure, real headings, and speed. For AI assistants there is one additional hard gate most sites fail silently: their crawlers generally do not run JavaScript, so a client-rendered site can be visible to Google and blank to ChatGPT or Perplexity. We ship prerendered static HTML and verify every page renders with JavaScript disabled.
Yes, and it is common. We start by auditing what exists — performance, security, indexability, and whether the current stack can carry where you are going — then tell you honestly whether the right move is improving it or rebuilding it. Sometimes the cheapest genuine fix is a week of performance and SEO work on the site you already have.
A domain (~$10–15/year), hosting (a few dollars a month upward, and very little for a static marketing site), and real maintenance for security patches and dependency updates. SSL should be free — Let's Encrypt is standard and most hosts automate it, so a charge for a certificate is worth questioning.
Tell us what the site needs to achieve. You'll get an honest recommendation — including "you need less than you think" when that's true — and a fixed written quote after a free scoping call. Fast, findable, and yours from day one.
We count page views anonymously — no cookies, no personal data, just which pages help. Our developer tools process all data locally and we never see your content. Decline to opt out of the counts entirely. Learn more