Company
A campaign goes out, the bounce rate spikes, and the sending domain takes the hit: mail starts landing in spam and someone has to explain why. BounceIntel exists to stop that list from ever being sent, and to show exactly why each address was removed.
01Why this exists
Every team that sends email eventually meets the same wall. A campaign goes out, the bounce rate spikes, the sending domain takes the reputational hit, and someone has to stand up in a meeting and explain what went wrong. The verification tool that was supposed to prevent it returns a colour and a percentage, and neither of those is an explanation.
BounceIntel is built on the opposite assumption: that the interesting part of a verification is the evidence, not the conclusion. Every check returns the reason codes that produced its verdict, the signals behind them, and a confidence level that is honest about how much the pipeline actually managed to learn.
That changes what the tool is for. It stops being a filter you trust blindly and becomes a source you can quote: in a deliverability review, in a data protection assessment, or to the colleague who wants to know why four thousand addresses were suppressed.
02How it works
The same pipeline runs behind the free checker on our home page, the dashboard, a bulk upload and the API. There is no premium path that checks harder.
The address is parsed against the rules of the standard, not against a regular expression someone found once. Where a domain looks like a near miss for a large provider we say so rather than failing silently. A typo caught here is a credit not spent downstream.
We resolve the domain and read its mail exchanger records. A domain with no MX cannot receive mail at all, which is a definitive answer available before a single connection is opened, and the cheapest class of bad address there is to remove.
Large providers enforce their own local-part rules (minimum lengths, disallowed characters, reserved words) and reject addresses that break them whether or not a mailbox could exist. Applying those rules before an SMTP conversation saves the round trip and produces a more specific reason code.
We open a conversation with the receiving server and ask about the recipient, without ever delivering a message. That is where catch-all configurations, disabled mailboxes and full inboxes reveal themselves, and where a server that answers identically for every address gives itself away.
The signals are combined into a score, a category, a recommended action and the reason codes that drove them. Where the evidence does not support a confident answer, the result says unknown. That is a finding, not a failure, and treating it as one is how you avoid suppressing good addresses.
03Principles
Most of these cost us something. That is roughly the test of whether a principle is real.
A catch-all domain genuinely cannot be resolved from the outside. We label it and tell you what that means for your risk, rather than guessing so the dashboard can look decisive.
Nothing you submit is used to build a directory, enrich a product we sell, or train anything. Bulk addresses are never written to our database, only to the report you download, which expires.
The free check on the home page runs the same code as an enterprise API call. There is no cheaper path with lower accuracy hiding behind a lower price.
Reason codes are part of every response, not a paid add-on and not a debug flag. If we cannot tell you why, the verdict has not earned your trust.
Scraped lists, purchased lists and address enumeration are refused, and accounts doing them are closed. A verification service that does not police its input becomes a tool for the people it should be defending against.
Reports expire on their own. Deleting one deletes the addresses, because there is no second copy sitting in a database waiting to be forgotten about.
04By the numbers
Volume is not a proxy for accuracy, but it is what makes accuracy measurable: the more mailboxes a pipeline has seen, the better it knows how each provider behaves.
05FAQ
An independent company operating in the European Union. The service you call at api.bounceintel.com is our own, built and run in-house rather than resold from another provider. Registration details are on the contact page and in our legal documents.
Not in our database. A bulk list exists only in the report files generated for you, which sit outside the web root, are reachable only through an authenticated download and are deleted when their retention window expires. What we keep is aggregate: how many came back safe, risky, invalid or unknown.
Because some mail servers refuse to disclose whether a mailbox exists, and some are temporarily unreachable. Inventing a verdict in those cases is how a verification tool ends up suppressing good customers. We tell you what happened and let you decide.
A catch-all accepts every address, so no external test can distinguish a real mailbox from an invented one. We identify the configuration, score the risk from the other signals available (provider behaviour, domain reputation, historical outcomes) and report the confidence honestly.
No. Our Acceptable Use Policy prohibits purchased, scraped and brokered lists, and we enforce it. Verification confirms that addresses you already hold lawfully are still live; it is not a way to make an unlawful list usable.
Yes, and it applies automatically as part of our terms; you do not need to sign anything for it to bind us. If your procurement process needs a countersigned copy, or the Standard Contractual Clauses executed separately, ask and we will arrange it.
100 credits on signup, no card required, and the same pipeline an enterprise API call runs on.