# Pylon SEO > Pylon SEO is an all-in-one SEO platform from Bytenovo, a company registered in Gurugram, > Haryana, India. It crawls a website, audits technical SEO, tracks keyword rankings, > researches keywords, measures brand visibility inside AI answers, and compares a site > against competitors. It is delivered as a browser dashboard plus native apps and plugins > for WordPress, Shopify, Wix, Squarespace and BigCommerce. This document is the full-text companion to [llms.txt](https://pylonseo.io/llms.txt). It is written for language models and agents that want Pylon SEO's substantive content in one request rather than page by page. **Two labelling rules apply throughout this document and to the product itself.** Measured figures — crawl results, keyword positions, audit scores, AI mention counts — trace to stored records. Modelled figures — search demand, competitor authority, competitor traffic — are estimates and are labelled as such in the product. Where this document gives a capability, it is one the product actually implements. Where a capability depends on a third party, the dependency is named. --- ## What Pylon SEO does Pylon SEO covers the four jobs an SEO team performs in sequence: 1. **Discover** — crawl the site and find what is technically wrong with it. 2. **Diagnose** — turn crawl output into a prioritized list of issues with a fix for each. 3. **Recommend** — identify keywords the site could rank for and competitors it is losing to. 4. **Measure** — track classic search positions and AI-answer visibility over time, and confirm whether the fixes worked. All modules share one site, one data model and one prioritization language, so an audit finding, a keyword, a ranking and an AI prompt are all connected rather than siloed. ### The crawl → measure → label workflow - **Crawl.** robots.txt is fetched and honoured before any page is requested. Crawl depth and URL budgets are configurable and respected. Outbound requests are SSRF-checked. - **Measure.** Rankings come from live SERP providers rather than simulation. Snapshots are archived so movement can be shown between runs. - **Label.** Any modelled value carries an "estimated" label. The product does not round a model output up and present it as a measurement. --- ## Smart Crawler - Breadth-first crawl that fetches and honours robots.txt before crawling. - XML and HTML sitemap discovery, so pages not linked internally are still found. - Canonical handling, so canonical and duplicate variants are not counted twice. - Configurable depth and URL budgets. - SSRF-safe outbound networking (private and link-local address ranges are refused). - Best-effort probe of the site's own `/llms.txt`, stored and used as an input to the AI readiness audit rules. - Crawl journal: every fetched URL with its status code, response time, payload size, depth, indexability verdict and failure reason. ## Technical SEO Audit The audit runs **35 rules across six categories**: technical (13), on-page (8), content (7), AI readiness (4), performance (2) and structured data (1). Every rule carries a severity and an effort estimate, so findings are ordered by impact per hour of development time rather than by raw count. Cross-page issues are counted once instead of once per affected URL. Severity tiers are Critical, High, Medium and Low. ### Coverage areas **Link structure** - Anchor text quality and over-optimization - Orphan pages with zero internal inbound links - Broken internal and external links - URL depth and crawl coverage **Indexation** - robots.txt directives and `noindex` meta tags - sitemap.xml consistency and coverage - Canonical chain resolution - `meta robots` and `x-robots-tag` handling **On-page quality** - Title length, presence and uniqueness - Meta description length and uniqueness - Heading hierarchy from H1 to H6 - Content depth and word count - Image alt text coverage **Server and security** - HTTPS enforcement and HSTS - HTTP/2 availability - SSL certificate expiry monitoring - SPF records and www canonicalization **Structured data** - JSON-LD syntax validation - Rich-result eligibility - Schema.org type coverage - Open Graph and Twitter card completeness **AI readiness** (the four rules in this category, by name) - `ai-llms-txt-missing` — no `/llms.txt` describing the site for LLMs - `ai-robots-blocked` — AI crawlers blocked by robots.txt - `ai-robots-missing` — robots.txt unavailable - `ai-structured-data-low` — low structured data coverage **Performance** (the two rules in this category, by name) - `no-compression` — response compression not enabled - `slow-response` — slow server response ### Reading an audit Each finding names the rule, the affected URL, the current value, the expected value, and a plain-language fix. Reports export as flat CSV or JSON. History is kept, so a re-crawl can be diffed against the previous run to show what was fixed, what is new, and what regressed. --- ## AI Visibility — GEO and AEO Pylon SEO measures how a brand appears in AI answers. Three distinct disciplines are reported separately rather than blended: - **SEO** — win the click. Classic rankings, click-through rate, SERP features. - **GEO** (Generative Engine Optimization) — win the recommendation: being named in an AI answer and cited as a source. - **AEO** (Answer Engine Optimization) — win the answer itself: being quoted as the definitive response rather than mentioned in passing. ### Metrics - **Mention rate** — the share of monitored prompts in which the brand appears at all. - **Citation coverage** — the share of prompts where the brand is cited as a source, tracked separately from mentions. - **Share of voice** — relative presence against named competitors, computed per provider. - **Answer position** — the role the brand played in a given answer: headline answer, supporting source, footnote, or absent. This is stored per response, not aggregated away. - **Trend score** — a composite 0–100 score per run, so movement between runs is visible. ### Answer engines | Engine | State | Notes | | --- | --- | --- | | DeepSeek V4 Flash | Live | Answered through OpenRouter; used for paid workspaces. | | DeepSeek V4 (free tier) | Live | OpenRouter free-model route used for trial workspaces. | | ChatGPT (GPT-4o) | Adapter shipped | Active when an OpenAI API key is configured. | | Anthropic Claude | Adapter shipped | Active when an Anthropic API key is configured. | | Google Gemini | Adapter shipped | Active when a Gemini API key is configured. | | Perplexity Sonar | Adapter shipped | Active when a Perplexity API key is configured. | | Microsoft Copilot | Planned | Waiting on a stable public Microsoft API. | | Google AI Overviews | Planned | Being sourced from Search Console and SERP data. | Every enabled engine adapter receives the same prompt set in parallel. Each answer is stored as its own row, so a per-engine breakdown is real data rather than an average. Recommended cadence is weekly. ### AI readiness checklist 1. AI crawler access is granted (GPTBot, OAI-SearchBot, PerplexityBot are checked). 2. Content answers the question directly near the top of the page. 3. Claims are sourced and carry statistics. 4. An `llms.txt` file is published and valid. 5. Structure is extractable without executing JavaScript. 6. The prompts that matter are monitored rather than assumed. --- ## Rank Tracking - **Markets:** United States, United Kingdom, Canada, Germany, Australia. - **Devices:** desktop and mobile, tracked as a split rather than a blended number. - **Granularity:** daily position updates. - **SERP feature detection:** AI Overviews, featured snippets, local pack, People Also Ask. - **Alerts:** configurable rank-drop thresholds per site or per keyword, with email delivery and a deduplicated trigger log so one drop does not send many emails. Rank tracking deliberately reports device splits, local SERPs and feature drift separately. A single blended "position 12" number hides all three. ## Keyword Research - **Up to 300 ideas per seed**, generated from topic templates — questions, comparisons, modifiers and local variants — and scored deterministically. - **Metrics per idea:** volume, difficulty, intent, AI-answer overlap. - **Intent classification:** informational, commercial or transactional. - **Opportunity score** (0–100): a composite of volume, intent, your domain's strength and AI-answer risk. This is deliberately not raw keyword difficulty — difficulty alone rewards queries nobody searches. - **Difficulty ceiling:** a control that filters results to a range the site can realistically compete in. - **Buckets:** Winnable, Contestable, Risky. - **AI-answer overlap (0–100):** whether an AI answer already saturates the query. Queries with high overlap are flagged for GEO/AEO work instead of classic ranking work. - **AI trending ideas:** optional LLM-generated keyword suggestions. AI supplies the keyword and its rationale only; demand figures attached to AI-sourced ideas remain modelled estimates and are never fabricated. - Works without an account. Results export to CSV. ## Competitor Analysis Four lenses, up to five competitors per project on the Growth plan: 1. **Ranking overlap** — shared, won and lost keywords, ordered by how far the competitor's ranking strength exceeds the site's. 2. **Keyword gaps** — queries a competitor ranks for that the site does not appear for, filtered by volume, difficulty and AI-answer risk. 3. **AI share of voice** — per-provider share of voice against competitors, with mentions and citations tracked as separate measures. 4. **Audit benchmarks** — technical health, schema coverage and AI readiness of competitor domains shown alongside the site's own. Gap types: **missing** (competitor ranks, site absent — highest priority), **weak** (site ranks 11–50 against a competitor's top 5), **AI-flagged** (routed to GEO/AEO instead of classic ranking), and **volume-verified** (demand cross-checked for the target market). Competitors are never invented. Auto-discovery requires stored AI-mention evidence that names the rival; a rival with no snapshot is shown as unavailable rather than as a zero. Competitor authority, traffic and organic keyword counts are snapshot estimates and are labelled as estimates. --- ## Free tools ### Free SEO Checker Paste any public URL. No account required; anonymous users get a limited number of checks per day. - **Title and meta:** title length and keyword presence, meta description quality, Open Graph title and image, Twitter card markup. - **Heading hierarchy:** single H1 enforcement, H1→H6 logical nesting, keyword in the first heading, heading count and distribution. - **Schema and Open Graph:** JSON-LD syntax validation, Schema.org type detection, rich-result eligibility, Open Graph completeness. - **Canonicals and indexability:** canonical presence and target, `noindex` flag, sitemap membership, hreflang conflicts. - **Image alt text:** missing alt attributes, descriptive versus empty alt, file size hints, lazy-load attributes. - **Performance hints:** time to first byte and response time, render-blocking resources, DOM size estimate, Core Web Vitals flags. Scores are returned for on-page, indexability, schema, performance and accessibility. Each check returns pass, warn or fail plus a plain-language fix. Note: the checker scores the HTML a server returns. Content that only appears after JavaScript execution is not scored, so JavaScript-heavy single-page applications are scored on raw HTML. Use server-side rendering or pre-rendering for those sites. --- ## Platform apps and plugins ### WordPress plugin Requires PHP 7.4+ and MySQL 8+. In-dashboard SEO for WordPress sites. - **On-page and content:** meta tags and SERP preview, content analysis, image SEO optimizer, local SEO, SEO A/B testing. - **Schema and sitemaps:** schema auto-generator, XML sitemaps, Google Search Console. - **Technical and crawl:** redirect manager, broken link checker, site health dashboard, full site audit. - **AI and automation:** AI engine, keyword clustering and AI briefs, autonomous SEO agent, workflow automation, version control. - **Tracking and migration:** rank tracker, one-click migrator, white-label reports. Mechanics: - The license key is emailed on purchase and entered under **Pylon SEO → License** in wp-admin. - Import runs from **Pylon SEO → Migrator**, covering Yoast, Rank Math, All in One SEO, SEOPress, Slim SEO, The SEO Framework, Squirrly SEO, SmartCrawl, WP Meta SEO and others. - Deactivating the plugin keeps all SEO metadata in the database. Redirects and sitemaps stop being served until it is reactivated. - Multisite: each site needs its own key, or a shared key within the plan's site limit. - The autonomous agent runs daily via Action Scheduler and emails a weekly report. - AI features call OpenAI using either your own key or purchased credits, server-side only. Page content is never used for training. - MCP server is enabled under **Pylon SEO → Settings → MCP Server**. - Optimizes for Google, Bing, ChatGPT, Perplexity and Gemini. ### Shopify app Bulk product meta optimization, JSON-LD, schema migration, automatic sitemaps, HTML sitemap and llms.txt, site verification, 404 monitoring and redirects, AI crawler log and referrals, rank tracking with SERP features, autonomous AI agent, workflow automation, GEO/AEO for AI search, AI Overview risk detection, LLM citation tracking, version control and rollback. ### Wix app Schema auto-generator, schema testing and validation, schema migration, multi-format sitemaps, meta tags and SERP preview, content analysis, image SEO, local SEO, robots.txt manager, redirect manager, link analyzer, site health dashboard, full site audit, AI crawler log, autonomous SEO agent, workflow automation, topic clusters and AI content briefs. ### Squarespace extension Scans product and store pages, generates meta tags and JSON-LD, and supplies paste-ready blocks. Includes AI meta tags, AI alt text generation, internal linking, a JSON-LD schema library, sitemap intelligence, Google Search Console, page speed analysis, multi-engine SERP preview and a keyword research hub. ### BigCommerce app Audits catalogs through the BigCommerce API, finds self-competing pages, and validates category and brand templates. Covers catalog-wide audits, template-level fixes, product and variant SEO, variant URL canonicalization, facet and filter URL control, redirect chain flattening, page speed analysis, 16+ JSON-LD schema types, offer and rich-result validation, schema migration, AI meta and content generation, keyword research, rank tracking, Google Search Console and an autonomous SEO agent. --- ## Integrations, API and automation ### Data and search consoles - **Google Search Console** — impressions, click-through rate and average position merged with crawl data. Read-only: Pylon never writes to a Search Console property. - **Bing Webmaster Tools** — Bing performance alongside Google. - **IndexNow** — publishes URL changes to Bing, Yandex and other supporting engines. ### API - Base URL: `https://api.pylonseo.io/v1/` - Bearer-token authentication. API keys are issued per workspace with workspace-scoped permissions. - Representative endpoints: `GET /v1/sites`, `POST /v1/sites/:id/scan`, `GET /v1/sites/:id/audit/latest`, `GET /v1/sites/:id/keywords`, `GET /v1/sites/:id/competitors`, `GET /v1/sites/:id/ai-visibility`. - REST API access is included on the Agency plan. ### Webhooks and exports - Webhooks deliver signed event payloads for crawl, verification and ranking events. A completed audit event carries the site id, crawl id, score, URLs crawled, error, warning and notice counts, a summary of high-impact issues including how many issues block AI crawlers, and an HMAC signature over the payload. - Exports are flat CSV or JSON. Client-facing reports are white-label, carrying the agency's own brand. --- ## Plans and pricing Prices are served from a live plan catalogue and rendered from it; they are not hardcoded in page components. | Plan | Monthly (USD) | Annual, per month (USD) | Monthly (INR) | Annual, per month (INR) | Included | | --- | --- | --- | --- | --- | --- | | Free | $0 | $0 | ₹0 | ₹0 | 1 connected site, 10 tracked keywords, 5 audits/month, 2 AI visibility runs/month, 2 reports/month, technical audits, issue tracking, instant on-demand scans, community support | | Growth | $59 | $47 | ₹4,939 | ₹3,949 | 10 connected sites, 500 keywords, 50k pages crawled/month, 50 audits/month, 25 AI visibility runs/month, 25 reports/month, daily rank tracking, AI visibility, competitor analysis, scheduled weekly audits, white-label reports, CSV and PDF export, priority support | | Agency | $149 | $115 | ₹12,699 | ₹10,149 | 30 connected sites, 1,500 keywords, 250k pages crawled/month, 200 audits/month, 100 AI visibility runs/month, 100 reports/month, REST API access, custom crawl schedules, multi-tenant client workspaces, white-label reports, role-based access control, priority support, SLA | - Paid plans include a **7-day trial with no credit card required**. - Annual billing is shown as the per-month equivalent, with the annual total charged. - **Allowances are enforced, not decorative.** Each metered dimension is blocked at its limit with an HTTP 402 upgrade prompt naming the dimension reached (`sites`, `plan_limit_exceeded`). Tracked keywords, monthly audits, AI visibility runs and reports are all gated; `pages_crawled` is reported but not gated, because it is a monthly aggregate of per-run crawl budgets and can only be checked before a run starts. - Currency is detected automatically rather than user-selected: the signed-in workspace country wins, then request IP for anonymous visitors, then browser timezone. India sees INR billed via Razorpay (UPI, cards, netbanking); other markets see USD billed via PayPal and international cards. - Platform plugins and apps are **licensed separately per product**, at the price shown on that product's page (for example the WordPress plugin has its own Starter, Pro and Agency tiers, as do the Shopify, Wix, Squarespace and BigCommerce apps). These are one-time product licenses for the site they are installed on, distinct from the SaaS plans above. - Refund eligibility windows and license handling are documented in the [Refund Policy](https://pylonseo.io/refund-policy). --- ## Company **Bytenovo** builds Pylon SEO. Registered office: C-58, Sector 12, Dharam Colony, Palam Vihar Extension, Gurugram, Haryana 122022, India. Four operating principles: 1. **Measured versus estimated figures are always labelled.** A modelled number presented as a measurement is a defect, not a rounding issue. 2. **A crawler that follows the rules.** robots.txt is read and honoured, and outbound requests are SSRF-checked. 3. **Your data stays yours.** Site data is owned by the workspace that produced it, scoped to that workspace, and never sold. 4. **Plain language over jargon.** Contact: [contact@pylonseo.io](mailto:contact@pylonseo.io) · +91 97179 30424 (Mon–Sat, 10:00–20:00 IST). --- ## Policies - **[Privacy Policy](https://pylonseo.io/privacy)** — accounts, usage logs, payments, and the rights available to request access, correction or deletion. - **[Terms of Service](https://pylonseo.io/terms)** — the terms governing use of the platform and the WordPress plugin. - **[Cookie Policy](https://pylonseo.io/cookies)** — Pylon stores functional data only. There is no cross-site advertising or tracking cookie. A consent choice is recorded locally for seven days; nothing on the site changes behaviour based on it. - **[Refund Policy](https://pylonseo.io/refund-policy)** — eligibility, licenses and requests. --- ## Verifying these claims Everything asserted above is checkable in the product or on this site: - The 35 audit rules are the live rule catalogue, grouped by category, behind every audit. - Engine availability follows the configured API key for each provider — an adapter is shipped but inert until its key is set. - Prices are read from the plan catalogue at request time. - `robots.txt`, `sitemap.xml`, `llms.txt` and `llms-full.txt` are all directly fetchable at the site root.