What an enterprise redesign costs in India, and what is actually at risk
An enterprise website redesign in India in 2026 costs Rs. 3,00,000 for a structural refresh, Rs. 8,00,000 to Rs. 15,00,000 for a full re-platform with content migration, and Rs. 25,00,000 upward once CRM, ERP and SSO work is in scope. GST at 18 percent applies on top.
That number is not the one that should keep you awake. If your site brings 30,000 organic sessions and 250 qualified enquiries a month, a migration that loses half your rankings for two quarters costs more than the project fee. Companies buying their first website fear an ugly result. Companies on their third fear a silent one, where the site looks better and the pipeline quietly shrinks for six months. Everything below exists to protect what already works.
| Scope | What is included | Cost, excluding GST | Timeline |
|---|---|---|---|
| Structural refresh | New design system, same CMS, same URLs | Rs. 3,00,000 to Rs. 6,00,000 | 8 to 12 weeks |
| Full redesign, same platform | New URLs, redirect map, partial rewrite | Rs. 5,00,000 to Rs. 9,00,000 | 12 to 16 weeks |
| Redesign with re-platform | New CMS, content migration, schema rebuild | Rs. 8,00,000 to Rs. 15,00,000 | 16 to 24 weeks |
| Redesign with deep integrations | CRM, ERP, SSO and payment flows retested | Rs. 12,00,000 to Rs. 25,00,000 | 22 to 32 weeks |
All bands assume a written scope with fixed deliverables, not an hourly retainer. If a re-platform proposal lands under Rs. 3,00,000, the migration work has been left out and you will pay for it in lost traffic.
The pre-redesign audit: what ranks, what converts, what must not change
The audit is not a formality and it is not free. Budget 2 to 3 weeks and roughly 10 percent of the project value. Its output states which pages earn the money and which behaviour of the current site is load bearing.
Three reports that set the scope
Pull these before a wireframe is drawn.
- Top pages by organic entrance, from Search Console across 16 months, with the queries each page ranks for. On most Indian B2B and manufacturing sites, 20 to 40 URLs carry about 70 percent of organic traffic.
- Top pages by assisted conversion, not last click alone. A specification page or case study rarely gets credit but appears in most winning journeys.
- A complete URL inventory from a crawl plus server logs plus the CMS export. Crawls alone miss orphan pages, PDFs, retired campaign pages and language variants.
The do-not-touch list
From those reports, write one list of URLs, titles, headings and elements that must survive unchanged: your best 30 to 50 URLs, the heading patterns that rank, the internal links feeding them, and any form or phone number printed in offline campaigns. Hand it to the design team as a constraint before they present concepts, which is far cheaper than discovering in month three that a new template dropped 600 words of ranking copy.
URL mapping and the 301 redirect plan
Skipping the redirect map is the single most expensive mistake in enterprise redesigns, and it is common because it is invisible in a design review. Nobody presents a redirect map to a director. It gets pushed into launch week, and the misses surface as a ranking drop in week six.
If your URLs change at all, every old URL needs a decision: redirect to the closest equivalent, keep as is, or retire deliberately. A blanket redirect of everything to the homepage is worse than nothing, because search engines treat it as a soft 404 and wipe out link value you spent years earning.
Building a redirect map that survives launch
- Start from the old URL inventory, not the new sitemap. The new site says where you are going, the old inventory says what you owe.
- Map one to one wherever possible: one old URL, one new URL, one hop. Chains of three or four hops leak speed and confuse crawlers.
- Preserve trailing slash, case and protocol conventions, and test www and non www variants.
- Handle the tail: ad query strings, retired campaign pages, AMP versions, feed URLs and PDF paths.
- Test on staging with an automated crawl of every old URL, and require a report showing status code and final destination per row before sign-off.
Ask your vendor to name the person who owns the redirect map, and to show 50 sample rows at proposal stage. A team that has done enterprise migrations produces it in a day. A team that has not will change the subject.
Content migration: what to prune, what to keep, what to rewrite
Established company sites in India carry 300 to 3,000 pages, and much of it is dead weight: old press releases, duplicate service pages built for a 2019 campaign, three versions of one product uploaded by three regional teams. A redesign is the only realistic moment to clean that up, but pruning without evidence is how traffic disappears.
The prune test
Ask four questions of every page. Did it receive organic entrances in the last 12 months? Does it hold external links? Does sales ever send it to a customer? Is it legally required? If all four answers are no, retire it and return a 410 rather than an irrelevant redirect. If any answer is yes, the page moves or gets consolidated into a stronger page with a redirect.
Keep the words. The most common self-inflicted wound is a new template with less text, because the layout looks cleaner with 150 words than 900. If a page ranks on detail, the template needs somewhere to put that detail, even inside accordions.
Legal pages and accessibility are migration items too
Privacy policy, terms, refund policy, grievance officer details required under Indian IT rules, and GST and CIN details in the footer must move with correct URLs and current dates, not be regenerated from a template. Budget for colour contrast, keyboard navigation, visible focus states, labelled form fields and alt text inside the build, not as a later patch. If you sell to government, banks or global enterprises, treat WCAG 2.1 AA as the default.
Structured data and analytics continuity
Two silent losses happen at launch. The first is structured data. Organisation, product, FAQ, breadcrumb and review markup often lives inside the old theme and never gets rebuilt, so rich results fade within weeks. Inventory the schema types you have now and make them acceptance criteria for the new templates.
The second is measurement. If the container ID changes or events get renamed, your before and after comparison becomes meaningless at the moment you need it. Keep the same analytics property, port every event and conversion with identical names, run old and new tracking in parallel on staging, and annotate the launch date. Verify server side tagging, call tracking numbers and lead attribution. A redesign you cannot measure is one you cannot defend at the next board meeting.
Integrations that must keep working
At this size the website is a front end for systems other departments depend on, and each is a possible launch day outage.
The integration inventory
- CRM and lead routing. Submissions must arrive with the same field names, owners and routing rules. Test against live routing, not a test inbox.
- Marketing automation. Tracking scripts, cookies, embedded forms and list mapping. One renamed field can break nurture flows for months.
- ERP and product data. Price, stock and specification feeds, plus the sync frequency both sides expect.
- SSO and dealer or customer portals. Session handling and redirect URIs need testing with real accounts before launch.
- Payments. If you collect money on the site, revalidate webhooks, callback URLs and refund flows on the new paths. Our note on payment gateway options for Indian businesses covers what to check.
- Search, chat, ticketing and internal tools your teams quietly built. Ask every department head directly, because nobody volunteers this.
Write the inventory into the contract as a test list, with an owner, a test case and a sign-off name per row.
Staging, UAT and sign-off across departments
Enterprise redesigns fail on people more often than on code. Marketing drives the project, sales sees the site two days before launch, and compliance then forces the enquiry form to be rebuilt.
Who signs off on what
Fix this at kickoff with a simple matrix. Marketing signs off on messaging, design and search requirements. Sales signs off on enquiry forms and routing. IT signs off on hosting, security, SSO and data flows. Legal signs off on policy pages, claims and consent. Finance signs off on payment and invoicing. One named person per area, one deadline, and a written rule that silence past the deadline counts as approval.
Run UAT on a password protected staging environment that mirrors production, blocked from indexing by robots rules and HTTP authentication. Give reviewers a structured form, not a WhatsApp group, otherwise you collect 200 unsorted opinions and zero decisions. Cap review at two rounds in the contract and price further rounds as change requests.
Phased launch or big bang, and the rollback plan
A big bang launch, where the whole site is replaced overnight, is right when the URL structure barely changes and the site is under roughly 500 pages. It is faster and cheaper.
A phased launch is safer when you are re-platforming, when integrations are numerous, or when one section such as a dealer portal carries revenue directly. Move the marketing pages first, keep the transactional section on the old platform behind the same domain, then migrate it in a second window. It costs more engineering time, but it limits the blast radius. Teams running web development projects of this size phase the revenue pages last.
The launch week runbook
Insist on a written runbook: a deployment window outside business hours, DNS TTL lowered 48 hours in advance, a full backup of database and files, redirect verification in the first hour, sitemap resubmission, a crawl for broken links and missing schema, and daily rank and lead monitoring for 30 days. Rollback should be one documented action that restores the old site within an hour, tested once on staging. If your vendor cannot describe rollback in two sentences, they do not have one.
Planning a redesign you cannot afford to get wrong?
Send us your current site and we will come back with a free risk read: the pages you cannot afford to lose, the redirect work involved, and a realistic budget.
Timelines, milestones and how to pay
A realistic enterprise redesign takes 4 to 7 months from kickoff to a stable launch. Anyone promising 6 weeks is quoting a template swap or has not read your integration list. The schedule below assumes your team turns around content and approvals within a week.
| Phase | What happens | Duration | Payment |
|---|---|---|---|
| Audit and discovery | Analytics review, URL inventory, do-not-touch list | 2 to 3 weeks | 15 percent |
| Architecture and design | Sitemap, draft redirect map, design system, key templates | 3 to 5 weeks | 25 percent |
| Build and migration | Templates, content move, schema, tracking, integrations | 5 to 9 weeks | 30 percent |
| UAT and testing | Department sign-off, integration tests, accessibility and speed | 2 to 3 weeks | 15 percent |
| Launch and stabilisation | Redirect verification, rank and lead monitoring, fixes | 4 weeks | 15 percent |
Pay against milestones tied to deliverables you can inspect, not against dates. A final 15 percent released 30 days after launch, once redirects and integrations are verified, aligns incentives better than any contract clause. Add GST at 18 percent to each invoice, confirm the vendor GSTIN so you can claim input credit, and sign the statement of work before the first payment leaves.
How to run the vendor selection
Shortlist three vendors, no more, and give all three the same brief with your do-not-touch list and integration inventory. Compare on evidence, not presentation quality: the best deck and the best migration are rarely the same company.
Ask for two references from redesigns of similar size, and ask them one question: what happened to organic traffic in the three months after launch. Ask to see a redacted redirect map from a past project, and ask who writes the code and where they sit, because subcontracting is common.
Then settle the terms that matter. You own the code, design files, content and repository from day one. The repository sits in your organisation account with the vendor as a collaborator. Hosting, domain and analytics accounts stay in your company name, with data residency confirmed in writing. Documentation and a handover session are deliverables, not favours. Change requests are priced from a published rate card. Support carries a written SLA. An exit clause transfers everything inside 15 days if the relationship ends. If a vendor resists code ownership, you already have your answer.
When a targeted fix programme beats a full redesign
Sometimes the honest answer is that you should not redesign at all. If traffic and conversion are healthy, if the complaint is that the site looks dated to internal stakeholders, and if the platform is supported and secure, a full redesign is an expensive way to solve a taste problem while risking your rankings.
A targeted fix programme of Rs. 1,00,000 to Rs. 4,00,000 across three months usually wins. That budget buys a visual refresh of header, hero and typography, a rebuilt enquiry flow, Core Web Vitals work, mobile fixes at 390 pixels, schema, and rewritten top 20 pages. You keep your URLs, rankings and integrations, and you measure each change on its own.
Redesign properly when the platform is unsupported or insecure, when nobody can publish a page without a developer, when it cannot be made fast or accessible without a rebuild, when the information architecture no longer fits the business, or when a merger forces the issue. Those are structural reasons. Looking old, on its own, is not one.
Our team is based in Pune and works with companies across India on redesigns in this range. The projects that go well begin the same way: an audit, a do-not-touch list, and a redirect map agreed before design starts.
Also read: our guides to mobile app development cost in India, ecommerce website development cost in India, and how we scope custom software builds.
