Zero schema. Zero organic footprint. And the domain isn't yours.
You're building a vendor's SEO, not your own.
edyson.isellfl.me is a white-label IDX page for Edyson Tavarez (REALTOR®, LPT Realty, FL license SL3420617) on the shared isellfl.me platform. The homepage is a React single-page app: ~62 words of real text, a generic <title>Home Search</title> shared with every other agent on the platform, zero JSON-LD structured data, and a robots.txt that blocks crawlers from everything except the homepage and /search. DataForSEO returns zero ranked keywords for this subdomain — and the parent platform itself ranks for only 39. Whatever authority this page earns flows to the vendor's root domain, not to Edyson.
Domain: edyson.isellfl.me · Business: Real estate agent · LPT Realty · FL Lic. SL3420617 · Shana Gates / Military Veteran Realtor Team · Service area: Tampa Bay, FL (Tampa 813 · Pinellas 727)
Read this if nothing else.
The tech hygiene is fine (HTTPS, CSP, stable layout). The SEO fundamentals for a personal-brand agent are not: you're on a vendor subdomain, the homepage title is generic and shared, there's no schema, and robots.txt blocks crawlers from your content. None of this builds Edyson's brand in search — it builds isellfl.me's.
The domain is the core problem
edyson.isellfl.me is a subdomain on a vendor's shared IDX platform. Every link earned, every page indexed, every signal Google records accrues to isellfl.me — not to you. You cannot build durable, portable search equity on a subdomain you don't own. If you switch IDX vendors or LPT relationships, the URL and any ranking it earned disappear. The fix is a custom domain (e.g. edysontavarez.com) that you own, with the IDX search embedded or pointed at it.
Zero structured data
Grep on the rendered DOM returns no JSON-LD @type at all. No RealEstateAgent, no Person, no LocalBusiness. For an agent, RealEstateAgent + Person schema (name, license SL3420617, phone, areaServed, sameAs to your Google profile + social) is the single highest-leverage on-page fix — it's how Google and AI engines connect this page to the real human Edyson Tavarez.
robots.txt blocks almost everything
The file reads User-agent: * → Disallow: / with only Allow: /$ and Allow: /search. That blocks crawlers from any agent bio, about, neighborhood, or content page. For a content-driven personal brand this is fatal — Google is only permitted to index your homepage and the search shell.
Generic, shared homepage title
<title>Home Search</title> — 11 characters, no name, no city, no brand. It's the platform default, identical across agents. The meta description ("Home Search and latest real estate listings. Find your perfect home with us!") is equally generic. Nothing tells Google this is Edyson Tavarez in Tampa Bay.
No real XML sitemap
/sitemap.xml, /sitemap_index.xml and /wp-sitemap.xml all return the 584 KB React app shell (HTTP 200, zero <urlset>/<loc> tags), and robots.txt declares no Sitemap: line. There is no machine-readable sitemap for any crawler to follow.
og:url leaks the port
<meta property="og:url" content="https://edyson.isellfl.me:443/"> — the explicit :443 is a platform bug. Social shares and canonical signals should never carry the port.
Mobile zoom disabled
<meta name="viewport" ... maximum-scale=1 user-scalable=no> blocks pinch-zoom — an accessibility failure (WCAG 1.4.4) Lighthouse flags, and a usability problem for older buyers.
Six fixes. The first is the whole ballgame.
| # | Action | Why it matters | Effort |
|---|---|---|---|
| 1 | Get on a domain you own. Register edysontavarez.com (or similar), and either (a) point the IDX vendor's white-label at your custom domain, or (b) build a real agent site there and embed/iframe the IDX search. Ask isellfl.me support whether custom-domain mapping is supported on your plan — most IDX vendors offer it. | This is the difference between building your own search equity and renting space on a vendor's. Everything below compounds 10x once it lives on a domain that's yours and that you can take with you. | 2-4 hr + DNS |
| 2 | Add RealEstateAgent + Person JSON-LD. Name (Edyson Tavarez), jobTitle REALTOR®, license SL3420617, phone, areaServed (Tampa Bay cities), worksFor LPT Realty, and sameAs links to your Google Business Profile, Zillow, and social profiles. | Zero schema means Google and AI search have no structured way to identify the human behind this page. This is what makes you eligible for rich results and AI-Overview citation for 'Realtor in Tampa' style queries. | 2-3 hr (vendor ticket) |
| 3 | Fix robots.txt — stop blocking your own content. Replace the blanket Disallow: / with targeted disallows (e.g. /search/map/) and Allow: / for everything else. Add a real Sitemap: line. | Right now Google is only allowed to index your homepage and the search shell. Any bio, neighborhood, or blog content you add will be invisible until this is fixed. | 30 min (vendor ticket) |
| 4 | Write a real title + meta description. Title: 'Edyson Tavarez, REALTOR® — Tampa Bay Homes for Sale | LPT Realty'. Meta: lead with name, city, military-veteran angle, and a call to action. | 'Home Search' tells Google nothing. A name + geo title is free ranking weight on every branded and local query. | 15 min (vendor ticket) |
| 5 | Ship a real XML sitemap + fix the og:url port. Generate a proper sitemap.xml (homepage + any indexable agent pages) and reference it in robots.txt. Strip :443 from og:url. | Crawl hygiene. No machine-readable sitemap = crawlers rely on links, and you have only 2 internal links on the homepage. | 1 hr (vendor ticket) |
| 6 | Add real, indexable content + AI-crawler access. Agent bio, Tampa Bay neighborhood guides, buyer/seller FAQs — on the owned domain. Add the 14 AI-crawler allows + an llms.txt once robots.txt is unblocked. | 62 words of body text can't rank for anything. Content is the only thing that earns the long-tail and AI-citation queries that bring buyers. | ongoing |
What the live HTML actually says.
Live capture from edyson.isellfl.me on 2026-06-02 (curl + JS-rendered Lighthouse + DataForSEO). Stack: Express (Node.js) backend serving a React single-page app, content from Contentful CMS, hosted behind AWS CloudFront/S3.
Stack detected
| Platform | isellfl.me white-label IDX (shared vendor subdomain) not an owned domain |
| Backend | Express / Node.js (x-powered-by: Express) |
| Frontend | React SPA (client-rendered) + Contentful CMS |
| Security headers | CSP, HSTS-adjacent, X-Frame SAMEORIGIN, nosniff present |
| robots.txt | Disallow: / (blocks all but homepage + /search) |
| Sitemap | none real — /sitemap.xml returns the SPA shell |
| Trackers | GTM, GA, Mixpanel, Facebook, TikTok, Bing Ads, DoubleClick, Adara, freshpaint, listtrac (10+) |
Vendor subdomain — no owned equity
All SEO authority flows to isellfl.me, not to Edyson. The URL is not portable across vendor/brokerage changes.
evidence: hostname edyson.isellfl.me; contentfulPartyWebsite domain = edyson.isellfl.me
Zero JSON-LD structured data
No RealEstateAgent, Person, or LocalBusiness schema anywhere in the rendered DOM.
evidence: grep 'application/ld+json' rendered DOM → 0 blocks
robots.txt blocks all but homepage + /search
Blanket Disallow: / with only Allow: /$ and Allow: /search permits crawlers to index almost nothing.
evidence: robots.txt: 'User-agent: *' then 'Disallow: /' + 'Allow: /$' + 'Allow: /search'
Generic shared title + meta
Title 'Home Search' (11 chars) and the meta description are platform defaults identical across agents. No name, city, or brand.
evidence: <title>Home Search</title>; description 'Home Search and latest real estate listings...'
No real XML sitemap
sitemap.xml / sitemap_index.xml / wp-sitemap.xml all return the React shell (200, no urlset/loc). robots declares no Sitemap.
evidence: curl /sitemap.xml → 584KB HTML, 0 <loc> tags
og:url contains :443
Open Graph URL carries the explicit port — a platform bug affecting share canonicalization.
evidence: og:url content='https://edyson.isellfl.me:443/'
Thin content — 62 words
Plain-text content is ~62 words / 390 bytes against a 600KB DOM. Only 2 internal links.
evidence: on-page: plain_text_word_count 62, internal_links_count 2
viewport disables zoom
maximum-scale=1, user-scalable=no blocks pinch-zoom (WCAG 1.4.4).
evidence: <meta name=viewport ... user-scalable=no>
Desktop LCP 4.74s on a 3.8 MB page. Mobile will be worse.
Lighthouse mobile lab data, captured 2026-06-02. Google's Good zone: LCP <2.5s, FCP <1.8s, CLS <0.1, TBT <200ms.
Heavy SPA + 10+ trackers, measured on desktop
Lighthouse ran desktop (form factor desktop, no CPU throttle) and still scored 56 with a 4.74s LCP and a 3.8 MB total payload. Layout is stable (CLS 0.01) and TBT is borderline OK (196ms), so the issue is sheer weight: a client-rendered React app plus 28 scripts and 10+ analytics/ad trackers (GTM, GA, Mixpanel, Facebook, TikTok, Bing, DoubleClick, Adara, freshpaint, listtrac). On a throttled mobile connection — how most buyers browse — expect materially higher LCP. Most of the tracker load is vendor-controlled.
Zero structured data is the biggest on-page miss.
No JSON-LD of any type. For a named real-estate agent, this is the difference between being a person Google understands and being an anonymous search box.
What's missing: RealEstateAgent, Person (name = Edyson Tavarez, license SL3420617), LocalBusiness / Organization (LPT Realty), PostalAddress, areaServed, AggregateRating, and sameAs links to Google Business Profile + Zillow + social.
Why it matters most for an agent: real-estate is a person-trust business. Schema is how Google's Knowledge Graph and AI engines (ChatGPT, Perplexity, AI Overviews) tie a web page to the real, licensed human. With zero schema, a query like 'Edyson Tavarez Realtor Tampa' has nothing structured to anchor to.
Are you findable by ChatGPT, Perplexity, Google AI Overviews?
As of 2026, 22% of all search queries route through an AI surface. Visibility there is now a measurable revenue lever.
robots.txt: AI crawler access
The blanket Disallow: / blocks AI crawlers from everything but the homepage and /search, and none of the 14 standard AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, etc.) are explicitly named. Fix the foundational robots.txt first, then add explicit AI-crawler allows.
llms.txt: citation hygiene file
/llms.txt returns HTTP 200, but with no real indexable content and a crawl-blocked site, there's nothing for an LLM to cite. Ship a real llms.txt (agent bio, service areas, citation guidance) once content + robots are fixed.
How your business appears in local search.
Verifiable on-page: Edyson Tavarez, REALTOR®, FL license SL3420617, LPT Realty, associated with the Shana Gates / Military Veteran Realtor Team. Phones (727) 300-6681 (Pinellas) + (813) 336-7967 (Tampa) place the business in Tampa Bay.
The raw signals are real — a licensed agent, a brokerage, a team, two Tampa Bay phone numbers, a veteran-owned angle that's a genuine differentiator. None of it is expressed in structured local-SEO data.
For local real estate, the wins are: (1) a RealEstateAgent entity with full NAP + license + areaServed; (2) a verified, optimized Google Business Profile linked via sameAs; (3) the military-veteran positioning surfaced as on-page content (VA-loan buyer guides are exactly the long-tail AI Overviews cite). All three depend on first getting onto an owned domain with crawlable content — which is why the domain fix leads the roadmap.
The order to do this work in.
Ranked by impact divided by effort. Estimates assume a competent WordPress developer.
| # | Action | Effort | Impact |
|---|---|---|---|
| 1 | Migrate to an owned custom domain (map IDX or embed search) | 2-4 hr | Extreme (foundational) |
| 2 | Add RealEstateAgent + Person JSON-LD (license, NAP, sameAs) | 2-3 hr | Very high |
| 3 | Fix robots.txt — stop blocking content + declare sitemap | 30 min | Very high |
| 4 | Real title + meta description (name + geo + brand) | 15 min | High |
| 5 | Ship a real XML sitemap + strip og:url port | 1 hr | Medium-high |
| 6 | Add indexable agent + neighborhood content | ongoing | High (long-tail + AI) |
| 7 | Verify + link Google Business Profile via sameAs | 1-2 hr | High (local pack) |
| 8 | Add 14 AI-crawler allows + llms.txt (after robots fix) | 45 min | Medium |