Engines can read, match, and cite this business — a few named gaps are still costing mentions. Every lost point below is traced to a check you can read.
https://savugaragedoorrepair.com/ · Checked Nov 2, 2026
Checked 10 pages — 24 checks, each scored on the evidence found:
Readiness score from on-site signals — not a measurement of live ChatGPT / Perplexity / Google AI answers.
Most of the 19 lost points come from review profile links, review / rating schema, and llms.txt published:
The full ranked list is in the fix list below; every point traces to a check you can read in the audit.
30 JSON-LD blocks found across the checked pages; readable structured data parsed on 10 of the 10 pages.
Schema uses the generic type LocalBusiness — valid, but it does not say what the business does.
Why it matters: A specific type ("Plumber", "RoofingContractor") tells an engine exactly what the business is and which local questions it answers. A generic Organization label leaves the engine to infer the trade.
Next step: Use a specific subtype (e.g. GeneralContractor, Plumber, RoofingContractor) so engines know the trade, not just that a business exists.
Source: the business entity’s @type in the crawled JSON-LD · Confidence: Direct read
The business schema carries 7 of 8 key fields (name, phone, street address, URL, opening hours, geo coordinates, service area) — missing: profile links (sameAs).
Why it matters: Each field is a question an engine can then answer without guessing — the phone for "call", the street for "near me", hours for "open now". Missing fields are questions the engine answers with someone else’s data.
Next step: Add the missing fields to the business schema: profile links (sameAs).
Source: the most complete business entity found in the crawled JSON-LD · Confidence: Direct read
FAQPage schema found — engines can lift the Q&A directly.
Service schema found — the services themselves are described as structured data.
No AggregateRating or Review schema found on any checked page.
Why it matters: Marked-up ratings give engines a citable proof number. A rating that exists only on Google’s own pages never reaches the answers assembled from the open web.
Next step: Once real reviews are published on the site, mark them up with Review / AggregateRating schema that matches them exactly.
Source: JSON-LD @type values across the pages read · Confidence: Direct read
robots.txt is live and readable (182 bytes).
No rule blocks the major AI crawlers — GPTBot: allowed · PerplexityBot: allowed · ClaudeBot: allowed.
Sitemap found at /sitemap_index.xml (an index) listing 23 pages.
Why it matters: The sitemap is the map handed to crawlers. Pages missing from it are found late, by accident, or not at all — including the service pages that carry local questions.
Next step: Grow the sitemap to cover every public page — services, cities, guides, and posts — so crawlers find the money pages.
Source: the sitemap file listed in the appendix · Confidence: Direct read
No llms.txt found (HTTP 404).
Why it matters: llms.txt is the short, plain-text brief written for AI tools: who the business is, what it does, and which pages matter. It is the cheapest signal on this list to publish.
Next step: Publish an llms.txt at the site root: what the business does, where it works, and links to the key pages.
Source: /llms.txt · Confidence: Direct read
No llms.txt to evaluate.
Why it matters: A stub llms.txt tells an AI tool almost nothing. One that links the money pages — services, cities, guides — steers the tool to the pages worth citing.
Next step: Give llms.txt real substance — a business summary plus links to services, cities, contact, and FAQ pages.
Source: the body of /llms.txt · Confidence: Direct read
Homepage declares a canonical URL; site serves clean HTTPS.
37 question-and-answer blocks found across the checked pages (37 question-style headings; no dedicated FAQ page found).
5 service+city / city pages found (0 combined service+city, 5 dedicated city pages) — e.g. /long-beach-ca/, /torrance-ca/, /glendale-ca/; 43 service pages without a city in the address.
Homepage runs 2,421 words; the main services page runs 1,516 words. Thin pages give engines little to cite.
Homepage has 1 H1 heading and 9 descriptive H2 sections.
10 of 10 checked pages carry a clean page title and a meta description in a sane length.
68 short factual blocks (tight paragraphs and lists) found on the homepage/services page — the shapes engines lift answers from.
Phone +1-213-600-0498 is identical in the schema and the visible pages checked. Business name "SAVU Garage Door Repair" matches the name used on the site. Street address "3549 Chesapeake Ave" matches in the schema and the visible pages.
tel: click-to-call links found — a phone tap on mobile reaches the business directly.
A contact page is published with a contact form and an email link and a phone link.
No links to review profiles (Google, Yelp, Houzz, BBB…) found on the checked pages.
Why it matters: Links to Google, Yelp, and the trade directories connect the site to its proof. Engines cite businesses they can corroborate — an unlinked profile is corroboration they may never find.
Next step: Link the Google and Yelp review profiles from the footer and contact page so engines can cite real proof.
Source: outbound links in the pages read · Confidence: Direct read
Testimonial content found on the site (128 testimonial blocks/sections) — visible proof, not just a rating number.
No Google Business Profile or Maps profile link found on the checked pages.
Why it matters: A published link to the Google Business Profile ties the website to the listing that powers maps and "near me" answers, closing the loop engines use to confirm the business is real and current.
Next step: Link the Google Business Profile from the footer and contact page.
Source: outbound links in the pages read · Confidence: Direct read
Review profile links
What we found: No links to review profiles (Google, Yelp, Houzz, BBB…) found on the checked pages.
What to do: Link the Google and Yelp review profiles from the footer and contact page so engines can cite real proof.
Review / rating schema
What we found: No AggregateRating or Review schema found on any checked page.
What to do: Once real reviews are published on the site, mark them up with Review / AggregateRating schema that matches them exactly.
llms.txt published
What we found: No llms.txt found (HTTP 404).
What to do: Publish an llms.txt at the site root: what the business does, where it works, and links to the key pages.
Business type named specifically
What we found: Schema uses the generic type LocalBusiness — valid, but it does not say what the business does.
What to do: Use a specific subtype (e.g. GeneralContractor, Plumber, RoofingContractor) so engines know the trade, not just that a business exists.
llms.txt has substance
What we found: No llms.txt to evaluate.
What to do: Give llms.txt real substance — a business summary plus links to services, cities, contact, and FAQ pages.
Google Business Profile link
What we found: No Google Business Profile or Maps profile link found on the checked pages.
What to do: Link the Google Business Profile from the footer and contact page.
Schema field completeness
What we found: The business schema carries 7 of 8 key fields (name, phone, street address, URL, opening hours, geo coordinates, service area) — missing: profile links (sameAs).
What to do: Add the missing fields to the business schema: profile links (sameAs).
Sitemap published and useful
What we found: Sitemap found at /sitemap_index.xml (an index) listing 23 pages.
What to do: Grow the sitemap to cover every public page — services, cities, guides, and posts — so crawlers find the money pages.
Email me this report (PDF) + the fix checklist. Everything above stays free and visible — the email is for delivery, the checklist, and a nudge to re-check after you fix things.
This list is the work. AltaPoint can take it from there — it starts with a free Snapshot, then the $999 Full Audit turns the fix list into a measured report with a ranked fix order.
Get your free SnapshotNot sure what else is leaking? The AI visibility score is one leak of several. Get your free Snapshot — a few minutes, and it covers the operational side too.
Developers & agencies: this checker is coming as an API. Join the API waitlist — one email when it opens, nothing before.
Fixed something? Re-check in 24h — results are cached for a day so the score reflects real changes, not refresh luck.
Rubric: AEO Score v2 — 24 checks in 4 sections, 100 points · Run date: Nov 2, 2026 · Read time: 10.1s. The audit fetches the pages and files listed above once, scores only what they contain, and shows the evidence next to every check.
This is a readiness score computed from on-site signals only — the schema, files, content, and links published on the pages listed above. It is not a measurement of live ChatGPT, Perplexity, or Google AI answers, and it does not claim the business does or does not appear in any engine’s responses. It measures whether the site publishes what those engines rely on when they choose what to cite.
AEO (Answer Engine Optimization) is how ready your website is for AI answer engines — ChatGPT, Perplexity, and Google AI Overviews. When someone asks an AI tool for a contractor or local service, the engine reads structured signals on business websites and cites the ones it can understand. AEO is making sure your site is one of them.
No. This is a readiness score from on-site signals — not a measurement of live ChatGPT / Perplexity / Google AI answers. It measures whether your site publishes the schema, Q&A, contact details, service pages, reviews, and machine files that answer engines rely on when they choose what to cite.
Up to 10: your homepage plus the key pages found from your sitemap and internal links — services, contact, FAQ, about, and service/city pages. The report states exactly how many pages were checked and lists them, and every one of the 24 checks shows the evidence it was scored on.
Most low scores come down to a few fixable gaps: no LocalBusiness schema, no FAQ page engines can lift answers from, a missing street address or phone in machine-readable form, no dedicated service and city pages, or no sitemap. Your report lists every gap found, in order of points recoverable.
llms.txt is a plain-text file at the root of your site (yoursite.com/llms.txt) written for AI tools: what the business does, where it works, and links to the key pages. Paired with a sitemap.xml, it hands crawlers and answer engines a map instead of making them guess.
Answer engines need some machine-readable location signal to match you to a place. A published street address in your schema earns full credit. If you work from home or cover an area, publish an explicit service area in your schema instead — the check explains which signal it found, and the fix list says exactly what to publish.
Re-check after you fix a gap, and then every few months. Results are cached for 24 hours, so a re-check inside a day shows the same report with its timestamp; after 24 hours the check fetches your site fresh.
You see the full report instantly — no signup. If you want it in your inbox, the email gate sends the report and the fix checklist. If you want the gaps fixed for you, that is the AltaPoint path: a free Snapshot, and the $999 Full Audit that turns the fix list into a measured report with a ranked fix order.
More free resources: see every free tool and guide we publish — including the Trades API Directory and the Long Beach Home Cost guides.
Curious what real checks look like? The AEO Scoreboard publishes actual scores from real runs of this checker — including our own site’s 86 and the biggest gap found in each report.
