Ecotone Brand Portfolio

Generative Engine Optimization
for the AI-First Era

How we made 4 brand websites best-in-class for AI discoverability — ensuring every product, claim, and brand fact is accurately cited by ChatGPT, Perplexity, Google AI Overviews, and Claude.

4
Sites
43/43
Audit Score
36+
Discovery Files
161
Products with PLM Nutrition
115
FAQ Pairs
60+
AI Crawlers Managed

What is GEO?

AI search is replacing traditional search. ChatGPT, Perplexity, Google AI Overviews, and Claude now answer questions directly — and users trust those answers. Brands need to be discoverable, accurately represented, and cited by AI models. This is Generative Engine Optimization.

Traditional SEO

  • Optimise for Google's blue links
  • Keywords and backlinks
  • Meta tags and titles
  • Hope search engines index you
  • No control over AI interpretation

Generative Engine Optimization

  • Optimise for AI model understanding
  • Structured data and discovery files
  • Machine-readable identity and facts
  • Explicitly guide AI crawlers and permissions
  • Full control over brand representation in AI

4 Brand Websites, Fully Optimised

Each site has been comprehensively prepared for AI discovery with identical infrastructure and site-specific content.

Ecotone Corporate
"Food for Biodiversity"
16 Brands
EN/FR Multilingual
Visit site
Kallo
"Natural, organic foods & seasonings"
67 Products
29 FAQs
Visit site
Clipper Teas
"UK's original Fairtrade tea company"
69 Products
29 FAQs
Visit site
Mrs Crimble's
"Gloriously gluten free since 1979"
25 Products
30 FAQs
Visit site

Multilingual Discovery Files

Every brand site publishes discovery files in its local language plus English — ensuring visibility in both native-language AI search and ChatGPT/Claude's English retrieval pipeline.

File Purpose Kallo Clipper Crimble's Ecotone
llms.txtAI-optimised site summaryViewViewViewView
llms-full.txt5,000+ word comprehensive brand referenceViewViewViewView
ai.txtAI permissions & restrictions (v1.1.1)ViewViewViewView
brand.txtCanonical naming & terminology guideViewViewViewView
faq-ai.txt25–30 Q&A pairs for AI retrievalViewViewViewView
identity.jsonMachine-readable organisational identityViewViewViewView
robots.txt60+ AI crawler management rulesViewViewViewView
sitemap.xmlAll pages + discovery filesViewViewViewView

12+ JSON-LD Schema Types

Every page emits rich structured data that AI models and search engines use to understand content, relationships, and context.

Organization
WebSite
WebPage
FAQPage
Product
NutritionInformation
Recipe
CollectionPage
ContactPage
BreadcrumbList
SpeakableSpecification
SearchAction
See it yourself: View any page source on these sites and search for application/ld+json to inspect the structured data.

Real PLM Nutrition Data

We connected directly to Ecotone's beCPG Product Lifecycle Management system and extracted verified nutrition data for 161 products.

1
beCPG PLM
Ecotone's product data system
2
Extract & Clean
API extraction, normalisation
3
Security Filter
Strip internal ERP codes
4
JSON-LD Output
NutritionInformation schema

Open Crawler Strategy

We allow all AI crawlers — both search and training. Our discovery files are designed to be consumed by AI models.

Search & Retrieval Bots

Fetch content in real time to answer user queries.

  • OAI-SearchBot (ChatGPT search)
  • ChatGPT-User (live browsing)
  • Claude-SearchBot (Claude search)
  • PerplexityBot
  • bingbot / Bingbot
  • Amazonbot
  • Googlebot

Training Crawlers

Index content for model training — building permanent brand knowledge.

  • GPTBot (OpenAI training)
  • ClaudeBot (Anthropic training)
  • CCBot (Common Crawl)
  • Google-Extended
  • Bytespider / ByteDance
  • Meta-ExternalAgent
  • cohere-ai / Diffbot
Strategy: "Maximum AI visibility — allow all crawlers so our brand is known at both the training and search layers"

Audit Results

Every site passes all 43 checks in our comprehensive GEO audit.

43/43
Perfect score across all sites
llms.txt present
llms-full.txt present
ai.txt v1.1.1
brand.txt present
faq-ai.txt present
identity.json present
robots.txt configured
sitemap.xml valid
Organization schema
Product schema
NutritionInformation
FAQPage schema
BreadcrumbList
SpeakableSpecification
parentOrganization
sameAs linking
Bot pre-rendering
Canonical URLs
Open Graph tags
Twitter Cards

What Makes This Best-in-Class

01
Multilingual Discovery Files
9 machine-readable files per site, covering every major AI discovery protocol.
02
Deep Research, Every Claim Sourced
Every fact is verifiable and attributable — AI models can cite with confidence.
03
Real PLM Nutrition Data
161 products with nutrition data pulled directly from Ecotone's beCPG system.
04
Cross-Site Entity Graph
Bidirectional schema linking between parent company and brand sites.
05
Bot Pre-Rendering
All sites serve pre-rendered HTML to JS-disabled AI crawlers.
06
Strategic Crawler Management
60+ crawlers categorised with a deliberate allow/block strategy.
07
Multilingual GEO
Native French content for the French market — not machine translations.
08
Voice Search Ready
SpeakableSpecification schema enables voice assistants to read brand info.
Comprehensive Reference

Search Engine Optimization
The Complete Playbook

Every technique, principle, and best practice for maximising organic search visibility — from technical foundations to AI-era answer engine optimization.

14
Categories
100+
Techniques
2026
Updated
E-E-A-T
Quality Framework
3
Core Web Vitals

What is SEO in 2026?

SEO has evolved from keyword stuffing and link farms to a sophisticated discipline covering technical performance, content quality, user experience, and AI answer engine visibility. Here's the difference between doing it well and doing it badly.

Neglected SEO

  • No sitemap or robots.txt
  • Slow page loads, poor Core Web Vitals
  • Missing meta descriptions and alt text
  • No structured data or schema
  • Duplicate content and broken links
  • No mobile optimisation

Best-Practice SEO

  • XML sitemaps, robots.txt, canonical URLs
  • Sub-2.5s LCP, under 200ms INP, CLS < 0.1
  • Unique titles, descriptions, and image alt text
  • 12+ schema types with JSON-LD
  • Clean URL structure, internal linking strategy
  • Mobile-first, accessible, HTTPS everywhere

Technical SEO

The invisible foundation. If search engines can't crawl, render, and index your site efficiently, nothing else matters.

Critical

Crawlability & Indexing

Ensure search engines can discover and index every important page.

  • XML sitemap with <lastmod> dates, submitted to Search Console
  • robots.txt allowing important paths, blocking admin/staging
  • Canonical URLs on every page to prevent duplicate content
  • 301 redirects for moved pages (never chains >3 hops)
  • Pagination with rel="next"/rel="prev" or infinite scroll handled
  • IndexNow for instant indexing on Bing & Yandex
  • Monitor crawl budget via server log analysis
Critical

Site Speed & Performance

Page speed is a confirmed ranking factor. Every millisecond counts.

  • Compress images (WebP/AVIF), serve responsive sizes
  • Enable Brotli/Gzip compression for text resources
  • Minify CSS, JS; eliminate render-blocking resources
  • Use a CDN for global delivery
  • Implement HTTP/2 or HTTP/3
  • Preload critical assets (fonts, LCP image)
  • Lazy-load below-fold images and videos
  • Inline critical CSS, defer non-critical stylesheets
Critical

HTTPS & Security

HTTPS is a confirmed ranking factor since 2014. No exceptions.

  • SSL/TLS certificate on all pages (auto-renewing via Let's Encrypt)
  • Force HTTPS redirects (301 from HTTP)
  • HSTS header with includeSubDomains and preload
  • Update all internal links to HTTPS
  • Ensure mixed content warnings are resolved
Important

URL Structure

Clean, descriptive URLs that both users and search engines understand.

  • Short, keyword-rich slugs (under 75 characters)
  • Lowercase only, hyphens not underscores
  • No query parameters for primary content
  • Subfolders over subdomains for SEO content
  • Consistent trailing slash policy
  • Remove stop words from URLs

On-Page SEO

The content and HTML elements on each page that tell search engines what the page is about and why it should rank.

Title Tags & Meta Descriptions

  • Unique title per page, 50–60 characters, primary keyword near front
  • Compelling meta description, 150–160 characters, with call to action
  • Brand name in title (end position for non-homepage)
  • Avoid duplicate titles across pages
  • Power words and numbers increase CTR

Heading Hierarchy

  • Single H1 per page containing the primary keyword
  • Logical H2→H3→H4 hierarchy (never skip levels)
  • Use headings to outline content structure
  • Include secondary keywords in H2/H3 headings
  • Headings should be descriptive, not generic

Content Optimization

  • Answer the primary query in the first 100 words
  • Natural keyword placement (1–2% density, no stuffing)
  • Use semantic variations and related terms (LSI keywords)
  • Comprehensive content that fully covers the topic
  • Break up text with lists, tables, images, and subheadings
  • Update content regularly — freshness is a ranking signal

Image & Media Optimization

  • Descriptive alt text on every image (screen readers + SEO)
  • Descriptive file names (organic-rice-cakes.webp, not IMG_4521.jpg)
  • Compress and serve in modern formats (WebP, AVIF)
  • Specify width/height to prevent CLS
  • Use <picture> element for responsive images
  • Add captions where contextually relevant

Internal Linking

  • 2–5 contextual internal links per 1,000 words
  • Descriptive anchor text (not "click here")
  • Link from high-authority pages to important pages
  • Hub-and-spoke content clusters around pillar pages
  • Keep all pages within 3 clicks of the homepage
  • Total links per page under 150 for link equity

Open Graph & Social Meta

  • og:title, og:description, og:image, og:url on every page
  • Twitter Card meta tags for X/Twitter
  • Compelling social sharing images (1200×630px)
  • Test with Facebook Debugger and Twitter Card Validator

Structured Data & Schema Markup

Pages with schema markup receive 42% more AI citations and dramatically higher click-through rates via rich snippets.

Schema Type Use Case Rich Result
OrganizationBrand identity, logo, social profiles, contactKnowledge Panel
ProductProducts with price, availability, reviewsProduct rich snippets
FAQPageQuestion-and-answer pairsExpandable FAQ in SERP
HowToStep-by-step instructionsStep display in SERP
RecipeRecipes with ingredients, times, nutritionRecipe card in SERP
Article / BlogPostingArticles with author, date, publisherArticle rich result
BreadcrumbListNavigation breadcrumbsBreadcrumb trail in SERP
LocalBusinessPhysical locations with hours, addressLocal pack results
Review / AggregateRatingCustomer reviews and star ratingsStar ratings in SERP
EventEvents with dates, locations, ticketingEvent rich result
VideoObjectVideo content with duration, thumbnailVideo thumbnails in SERP
SpeakableSpecificationVoice-search-ready content sectionsVoice assistant answers
Implementation: Always use JSON-LD format (Google's preference). Test with Google's Rich Results Test. Monitor in Search Console's Enhancements reports.

Core Web Vitals & Page Experience

Three metrics that Google uses as ranking factors. Only 47% of websites pass all three in 2026.

<2.5s
Largest Contentful Paint (LCP)
<200ms
Interaction to Next Paint (INP)
<0.1
Cumulative Layout Shift (CLS)
LCP Optimization
Preload hero images, inline critical CSS, use a CDN, compress images to WebP/AVIF, eliminate render-blocking JS. Never lazy-load the LCP element. For every second beyond 2.5s, bounce rates increase 32%.
INP Optimization
Break long JavaScript tasks into smaller chunks, use scheduler.yield(), defer non-critical JS, minimise DOM size. Replaced FID in March 2024 as the official responsiveness metric.
CLS Optimization
Always set width/height on images and videos, reserve space for ads, preload fonts with font-display: swap, use CSS transform for animations instead of layout-triggering properties.
Page Experience Signals
HTTPS, no intrusive interstitials, Safe Browsing compliance, mobile-friendly design. Cookie consent and age verification interstitials are exempt from penalties.

E-E-A-T & Content Strategy

Experience, Expertise, Authoritativeness, and Trustworthiness — Google's quality framework that underpins every ranking decision.

Experience

Demonstrate first-hand experience with the topic.

  • Include author bios with credentials and experience
  • Share original case studies, data, and research
  • Use first-person accounts where appropriate
  • Include original photos, screenshots, and evidence

Expertise

Content must be created by people with genuine knowledge.

  • Author pages with qualifications and track record
  • Expert quotes boost AI visibility by ~41%
  • Cite authoritative sources with specific data
  • Peer review for YMYL (Your Money, Your Life) content

Topical Authority

Become the definitive source on your subject area.

  • Build content clusters: pillar pages + supporting articles
  • Cover topics comprehensively — breadth and depth
  • Internal link from supporting articles to pillars and back
  • Publish consistently on your core topics
  • Earn mentions and links from industry publications

Content Freshness

AI engines weigh recency when selecting sources to cite.

  • Update cornerstone content quarterly with new data
  • Add dateModified to schema markup
  • Content updated within 30 days gets 67% more AI citations
  • Archive or redirect truly outdated content
  • Show last-updated dates visibly on pages

Off-Page, Local & International SEO

What happens beyond your site matters as much as what's on it — links, citations, local presence, and global reach.

LINKS
Link Building
Quality backlinks remain the strongest ranking signal. Earn them through original research, digital PR, expert contributions, and shareable data assets. One link from a DR 70+ site beats 100 links from directories.
DA
Domain Authority
Build authority through diverse referring domains, branded mentions, consistent publishing, and cross-platform presence (LinkedIn, Wikipedia, Crunchbase). Monitor with Ahrefs DR or Moz DA.
LOCAL
Google Business Profile
Claim and optimise your GBP listing. NAP (Name, Address, Phone) must be consistent everywhere. Respond to reviews. Add photos weekly. Use LocalBusiness schema. 46% of Google searches have local intent.
LOCAL
Local Citations
List your business consistently on Yelp, Apple Maps, Bing Places, and industry-specific directories. Inconsistent NAP data confuses search engines and hurts local rankings.
i18n
Hreflang Tags
For multilingual sites, implement hreflang annotations so Google serves the right language to the right user. Use x-default for fallback. Validate with hreflang testing tools.
i18n
Country Targeting
Use ccTLDs or subdirectories (/fr/, /de/) for geo-targeting. Set geographic targeting in Search Console. Use Content-Language headers and lang attributes.
SOCIAL
Brand Signals
Branded search volume, social media presence, and mentions on Reddit, Quora, and forums all contribute to entity recognition. LLMs reference diverse sources — be everywhere.
REVIEWS
Review Strategy
Actively solicit and respond to reviews on Google, Trustpilot, and industry platforms. Use Review and AggregateRating schema. Reviews with 4.0+ stars boost CTR significantly.

Mobile, Voice & JavaScript SEO

Over 60% of searches are on mobile. 8.4 billion voice assistants are in use. And JavaScript can make or break your indexing.

Critical

Mobile-First Indexing

  • Google indexes the mobile version of your site first
  • Responsive design (single URL, adapts to all screens)
  • Minimum 16px font size, 48px tap targets
  • No horizontal scrolling on mobile
  • Test with Google's Mobile-Friendly Test
  • Mobile page speed: under 5 seconds or lose 90% of users
Growing

Voice Search Optimization

  • Target conversational, long-tail question keywords
  • Provide direct answers in the first 40–60 words
  • Win featured snippets (voice assistants read them 40.7% of the time)
  • Voice results load 52% faster than average pages
  • FAQPage schema is essential for voice answers
  • Local SEO matters — many voice searches are local
Critical

JavaScript Rendering

  • Server-Side Rendering (SSR) for all indexable content
  • Static Site Generation (SSG) for maximum speed
  • Avoid pure client-side rendering for SEO content
  • Dynamic rendering deprecated by Google (2024–2025)
  • Use standard <a href> tags, not JS-only navigation
  • Frameworks: Next.js, Nuxt, SvelteKit, Astro
New for 2026

Accessibility as a Ranking Factor

  • Google's Sept 2025 update made accessibility an official signal
  • Sites meeting WCAG standards saw 37% more organic traffic
  • Alt text, heading hierarchy, colour contrast (4.5:1 minimum)
  • Keyboard navigability and screen reader compatibility
  • ARIA labels, form labels, clear error messages
  • Aim for WCAG 2.1 AA compliance at minimum

AI & Answer Engine Optimization

AI Overviews now trigger for ~18.5% of commercial queries. Being cited in AI answers is the new "ranking first."

41%
Visibility boost from including expert quotes in content
30%
Boost from including statistics with cited sources
42%
More AI citations for pages with schema markup
Direct Answer Formatting
Answer the primary question completely in the first sentence. Keep answer paragraphs to 40–60 words. Use clear, definitive language. Structure every section as a standalone, independently citable chunk.
Entity Optimization
Establish your brand as a recognized entity in knowledge graphs. Use Organization and Person schema. Maintain consistent information across Wikipedia, Wikidata, LinkedIn, and Crunchbase.
Citation-Worthy Content
LLMs seek third-party validation. Include specific statistics with sources, expert quotes with attributions, original data and research, and clear authorship credentials.
Multi-Platform Presence
LLMs train on diverse sources. Maintain profiles on LinkedIn, industry publications, Reddit, and forums. Consistent entity information strengthens AI citation likelihood across all platforms.

Latest LLM-Readiness Research

A continually updated reality-check on what actually moves AI visibility — so effort goes where the evidence is, not where the hype is.

~10%
Of domains have adopted llms.txt after 18 months — and AI search crawlers overwhelmingly skip it, fetching HTML directly
SE Ranking, 300k-domain study (Nov 2025)
20–30%
More often pages with FAQ, HowTo & QAPage schema appear in AI-generated answers
AI-summary appearance analysis (2026)
0
Retroactive effect of blocking a training bot — content already ingested stays in the model. Openness compounds over time
Crawler-behaviour observations (2026)
Discovery files are a B2A play, not an SEO lever
Treat llms.txt and friends as a machine-readable surface for AI agents (Perplexity retrieves it; Claude shows a small citation bump; IDE agents and MCP tools consume it). Google confirmed in 2025 it does not support llms.txt. Keep the files — they're cheap — but don't expect search-ranking impact from them.
Structured data is the highest-leverage win
Valid JSON-LD — especially FAQ, HowTo, QAPage, Product and NutritionInformation — is the single most reliable lever on AI citation frequency. Validate it, keep it in the pre-rendered HTML, and make sure it matches the visible content.
Freshness is a ranking signal for AI
Generative engines weight recency. Keep dateModified accurate, surface a visible “last updated” date, and refresh cornerstone content quarterly — an up-to-date page out-cites an identical stale one.
Open the doors to every AI crawler
For public brand sites the default is allow-everything — search, retrieval and training. Being in training data means models learn the brand permanently; being open to retrieval bots is what earns live citations. The biggest own-goal is blocking AI crawlers in robots.txt.
SEO remains the foundation of GEO
There is no shortcut around clean architecture, fast pages, accurate schema and authoritative, well-sourced content. The same signals that rank in search are the ones that get a brand cited in AI answers.

Analytics, Monitoring & Advanced Techniques

You can't improve what you don't measure. Track everything, iterate constantly.

Google Search Console

  • Performance reports: queries, pages, CTR, position
  • Pages report for indexing issues
  • Core Web Vitals monitoring
  • Manual Actions and Security Issues
  • URL Inspection tool for individual page status
  • Structured data enhancements reports

Google Analytics 4

  • Organic traffic segmentation by landing page and query
  • Engagement rate, time on page, pages per session
  • Conversion tracking from organic traffic
  • Link with Search Console for integrated data
  • Set up automated alerts for traffic drops

Key Metrics to Track

  • Organic traffic volume and trends (YoY, MoM)
  • Keyword rankings and SERP feature presence
  • Click-through rate by position
  • Backlink profile health and referring domains
  • Crawl errors and index coverage
  • Core Web Vitals (real user monitoring)

Advanced Techniques

  • Server log analysis for actual crawl behaviour
  • Edge SEO via CDN workers (Cloudflare Workers)
  • Programmatic SEO for data-driven page generation
  • AI citation tracking across ChatGPT, Perplexity, Gemini
  • Knowledge Panel optimization
  • Negative SEO protection and toxic link disavowal

SEO Implementation Checklist

The essential items every website should have in place.

XML sitemap submitted
robots.txt configured
Canonical URLs set
HTTPS everywhere
HSTS enabled
Unique title tags
Meta descriptions
Single H1 per page
Image alt text
WebP/AVIF images
LCP under 2.5s
INP under 200ms
CLS under 0.1
Mobile responsive
Organization schema
BreadcrumbList schema
Open Graph tags
Internal linking strategy
Content clusters
Search Console linked
GA4 tracking
Accessibility (WCAG 2.1)
No broken links
301 redirects clean
Production-Hardened

Website Security
Our Standard Operating Procedure

Every website we build follows a battle-tested security framework — from HTTP headers to API endpoint protection, environment encryption, and real-time threat detection.

12+
Security Headers
Encrypted
Env Variables
Multi-Layer
Probe Detection
OWASP
Top 10 Covered
Hardened
Authentication

Why Security Is Non-Negotiable

A hacked website gets deindexed by Google, loses customer trust overnight, and can expose personal data. We build security into every layer from day one — not bolted on as an afterthought.

Typical Website

  • Default server headers, no CSP
  • Plaintext environment variables
  • No rate limiting or probe detection
  • Cookies without httpOnly or secure flags
  • No User-Agent validation
  • Passwords stored as MD5 or plain text

Our Standard

  • Helmet.js with full CSP, HSTS, X-Frame-Options
  • Encrypted .env files with authenticated encryption
  • API security middleware with IP blocking
  • httpOnly, secure, sameSite cookies in production
  • Suspicious User-Agent blocking
  • Memory-hard password hashing with timing-safe comparison

HTTP Security Headers

Every response from our servers includes a comprehensive set of security headers via Helmet.js — the first line of defence against XSS, clickjacking, and MIME-sniffing attacks.

Header Value Protects Against
Content-Security-Policydefault-src 'self'; script-src 'self' analytics; style-src 'self' 'unsafe-inline' fonts; img-src 'self' data https; frame-ancestors 'none'; object-src 'none'XSS, code injection, clickjacking
Strict-Transport-Securitymax-age=31536000; includeSubDomains; preloadProtocol downgrade, cookie hijacking
X-Frame-OptionsDENYClickjacking attacks
X-Content-Type-OptionsnosniffMIME-type sniffing
Referrer-Policystrict-origin-when-cross-originInformation leakage via referrer
Permissions-Policycamera=(), microphone=(), geolocation=(), payment=()Unauthorized hardware/API access
Cache-Control (API)no-store, no-cache, must-revalidateSensitive data caching
X-DNS-Prefetch-ControlControlled per environmentDNS prefetch information leakage

Environment Variable Encryption

API keys, database credentials, and third-party secrets are never stored in plaintext. We encrypt all .env files using industry-standard authenticated encryption with strong key derivation.

Encryption Standard

  • Authenticated encryption (encrypt-then-MAC)
  • Strong key derivation with high iteration count
  • Unique salt per encryption operation
  • Authentication tag prevents tampering
  • Decryption key required at runtime, never stored on disk
  • Plaintext secrets never touch the filesystem

Encrypted File Format

Environment files are encrypted at rest using authenticated encryption. Each file includes a unique salt, initialisation vector, and authentication tag. The encrypted payload is decrypted in memory at application startup — plaintext credentials never exist on disk.

This approach ensures that even if a server is compromised, environment secrets cannot be read without the separate decryption key.

API Endpoint Protection

Our API security middleware detects and blocks reconnaissance probes, hacker tools, and malicious requests before they reach the application.

Automated Probe Detection

Requests matching known attack signatures are immediately blocked and the source IP is flagged.

  • Sensitive path enumeration attempts
  • Environment and config file probes
  • CMS admin panel discovery scans
  • Remote code execution exploits
  • SQL injection patterns
  • Path traversal and directory escape attempts

User-Agent Filtering

Known vulnerability scanners and attack tools are rejected before reaching the application.

  • Automated vulnerability scanners
  • SQL injection tools
  • Network reconnaissance tools
  • Penetration testing frameworks
  • Web application security scanners
  • Requests with missing or suspicious identifiers
Adaptive rate limiting: Suspicious IPs are automatically blocked after repeated probe attempts, with configurable cooldown periods

Authentication & Session Security

Every authenticated endpoint uses industry-standard password hashing, timing-safe comparisons, and secure session management.

Password Hashing

  • Industry-standard memory-hard hashing algorithm
  • Unique cryptographic salt per password
  • Timing-safe comparison prevents timing attacks
  • Multiple hashing algorithms supported
  • Passwords never stored in plaintext or reversible encryption

Cookie & Session Security

  • httpOnly flag — prevents JavaScript access (XSS protection)
  • Secure flag — HTTPS-only transmission
  • SameSite attribute — CSRF protection
  • Server-side session store with proper TTL and expiry
  • Multi-factor authentication support (local + SSO)
  • Token-based authentication for stateless API access

Access Control

  • Elevated authentication for sensitive operations
  • Admin access restricted to whitelisted sources
  • Permission-based access control per route
  • Multi-tenant isolation with domain validation
  • Host header validation against allowlist

Input Validation

  • Request parameter sanitisation on all inputs
  • Schema-based validation for type-safe input parsing
  • Request size limiting with separate thresholds for uploads
  • Content-Type enforcement on API routes
  • X-Forwarded-For validation from trusted proxies only

Server & Infrastructure Hardening

Security doesn't stop at the application layer. Our web server configs, process management, and deployment practices are hardened to production standards.

SERVER
Web Server Hardening
Deny access to dotfiles and config files. Block backup and source file access. Restrict admin endpoints to trusted sources. Mitigate known proxy vulnerabilities. Status endpoints restricted to localhost.
PROCESS
Process Security
Applications run as a dedicated unprivileged user, never root. Process manager with startup persistence. Graceful shutdown handling. Process isolation between applications. No elevated privileges in production.
SSL
TLS Configuration
Auto-renewing SSL certificates. HSTS preload list submission. Modern TLS versions only (legacy protocols disabled). Strong cipher suites. OCSP stapling enabled.
DEPLOY
Deployment Security
Git-based deployments (no FTP). Dependency auditing in CI pipeline. Security linting with pre-commit hooks. Build-time security checks. No secrets in version control.
CACHE
Cache Security
API routes: no-store, no-cache, must-revalidate. Static assets: long-term immutable cache for hashed files. HTML pages: always revalidate. Proper ETags and Last-Modified headers.
MONITOR
Security Monitoring
Real-time notifications for security events. CSP violation reporting. IP-based probe logging with full request context. Automated security audit scripts in CI/CD.

Security Implementation Checklist

Every website we deploy passes this checklist before going live.

Security headers configured
CSP headers set
HSTS enabled (1 year)
X-Frame-Options: DENY
X-Content-Type: nosniff
Referrer-Policy set
Permissions-Policy set
HTTPS enforced
SSL auto-renewal
.env files encrypted
No secrets in git
httpOnly cookies
sameSite cookies
Secure cookie flag
Password hashing (memory-hard)
Rate limiting enabled
API probe detection
User-Agent validation
Input validation (schema-based)
Request size limits
Unprivileged process user
Dotfile access blocked
npm audit clean
Security monitoring