Free tool · 30 seconds

How citable is your site?

We score your site across 6 pillars and ask ChatGPT, Claude, and Perplexity if they name you.

We'll email the results. One audit per email per 24 hours.

No credit card · Results on-screen and in your inbox · One audit per email per 24h

0/ 100
Sample report — composite score

What you get

A score, a snapshot, a fix list.

  • ›Composite 0–100 across 6 pillars
  • ›Live citation snapshot from ChatGPT, Claude, Perplexity
  • ›Ranked action list — impact-ordered, not generic
  • ›Schema findings, real Core Web Vitals, Position recommendation

What we test

Six pillars, mapped to The Citation Stack.

Technical SEO

Foundation

Measure

HTTPS, status codes, canonical, sitemap, robots.txt, AI-bot allowlist (GPTBot, ClaudeBot, PerplexityBot), llms.txt, lang attribute.

Why

If a model's crawler cannot fetch the page, every other signal is moot. AI-bot allowlists are the single most-skipped Foundation item on contractor sites in 2026.

Common failure

robots.txt with a wildcard Disallow that blocks GPTBot. The owner has no idea the file even exists.

Tested query

“24-hour emergency HVAC Hicksville NY”

On-Page SEO

Foundation

Measure

Title tag length, meta description, H1/H2/H3 hierarchy, image alt coverage, OpenGraph and Twitter cards, internal link count.

Why

On-page is the second-cheapest layer of the Stack. It does not win citations on its own — but a broken H1 or missing meta description tanks the rest of the diagnostic.

Common failure

Three H1s on the homepage (one in the logo, one in the hero, one in the nav). Title tag set to “Home — Acme Plumbing.” Six words. No location, no service, no qualifier.

Tested query

“sewer line replacement Suffolk County cost”

Content & E-E-A-T

Authority

Measure

Word count, readability, freshness signals, author byline and bio, topical depth, source-grounded content density.

Why

Authority is what gives a model a reason to quote you over a competitor. A page can be perfectly crawlable and still uncitable if nothing on it has the texture of expertise.

Common failure

An About page with a stock photo and three vague paragraphs. No license number, no years in trade, no certifications, no Person schema. Invisible to a model.

Tested query

“who is the most experienced kitchen remodeler in Suffolk County”

Performance & UX

Foundation

Measure

Real Core Web Vitals from Google PageSpeed Insights — LCP, CLS, INP. HTML payload size, inline JS/CSS bytes, compression, request count.

Why

Slow sites get retrieved less. Less retrieval means fewer citations. LCP under 2.5s, CLS under 0.1, INP under 200ms is the floor.

Common failure

A 14MB hero video that pushes LCP to 7 seconds on mobile 4G. The site looks premium and ships zero traffic.

Tested query

“best kitchen remodeler Long Island under $80K”

Local SEO

Citations

Measure

LocalBusiness JSON-LD, NAP detection, click-to-call links, embedded Google Map, Service schema, FAQPage schema, BreadcrumbList.

Why

Local-intent queries pull heavily from Google Business Profile and structured local signals. Without LocalBusiness schema and a clean NAP, you do not exist to the local-aware model.

Common failure

Phone number is rendered as an image. Address is in the footer as plain text. No LocalBusiness JSON-LD anywhere. The model has no anchor to attach the business to a place.

Tested query

“electrician Huntington NY for panel upgrade”

AEO / GEO

Tracking + Authority

Measure

llms.txt published, FAQPage / HowTo / Article schema density, question-format heading ratio, citation-worthiness, entity clarity, sameAs links, extractable formats (lists, tables).

Why

This is the wedge. AEO/GEO is the layer almost no Long Island contractor scores above 30 on right now. The contractors who deploy this in 2026 own the citation surface for a decade.

Common failure

No llms.txt. No FAQPage schema. No comparison tables. No question-format H2s. The page reads as prose with no structure a model can extract.

Tested query

“ductless mini-split installer Nassau County financing”

How it works

Six steps. About 30 seconds.

  1. 01

    Submit your domain, email, subcategory, and location

    Fill the four fields at the top of the page. The subcategory dropdown picks a buyer-intent query template set tuned to your trade. Location is interpolated into the queries — be specific (“Suffolk County” or “Hicksville NY” beats “New York”).

  2. 02

    We crawl your site

    We fetch the page, parse HTML structure, schema, llms.txt, robots.txt, headers, and OG tags. Real Core Web Vitals come from Google PageSpeed Insights against your live URL. No cached data, no synthetic environment.

  3. 03

    We query Anthropic and OpenAI live with your buyer-intent queries

    We run a tested template set for your subcategory against ChatGPT and Claude in real time. We parse the responses for your business name and source URL, then compute a citation-rate snapshot.

  4. 04

    An AI executive summary scores your site

    Anthropic synthesises the crawl plus the citation snapshot into a numbered score across all 6 pillars and the 5 layers of The Citation Stack. The output is opinionated, not a checklist dump.

  5. 05

    Results render on-screen and arrive in your inbox

    The full report appears on a token URL. The same link emails to you so you can re-open it later or forward it to a partner. PDF export ships in the next two weeks.

  6. 06

    Optional: book a 15-minute call to walk through the fix roadmap

    If the score is below 60 and you do not want to fix it yourself, we run through the prioritized actions on a quick call and price the work. No pitch deck, no follow-up sequence — just the roadmap.

Common questions

Audit FAQ.

Yes. Submit a URL, an email, your subcategory, and your service area, and we run the full 6-pillar audit plus a live Anthropic and OpenAI citation snapshot. The report renders on-screen in about 30 seconds. No credit card. No demo required. Each audit costs us roughly ten cents in API calls, which is why we rate-limit to one audit per email per 24 hours and fingerprint by IP — to keep the tool free without subsidising scraping bots.

We email the results immediately and add you to a low-volume Long Island AEO list — typically two emails per month covering audit-score patterns, trade-specific case studies, and Citation Stack changes. Unsubscribe is one click and instant. We do not sell, share, or rent the list. We do not call unless you book a call yourself. The full privacy posture lives at /privacy.

We query Anthropic and OpenAI live, in real time, with the exact buyer-intent queries you would compete on for your subcategory and location. The snapshot reflects what the models said on the day of the audit. Citation rate moves daily as models refresh retrieval indexes, so treat the snapshot as a directional reading, not a permanent score. Re-running the audit 30 days into a Foundation deploy is the cleanest way to see lift.

Yes. The audit runs against any public URL. Drop in a competitor domain, pick the subcategory and location they actually compete on, and you will see the same six-pillar breakdown plus the citation snapshot. We see this used most often by owners doing competitor research before a sales call or before locking their own Position. Just be aware that the email field still goes to your inbox — the prospect being audited gets nothing.

Because a generic SEO audit run against “best contractor near me” is useless. Citation rate only matters against the queries a real buyer would type. The subcategory dropdown picks a tested set of templates — “best HVAC {location} NY”, “{location} kitchen remodel cost”, “emergency plumber {location}” — and the location field is interpolated into them. The query set you get tested against is the one a Suffolk-County homeowner actually runs in ChatGPT this month.

Read the snapshot in two passes. Pass one: which buyer-intent queries from your real Position are you missing on? Those are the ones that map to revenue. Pass two: which queries are you ranking on but do not actually want? Those usually point to a Position drift — a “bath remodel” mention bleeding into kitchen-only positioning, for example. Both passes are useful. The pillar scores tell you what to fix; the citation snapshot tells you what to defend and what to abandon.

Yes. Results live in our database keyed to a token URL — that is what powers the on-screen report and the link in your email. We retain audit results so you can re-open the report later, share it with a partner, or compare against a follow-up audit in 30 days. We do not publish, index, or share individual audit results. If you want yours deleted, email logansawyer108@gmail.com and we wipe the record within 24 hours.

Yes. Every audit gets a unique view link delivered in the results email. Forward the email or copy the link — your partner sees the same on-screen report without re-running the audit. A downloadable PDF version is on the roadmap and ships in the next two weeks. Until then the view link is the canonical share artefact, and it stays live indefinitely so it works for board reviews and contractor-to-marketing handoffs.

Run yours now.

Free, on-screen results in 30 seconds. If you score below 60, the next steps are our process and pricing.

Run the auditBook a 15-min walkthrough

By Logan Hoffman, founder of Silvo Systems · April 29, 2026