← Blog
SEOFeb. 2026· 14 min read

JSON-LD schema: a practical guide for Swiss SMEs

JSON-LD structured data improves CTR and visibility in Google. The schemas that matter for Swiss SMEs, with code examples and the mistakes to avoid.

Take two identical listings in Google's results. One shows a title and a URL. The other also shows the customer review rating (4.7 stars / 84 reviews) and the starting price of a service. Both pages hold the same organic position, but the second one will clearly get more clicks. That is what structured data in JSON-LD format makes possible. Yet many Swiss SME websites have none at all, or only what a plugin generated by default.

Structured data, without the jargon

The web is made of text that humans understand and machines interpret poorly. When your website says "Monday-Friday 9am-6pm", you and your customers know these are your opening hours. Google has to guess. Sometimes it gets it wrong, sometimes it simply decides not to show the information.

Structured data fixes this: your website tells Google explicitly what each piece of information represents. "This text is an address. This rating is 4.7 out of 5 based on 84 reviews. This list is an FAQ. This article was published on 15 September 2024."

JSON-LD (JavaScript Object Notation for Linked Data) is Google's recommended format for this data. It is a block of JSON code placed in the page's <head>, invisible to visitors but readable by search engines.

Schema markup and JSON-LD: what is the difference?

Schema.org is the vocabulary: the dictionary that defines object types (LocalBusiness, Article, FAQ, Product, etc.) and their properties (name, address, telephone, etc.). JSON-LD is the technical format in which that vocabulary is written. Google also accepts microdata and RDFa, but JSON-LD is by far the simplest to set up and the only one Google officially recommends.

The effect on CTR and rankings

Rich snippets (the enhanced results that structured data makes possible) have a direct effect on click-through rate: your result takes up more space, shows useful information before the click and stands out from the links around it.

As for rankings themselves, Google officially states that structured data is not a direct ranking factor. That is accurate, but the wording is deliberately narrow. The effect is indirect: a higher CTR (which Google measures) helps rankings, rich snippets take up more space in the SERPs, and some types (LocalBusiness above all) are used explicitly in the Knowledge Graph and the local pack.

What structured data does not do

It does not improve your content. A poorly written article with a perfect Article schema will still rank poorly. Structured data amplifies good content and a technically sound website; it does not replace them. It gives more SERP visibility to sites that already have good content.

The 5 priority schemas for Swiss SMEs

Here are the schema types that are most useful for SMEs in Geneva and across Switzerland, in order of priority.

1. LocalBusiness (and its subtypes)

This is the most useful schema for any business with premises or a service area. It gives Google your name, address, phone number, opening hours, service area and type of business: exactly what Google uses for the local pack and the Knowledge Panel.

{
  "@context": "https://schema.org",
  "@type": "LegalService",
  "name": "Cabinet Dupont Avocats",
  "url": "https://www.cabinet-dupont.ch",
  "telephone": "+41229000000",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "Rue du Rhône 14",
    "addressLocality": "Geneva",
    "postalCode": "1204",
    "addressCountry": "CH"
  },
  "openingHoursSpecification": [
    {
      "@type": "OpeningHoursSpecification",
      "dayOfWeek": ["Monday", "Tuesday", "Wednesday", "Thursday", "Friday"],
      "opens": "09:00",
      "closes": "18:00"
    }
  ],
  "geo": {
    "@type": "GeoCoordinates",
    "latitude": 46.2044,
    "longitude": 6.1432
  },
  "areaServed": {
    "@type": "City",
    "name": "Geneva"
  },
  "priceRange": "CHF CHF"
}

Schema.org offers many LocalBusiness subtypes. The precise type (LegalService, AccountingService, ArchitectOffice, Dentist, Restaurant) sends a stronger relevance signal than the generic one. Google can then show your listing in searches that the generic type would not trigger.

2. Article (and BlogPosting)

For any site publishing editorial content (blog posts, news, guides), the Article or BlogPosting schema tells Google the headline, author, publication and modification dates, and main image. The most concrete benefit: eligibility for Top Stories (the carousel of articles shown at the top of the SERPs for news queries) and a better presence in Google Discover.

The dateModified property matters in particular: a recently updated article is seen as fresher, which helps it on queries where freshness counts (legal, tax or technology news).

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "New accounting obligations for Swiss SMEs in 2025",
  "author": {
    "@type": "Person",
    "name": "Marie Favre"
  },
  "publisher": {
    "@type": "Organization",
    "name": "Fiduciaire Exemple"
  },
  "datePublished": "2024-09-15",
  "dateModified": "2024-11-20",
  "image": "https://www.exemple.ch/images/comptabilite-pme-2025.jpg",
  "description": "An overview of the accounting changes that apply to Swiss SMEs from 1 January 2025."
}

3. FAQPage

One point first: since August 2023, Google only shows FAQ accordions in its results for well-known government and health websites. An SME should therefore not count on this rich result. FAQ markup is still worth having: it tells Google and AI assistants clearly which questions your page covers and which answers it gives. That clarity is what counts today, not the accordion.

The rule: FAQs must reflect real questions from your prospects, not questions invented for SEO. The markup must also match, word for word, the questions and answers visible on the page.

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "How long does it take to set up a Sàrl (LLC) in Switzerland?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Setting up a Sàrl in Switzerland generally takes 1 to 3 weeks after the notarised articles of association are filed. The minimum capital is CHF 20,000, at least half of which must be paid up."
      }
    },
    {
      "@type": "Question",
      "name": "Can a Geneva lawyer practise in neighbouring France?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Yes, under certain conditions. A lawyer registered with the Geneva Bar can act in France in certain proceedings, notably under the European directive on established lawyers, even though Switzerland is not an EU member."
      }
    }
  ]
}

4. BreadcrumbList

The BreadcrumbList schema shows the page's navigation path in Google's results, beneath the title. Example: exemple.ch › services › audit-comptable. It serves two purposes: it helps Google understand your site structure, and it improves CTR by showing at a glance where the page sits.

For an SME with a structured website (services → sub-services, blog → categories), BreadcrumbList belongs on every page.

{
  "@context": "https://schema.org",
  "@type": "BreadcrumbList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "name": "Home",
      "item": "https://www.exemple.ch"
    },
    {
      "@type": "ListItem",
      "position": 2,
      "name": "Services",
      "item": "https://www.exemple.ch/services"
    },
    {
      "@type": "ListItem",
      "position": 3,
      "name": "Financial audit",
      "item": "https://www.exemple.ch/services/audit-comptable"
    }
  ]
}

5. Organization and WebSite

These two site-wide schemas are often forgotten, even though everything else builds on them. Organization defines your legal entity (official name, website, logo, social profiles, address) and feeds your company's Knowledge Panel on Google. WebSite enables the SearchBox in the results (the internal search field shown in the SERP for major brands) and sets the site's canonical name.

They belong on the homepage, ideally in the main layout.

Why CMSs handle structured data badly

Most WordPress, Wix, Squarespace and Jimdo sites have structured data, but it is incomplete, incorrect or even counterproductive.

Common problems with WordPress plugins

Popular plugins such as Yoast SEO or RankMath generate basic schemas, with several significant limits:

Missing data: Yoast's LocalBusiness schema includes neither GPS coordinates (geo), nor detailed openingHoursSpecification, nor areaServed. Yet these fields matter a great deal for local SEO.

Conflicting data: when several plugins generate schema (Yoast + WooCommerce + a booking plugin), the blocks can contradict or duplicate each other. Google flags these errors in Search Console, but many site owners never open it.

Invalid markup: plugin-generated schemas often contain validation errors. The most common: missing required properties, invalid types, malformed URLs.

No page-specific schema: a plugin applies the same schema to every page of a given type. With Yoast, every blog post gets the Article schema, with no distinction for those that would deserve a NewsArticle or FAQPage. Relevance suffers.

What Next.js allows natively

A website built with Next.js allows a JSON-LD implementation that a CMS cannot match:

Injection into the <head> via the layout: with the Next.js App Router, JSON-LD is injected into the server render by a dedicated component. It is therefore present in the HTML sent to crawlers, with no dependence on client-side JavaScript.

// components/JsonLd.tsx
export function JsonLd({ data }: { data: Record<string, unknown> }) {
  return (
    <script
      type="application/ld+json"
      dangerouslySetInnerHTML={{ __html: JSON.stringify(data) }}
    />
  )
}

// app/layout.tsx
import { JsonLd } from '@/components/JsonLd'

export default function RootLayout({ children }) {
  return (
    <html>
      <head>
        <JsonLd data={organizationSchema} />
        <JsonLd data={websiteSchema} />
      </head>
      <body>{children}</body>
    </html>
  )
}

Schemas built from your data: on a dynamically generated service page, the schema is built from the service's data (name, price, service area), with no manual input. Each page gets its own exact schema.

Validation at build time: with strict TypeScript, schema types are defined and checked. Missing or malformed properties, which go unnoticed in a CMS, are caught before going live.

No plugins, no conflicts: there is no intermediate layer between the developer and the final markup. What is coded is exactly what is delivered.

Validating and monitoring structured data

Writing a schema is not enough. You need to check that Google reads it correctly and track when rich snippets appear.

Validation tools

Comparative data

ToolUseAvailability
Google Rich Results TestTest a URL or JSON-LD codeFree, online
Schema Markup Validator (schema.org)Strict validation of types and propertiesFree, online
Google Search Console → EnhancementsMonitoring errors and warnings in productionFree (verified property)
Ahrefs / Semrush Site AuditSite-wide auditPaid

Start with Google's Rich Results Test. It tells you whether the schema is valid, and also whether the page is eligible for rich results for that schema type.

How long rich snippets take to appear

After implementation, Google can take 2 to 6 weeks to crawl the page, validate the schema and decide whether to show a rich snippet. The delay depends on how often the site is crawled (frequently updated sites are revisited more often) and on the quality of the page's content.

A valid schema on a mediocre page will not produce a rich snippet: Google reserves the right not to show it if it judges the page insufficiently relevant or reliable.

Mistakes to avoid

Schema spam: adding a LocalBusiness schema to every blog post, or declaring properties that do not match the visible content. Google's structured data guidelines forbid misleading markup, and a site that breaks them can lose its rich snippets through a manual action.

Inconsistent data: if your LocalBusiness schema states different opening hours from your contact page, Google may ignore the schema or show users the inconsistency.

Forgetting @context and @type: these two properties are mandatory in every JSON-LD schema. Without them, the markup is invalid.

Duplicate schema: several JSON-LD blocks of the same type (Organization) on one page cause confusion. One schema per type per page.

The long-term effect on SEO

Structured data is one of the few technical SEO measures whose effect builds up over time. Every correctly marked-up page gains visibility in the SERPs. A site with 50 blog posts, each with a relevant BlogPosting schema and a FAQPage that makes its questions and answers readable for Google and AI assistants, starts with a clear lead over competitors without structured data, at equivalent content.

The trend is clear: Google shows more and more information within the results themselves (People Also Ask, Featured Snippets, Knowledge Panels, Shopping Graph). Sites that use structured data can be included in these formats. The rest stay limited to the classic blue link.

In 2026, ignoring structured data is like ignoring responsive design in 2014: it is not mandatory yet, but every month without it widens the gap with those who have adopted it.

Advanced schemas for needs specific to Swiss SMEs

On top of the five core schemas, several markup types serve sectors and needs that come up often among Geneva SMEs.

Service and Offer: for service providers

The Service schema describes each service: name, description, service area, provider. Combined with Offer, it can show a price range in the results. For a fiduciary firm offering accounting, tax and legal advice, one Service schema per service adds a level of detail that LocalBusiness alone cannot provide.

{
  "@context": "https://schema.org",
  "@type": "Service",
  "name": "Personal tax returns in Geneva",
  "provider": {
    "@type": "AccountingService",
    "name": "Fiduciaire Exemple"
  },
  "areaServed": {
    "@type": "AdministrativeArea",
    "name": "Canton of Geneva"
  },
  "offers": {
    "@type": "Offer",
    "priceRange": "CHF 180-450"
  }
}

Event: for those who organise events

Law firms running webinars, architects giving talks, fiduciary firms hosting tax workshops: the Event schema lets them appear in rich results for local event searches, with the date, venue and registration link visible.

Person: to highlight individual expertise

For independents and professionals whose personal reputation drives the business (law firm partners, independent architects, consultants), a complete Person schema (name, education, areas of expertise, professional affiliations) feeds the personal Knowledge Panel and can appear in results for searches on their name.

Structured data and AI: preparing your website

With Google's AI Overviews and the spread of AI assistants that "read" the web to answer questions, a practical question arises: does structured data have a role to play?

Yes, and probably an increasing one. The LLMs (Large Language Models) that index the web, such as those behind AI answer engines or Google's AI Overviews, use structured data to understand the context and nature of information. A fact presented in a FAQPage schema, with a clearly delimited question and answer, is easier to extract and cite than an unstructured paragraph.

Schema.org is extending its vocabulary for these uses, with new types such as DefinedTerm and properties that better describe causal and temporal relationships between entities. SMEs that invest in complete structured data now will be better placed: their content will be easier for these systems to read, extract and cite.

The cost of implementation, in perspective

Structured data is often presented as complex and expensive. Here is a realistic estimate.

On a custom-built Next.js website, JSON-LD is part of the initial development. It is not a separate extra but part of the standard delivery. The JsonLd component, the Organization and WebSite schemas and the page-level schemas are set up once, then maintained automatically.

On an existing website (WordPress or another CMS), adding the essential structured data takes anywhere from a few hours for a basic LocalBusiness schema to a few days for a complete implementation on a site with 50+ pages and several content types. The return can be measured through CTR growth in Search Console, over a few months.

What costs money is not doing it: every month without rich snippets, better equipped competitors capture clicks that could have been yours.

Frequently asked questions

Is structured data a Google ranking factor?

Not directly, according to Google. Its effect is indirect: rich snippets take up more space in the results and improve CTR, which helps rankings.

Which schemas should a Swiss SME set up first?

LocalBusiness (with the precise subtype), Article or BlogPosting, FAQPage, BreadcrumbList, then Organization and WebSite on the homepage.

How long does it take for rich snippets to appear?

Google can take 2 to 6 weeks to crawl the page, validate the schema and show a rich snippet. On a mediocre page, it may never show one.

At INFUSE, every website is delivered with complete, validated structured data (LocalBusiness, BreadcrumbList, Article, FAQ), written directly into the Next.js code, with no plugins. Let's talk about your project.

One point of contact, from design to code.

Request a quote