Aayushman Vidhya Society

Aayushman Vidhya Society is a registered society in Jaipur, Rajasthan. It gives away health education, free, in three languages. Startup Savera did the website development for aayushmanvidhyasociety.org. At 112 pages it is the largest site we have shipped, and every one of them can be read in English, Hindi or Sanskrit.

This is a long case study because it is a big build. If you are here to judge whether we can do this for you, the short version is below and everything under it is the evidence.

This page shows the work, and it shows you where to check every claim on the live site yourself.

The client

What the society does

The motto is सर्वत्र आयुष्मान्, सर्वत्र विद्या. Health everywhere, knowledge everywhere. In the homepage screenshot above it sits in the hero in Sanskrit, with the English reading right below it, so a reader in either language meets the same idea before scrolling once.

The work is health literacy. Most people with a health question are served by a rumour, a forwarded video, or a neighbour who means well. The society answers those questions properly, in the language the reader thinks in. Its own framing is ancient wisdom, modern science, universal access.

Why the paperwork matters here

This is a registered society, not an informal group. Registration number COOP/2025/JAIPUR/500734, registered under the Rajasthan Societies Registration Act 1958 with the Registrar of Cooperative Societies in Jaipur. The society publishes all of it on the site instead of keeping it in a drawer. Its work has also been presented at the World Health Summit regional meeting in New Delhi in 2025.

That matters for search as much as for funders. Google is trying to work out whether a health site is a real organisation with a name and an address, or three people and a template. Published registration details are one of the clearest signals you can give it.

Starting a nonprofit yourself? The legal structure comes first and the website second. Our guide to Section 8 company registration in India explains the choice, and we do the filing on the Section 8 company registration page.

Who actually reads this site

The person with a worry

Someone searches in Hindi, often in Sanskrit, usually on a cheap Android phone, on a slow connection, about something frightening them today. There is no funnel here and nothing to convert. If the page is slow or hard to read, it has failed, and there is no second chance. The person goes back to the search results.

The funder checking you out

The second reader behaves nothing like the first. Donors, grant committees and institutional partners judge a nonprofit on governance before they judge it on intentions. That reader opens the registration and the disclosures first, and the mission statement last, if ever.

One site has to satisfy both. Most do not.

The problem

Brochure sites fail the first reader

Four pages and a donate button. Nothing for a searcher to find, so the organisation is invisible on the exact questions it exists to answer. It ranks for its own name and nothing else.

Content dumps fail the second reader

Plenty of pages, no governance anywhere. A serious funder cannot confirm the organisation is properly constituted, so they quietly move on and never say why.

Health content raises the stakes

A health education body has to do both jobs at once, in more than one language, fast enough to open on a weak connection in a small town. And it has to be careful. Health content that is wrong is not a marketing problem. It is a harm.

How we helped

Three languages, not one language with translations bolted on

English, Hindi and Sanskrit run across the whole site. Not a few translated landing pages. The whole thing.

Sanskrit as substance, not ornament

The society is named in Sanskrit and its motto is in Sanskrit. Almost the whole vocabulary of Indian health knowledge sits in that language. So the Sanskrit edition is a real interface with its own navigation, its own headings and its own web addresses under /sa/. Not a banner pasted on an English page.

This is the detail that tells a reader in ten seconds that the organisation is what it claims to be, and it is a large part of why the site does not read like a template with a logo dropped in.

The library built as real pages

The free health library carries guides, fact sheets, quick tips, a questions and answers section and annual publications, sorted into health and wellbeing, education and health literacy, climate awareness and disease prevention. Each one is an indexable page with its own URL and its own headings.

A PDF library would have been faster to build and completely invisible to search. That is the single most expensive mistake in this sector, and it is usually made to save a fortnight.

Dark mode, and a theme that follows the reader

Health material gets read late at night, on a phone, in bed. The theme control offers dark, light and automatic, and automatic follows the setting the reader already chose on their own device, so most people never have to touch it.

A CMS the organisation actually owns

Content lives in a git backed editor, so a staff member or volunteer can publish a library article, a report or a news update in any of the three languages without a developer, and every change is version controlled and reversible. The site is static, so what a visitor downloads is plain HTML with nothing to fall over under traffic. That is the same approach we take to every build on our website development service.

Governance treated as a headline, not a footer link

This is the page a grant committee opens first. Hiding it behind a footer link is the most common self inflicted wound in the sector, and it costs money that nobody ever traces back to the website.

Scale without cutting scope

At 112 pages this is the largest build we have delivered, and nothing was trimmed to make it fit a budget conversation. That is the specific thing this case study is here to demonstrate.

The SEO work, in plain words

Search engine optimisation gets talked about as a dark art. On a site like this it is mostly plumbing: making sure a machine can tell what each page is, who wrote it, and which language it is in. Here is what we actually did, and how you can check each one.

Telling Google that three languages are not three copies

This is the one most multilingual sites get wrong, and it is expensive when they do.

If you publish the same page in English, Hindi and Sanskrit and say nothing else, Google sees three pages that look alike, decides they are duplicates, keeps one and drops the rest. Two thirds of the work disappears.

The fix is a tag called hreflang. Each page points at its own translations and says which language each one is. Every page on this site carries the full set, including x-default, which tells Google what to show a reader whose language you do not publish:

<link rel="alternate" hreflang="en"
      href="https://aayushmanvidhyasociety.org/en/">
<link rel="alternate" hreflang="hi"
      href="https://aayushmanvidhyasociety.org/hi/">
<link rel="alternate" hreflang="sa"
      href="https://aayushmanvidhyasociety.org/sa/">
<link rel="alternate" hreflang="x-default"
      href="https://aayushmanvidhyasociety.org/en/">

That is copied from the live homepage. View source on any page of the site and you will find the same block. Now one health question can be answered once per language and rank once per language, instead of three versions fighting each other.

One official address per page

Every page also names itself with a canonical tag. If the same page can be reached by more than one web address, and on a real site it always can, the canonical tag says which address is the true one. Without it, Google splits the credit for a page between its duplicates and the page ranks worse than it should.

Letting the site describe itself to a machine

The homepage carries structured data, which is a small block of machine readable facts sitting quietly in the page. It declares the organisation as a NonprofitOrganization, with its legal name, its founding date, its logo and its website, plus a BreadcrumbList so search engines understand where each page sits in the structure.

This is what makes a site eligible to be quoted rather than merely listed, which matters more every year as answers get written by AI assistants instead of clicked through in a list of ten blue links. A machine cannot infer that a society is a registered nonprofit from a nice paragraph. It can read it from structured data in one pass.

Pages, not PDFs

Every guide, fact sheet, tip and answer is a real page with its own address and its own headings. A library of PDF downloads would have been quicker to build and almost invisible to search. This is the single most expensive mistake in the sector, and it is usually made to save a fortnight.

Two kinds of sitemap

There is an XML sitemap index for search engines, listing all 112 pages, and a human sitemap page for people. Most sites have the first and forget the second. The second one is also what let us verify the page count for this case study without asking anyone.

What E-E-A-T means, and how this site earns it

Google judges pages that can affect someone’s health, money or safety far more harshly than ordinary pages. The framework its own raters use is called E-E-A-T. Four words. Here they are without the jargon, and what each one looks like on this site.

Experience: has the author actually done this

Not theory, first hand involvement. The society runs community health workshops and took part in the World Health Summit regional meeting in New Delhi in 2025. The initiatives pages describe work that happened and mark it completed, rather than listing services on offer. That is the difference between an organisation and a brochure, and Google has spent three years learning to tell them apart.

Expertise: does the author know the subject

The library is organised the way someone who understands the field would organise it. Health and wellbeing, health literacy, climate awareness, disease prevention. Guides for depth, fact sheets for quick reference, questions and answers for the things people actually ask.

Authoritativeness: does anyone else recognise them

A registration number issued under a named act, by a named registrar, in a named city. A summit appearance. Eight press mentions collected on their own page, across named Hindi and English outlets including Dainik Bhaskar, Amar Ujala, Hindustan and Dainik Jagran, each dated. A stranger can check every one of those without taking the site’s word for anything.

Trust: would you act on what this page says

Trust is the one Google weighs heaviest, and it is where most sites are thin. This site publishes its governance, its registration and its annual reports as downloadable documents. It states its privacy position on every page: no tracking, no ads, no data selling, no AI training, essential cookies only. And it never asks for a login or an email address before letting anyone read a word.

A health charity that quietly sold its visitors’ behaviour would have destroyed the only thing it was ever selling. This one does not, and it says so where everyone can see.

What accessible actually means here

What WCAG is

Most sites print the word accessible as a badge. Almost none of them state a promise you can test. So here is what the words mean.

WCAG stands for the Web Content Accessibility Guidelines. It is the international standard, published by the W3C, for making a website usable by people who cannot use a website the way a designer pictures them. Someone with low vision. Someone who cannot use a mouse and moves through a page with the keyboard. Someone using a screen reader that speaks the page out loud. Someone with a tremor, or a cracked screen, or bright sunlight.

The three levels

A is the floor. AA is what most governments and serious institutions require. AAA is stricter than most real content can hold. This site commits publicly to WCAG 2.2 Level AA.

A commitment with a date on it

What Level AA asks for

Real, measurable things. Text has to stand out from its background by a set ratio, 4.5 to 1 for normal text. Every control needs a name a screen reader can say out loud, which is why the language switcher is labelled instead of being a bare icon. The page has to work by keyboard alone. It has to survive being zoomed in.

Why this matters for search too

The things that make a page usable by a screen reader are largely the things that make it readable by a search engine. Proper headings in order. Real text instead of words baked into pictures. Links that say where they go. Accessibility work and SEO work are the same work about half the time, and nobody sells it that way.

That tool is the part we are proudest of. It is the difference between claiming accessibility and proving it on a Tuesday afternoon. It also means the next person to touch the site cannot slip in an unreadable colour without the page telling them so.

How fast it actually is

Speed is an accessibility feature

Nobody files it under that heading. They should. A person on a weak connection in a small town does not experience a heavy page as slow. They experience it as broken, and they leave. Google measures roughly the same thing when it decides how to rank you.

The measurements

We measured the live site from our own machine on 27 July 2026, one visit per page, in a real Chromium browser. These are the numbers, not a score out of a hundred:

Across the English, Hindi and Sanskrit homepages, server response ran between 155 and 226 milliseconds, first paint between 1.02 and 1.32 seconds, and compressed HTML between 52 and 54 KB. The interior pages, which is where a search visitor actually lands, were faster on every measure.

Why it is fast

Not tuning. Architecture. The pages are static HTML, built before anyone asks for them, so a visitor never waits on a database, a plugin, or a server assembling a template. None of this cost the site any content. It carries 112 pages and still opens like a leaflet.

We would rather publish the method and the raw figures than a screenshot of a green circle. Measure it yourself and you should see the same shape.

Evidence

Everything on this page is checkable without asking us, which is the only kind of evidence worth publishing.

The site is live at aayushmanvidhyasociety.org. Its sitemap index publishes 112 URLs, verified 27 July 2026: 34 English, 41 Hindi and 36 Sanskrit pages, plus the root address. The society also publishes a human readable sitemap page, so the whole structure fits on one screen.

Every screenshot above is a page you can open right now. Switch the language in the header, toggle the theme, open the search, read the registration number on the transparency page, run the contrast checker. If anything here has changed since 27 July 2026, the live site wins and we would like to know.

Built to be quoted by AI, not just clicked

Search is changing under everyone’s feet. A growing share of questions never reach a list of ten blue links at all. They are answered inside ChatGPT, Gemini, Perplexity or Google’s own AI overview, and the answer names a source or it does not.

Two rules decide whether you get named.

Answer engines lift clean answers

A question needs a heading, and the heading needs a short, complete answer directly beneath it. This is why the health library is built as questions and answers, why the site carries its own FAQ page, and why the headings on this page you are reading now are phrased as questions a person would actually ask. That practice has a name, answer engine optimisation, and it is mostly just writing clearly and not burying the answer in paragraph nine.

Machines trust what they can verify

An assistant summarising a health charity will lean on the things it can confirm: a registration number, a governing act, a named registrar, a dated accessibility statement, structured data that says NonprofitOrganization in a machine readable way. This site publishes all of it. That is not a trick. It is the same evidence a funder wants, formatted so software can read it too.

What this project proves we can build

Most agencies show you a homepage. This is the whole thing, and it is a fair sample of what we do at Startup Savera. If any of the following is on your list, this is the case study to point at.

• Multilingual website development. English, Hindi, Sanskrit or any Indian language, with hreflang done properly so the versions help each other instead of competing. Most Indian sites that offer a Hindi toggle are quietly wasting it.
• NGO and nonprofit websites. Governance, transparency, annual reports and donor trust treated as primary content. If you are still setting the organisation up, we also handle Section 8 company registration and 12A and 80G registration, which is what lets your donors claim tax exemption.
• Accessible builds. WCAG 2.2 Level AA as a target you can test, not a badge in the footer.
• Fast static sites. No database to wait on, nothing to fall over, low bills, and page weights measured in tens of kilobytes.
• Content the client owns. A git backed CMS, so your team publishes without a developer and no change is ever unrecoverable.
• SEO built in from the first page, not bolted on later. Structure, schema, sitemaps and internal linking, plus the digital marketing that follows once the site can actually hold a ranking.

That is the website development service, and the rest of the work sits on the clients page.

How we worked together

Closely, and over a long stretch rather than as a single handover. We work out of Ahmedabad and the society is in Jaipur, which mattered not at all: the whole build, the reviews and the ongoing publishing run remotely, the way most of our work with founders across India does.

That closeness is exactly why the scope is as wide as it is. There was no negotiation about which sections to drop, so this is the clearest available answer to the question of what we produce when the page count is large, the content is trilingual, the subject is sensitive, and nobody is cutting the brief down to a budget.

For a commercial build with a sales outcome attached, ThermDry, Aburaj and Aura Farmers are the closer comparison. If your project is large, multilingual, or public interest work that has to survive being checked, this is the one to read.

Related reading

• Section 8 company registration in India, the structure most nonprofits should start with
• Website development, the service behind this build
• Every client we have shipped for
• Talk to us about a project