Domain checks
How scoring screens every domain, and the one-word reason returned for each kind of rejected input.
Before anything is scored or charged, every company domain is screened, whether you send it to /score/domain or we find it from an email or LinkedIn URL. Inputs that aren't a company's own website are rejected with HTTP 400 and a one-word reason you can switch on in code:
{ "error": "invalid_domain", "reason": "social_media" }In a bulk call, each rejected domain is listed in rejected with its reason, and the rest are still scored:
{ "rejected": [ { "domain": "logo.png", "reason": "file_name" } ] }Rejected domains are never charged.
Reasons
Instant checks
reason | Rejected input | Examples |
|---|---|---|
empty | Nothing entered | "", " " |
invalid_format | Not shaped like a domain | acme, hello world.com, -bad-.com |
invalid_tld | The ending isn't a real domain ending | acme.con, acme.comm |
email | An email address | jane@acme.com |
ip_address | An IP address, with or without a port or path | 192.168.1.1, http://10.0.0.1:8080/x, 2001:db8::1 |
file_name | A file name or path | logo.png, Document.pdf, Presentation.pptx, C:\files\deck.pptx |
social_media | A social network | linkedin.com/company/acme, instagram.com/acme, facebook.com, x.com, youtube.com |
messaging | A chat or meeting link | wa.me/9198…, chat.whatsapp.com/…, t.me/acme, discord.gg/… |
common_site | A well-known site, not a prospect | google.com, google.co.in, wikipedia.org, chatgpt.com, example.com |
email_provider | A generic or temporary email service | gmail.com, outlook.com, mailinator.com |
link_shortener | A short-link or link-in-bio service | bit.ly/x, tinyurl.com/x, linktr.ee/acme |
marketplace | A marketplace, directory or app store | indiamart.com, amazon.de, crunchbase.com, play.google.com |
free_host | A free website on a hosting or site-builder domain | acme.myshopify.com, acme.wixsite.com, acme.vercel.app |
Live checks
These look at the domain on the internet before scoring.
reason | Rejected input |
|---|---|
unregistered | The domain doesn't exist: expired or never registered |
no_website | Registered, but no website answers, the page is blank, or it shows a default server page |
parked | Parked or for sale (redirects to a domain marketplace, or a "this domain is for sale" page) |
coming_soon | A "Coming soon" or "Under construction" page |
Bulk only
reason | Meaning |
|---|---|
already_processing | A single scoring job for this domain is already running |
none_valid | Whole call: no domain in the list passed the checks |
How the checks work
- Cleaning first:
https://www.Acme.com/about?x=1is checked asacme.com. Paths andwww.are fine. - Subdomains count:
in.linkedin.com,m.facebook.comandx.tradeindia.comare recognised as their parent site. - Real endings only: the domain ending is checked against the official list of every top-level domain.
- File names vs. domains: endings that are both a file type and a real domain ending, such as
.aiand.app, are treated as domains, sox.aipasses. - Live checks never block a real company by accident: a slow site, a site behind bot protection, or a site with a temporary server error is accepted. A domain is rejected only on clear evidence.
Handling reasons in code
const res = await fetch(`${BASE}/score/domain`, { method: "POST", headers, body: JSON.stringify({ domain, template_id }) });
const body = await res.json();
if (res.status === 400 && body.error === "invalid_domain") {
switch (body.reason) {
case "social_media": case "messaging": return "That's a social or chat link. Enter the company's website.";
case "file_name": return "That looks like a file name.";
case "unregistered": case "no_website": return "This domain has no website.";
case "parked": return "This domain is parked or for sale.";
case "coming_soon": return "This website isn't live yet.";
default: return "Enter the company's own website, like acme.com.";
}
}