A proposed convention lets a domain owner publish a “for sale” signal as a DNS record, so buyers can query it directly instead of relying on parking pages, WHOIS contacts or brokers. It is a community idea, not an adopted standard.
Key takeaways
- The idea is to place a machine-readable “this domain is for sale” signal in the Domain Name System, alongside the records that already handle mail, verification and text metadata.
- DNS is attractive for this because it is queryable, cacheable, delegated to the domain owner and readable by any script without scraping a web page.
- The signal typically carries a contact method or a pointer to a sales page, and may carry an asking price, but the details vary between proposals and none is universally implemented.
- Nothing in DNS proves that the party publishing a record is legally entitled to sell the domain, so the signal is a hint rather than a guarantee.
- Adoption, not technical difficulty, is the hard part: a record that only a handful of registrants publish and only a handful of tools read has little practical value.
What is actually being proposed
The proposal is a convention for publishing domain sale intent as a DNS record. Instead of a would-be buyer loading a website, hunting for a contact form, or filing a WHOIS enquiry that may reach a privacy proxy, they would issue a DNS query against the domain and read a structured answer.
In practice this means using an existing DNS record type — most commonly TXT, the general-purpose text record — at a defined name under the domain, containing a defined set of fields. That pattern is familiar: SPF, DKIM, DMARC and countless site-verification tokens all work by putting an agreed string in a TXT record at an agreed name. A “for sale” record would follow the same shape, with fields for something like a contact address, a link to a listing, and possibly a price or currency.
The point is machine readability. A parking page is designed for a human with a browser. A DNS record can be read by a registrar’s interface, a browser extension, a marketplace crawler, a command-line tool or a script that checks a list of thousands of candidate names in seconds.
Why it is being discussed now
Ideas of this kind surface on technical forums periodically, and they gain attention when someone writes up a concrete specification rather than a vague suggestion. The current discussion follows that pattern: a published convention, a reference to how it would be formatted, and a comment thread arguing over whether it solves a real problem.
The wider context is that buying a domain from its current holder remains an awkward process. Registration data that was once broadly public is now commonly redacted for privacy reasons, which removed a channel buyers used. Parking pages, brokerage platforms and unsolicited enquiry emails filled the gap, and none of them is a neutral, open mechanism. A DNS convention is a response to that gap. Whether the response is proportionate is precisely what people are arguing about.
The background a newcomer needs
DNS is the system that turns names into addresses and other metadata. When you register a domain, you control the records published for it — that control is the whole point of delegation. Anyone on the internet can query those records, and answers are cached by resolvers for a period the owner sets.
Several important internet conventions are built purely on TXT records rather than on new record types. Email authentication is the best-known example: a mail server checks a TXT record to decide whether a sending host is authorised. Domain-control verification for cloud services works the same way. These conventions succeeded because a small number of large operators agreed to read the records, which gave every registrant a reason to publish them.
The DNS also has record types intended for richer metadata, and there is a general reluctance in the standards community to overload TXT indefinitely. That tension — reuse a general-purpose record and ship today, or define something specific and wait — runs through most discussions of new DNS conventions.
Who is affected and how
Domain holders who want to sell are the intended beneficiaries. Publishing a record costs one DNS change and no listing fee, and it works even if the domain is not serving a website at all. For a holder with a large portfolio, it can be scripted.
Buyers and their tooling gain a fast, low-friction way to test intent across many names. Anyone building a discovery tool, a registrar search interface or a portfolio monitor could read the record without scraping.
Marketplaces and brokers sit more awkwardly. An open convention reduces dependence on any single platform, though platforms could also adopt it as an additional channel and use it to feed their own listings.
Ordinary registrants who have no intention of selling are affected in one specific way: if the convention becomes widespread, absence of a record becomes a weak signal in itself, and it does nothing to reduce the volume of speculative enquiries that domain owners already receive.
Where informed people disagree
The strongest objection is that DNS is a naming and resolution system, not a classified-ads channel, and that loading commercial metadata into it is a misuse of infrastructure. Supporters answer that TXT records are already used for arbitrary metadata and that this is a difference of degree, not of kind.
A second disagreement is about trust. A DNS record proves control of the zone at the moment of the query, and nothing more. It does not prove the publisher is the registrant, that they have authority to transfer, or that any stated price is honoured. Control of a zone can also be delegated to a hosting provider, so the party who can publish the record is not always the party who owns the name. Advocates accept this and argue the record is only ever a starting point for a conversation that still ends in an escrowed transfer.
A third is whether it invites abuse. A machine-readable market signal is also a machine-readable target list. Critics expect it to be harvested for spam, or published falsely on domains the publisher does not intend to sell.
Finally there is the adoption argument, which is the most practical. Without registrar interfaces that expose the record and clients that read it, it remains a convention observed by a handful of enthusiasts.
What this means in practice
If you want to try it, the mechanics are ordinary DNS work. You add a TXT record at your provider’s control panel or through its API, at whatever name the convention you are following specifies, with the field values it defines. You choose a time-to-live; a shorter TTL means changes propagate faster but generates more queries. You then verify it with a standard lookup tool such as dig or nslookup, querying for TXT records at that name, and confirm the string appears as you intended.
Two cautions are worth stating plainly. First, whatever contact address you publish becomes public and will be scraped, so use an address you can filter or discard rather than a personal one. Second, publishing a price commits you to nothing legally but does anchor every subsequent negotiation, and it cannot be un-seen once caches and crawlers have picked it up.
If you are on the buying side, treat the record as an invitation to make contact and nothing more. Verify who actually holds the domain through the registrar, and complete any transfer through an escrow arrangement, exactly as you would with a listing found anywhere else.
What to watch next
The signals that would matter are adoption ones. Watch for registrars or DNS hosting providers adding a “for sale” toggle to their control panels, since that is what turns a convention into something non-technical holders use. Watch for search and discovery tools reading the record and surfacing it in results. And watch whether the convention is written up and submitted through a standards process, which would suggest an intent to make it interoperable rather than a single project’s format.
The alternative outcome is equally plausible: the idea is discussed, a few thousand domains publish a record, tooling never materialises, and it joins the long list of proposed DNS conventions that were technically fine and socially inert. At present there is no public evidence of broad implementation, and any claim about how many domains carry such a record should be treated sceptically unless it comes with a described measurement method.
Frequently asked questions
What DNS record type would a domain-for-sale signal use?
Most proposals of this kind use a TXT record, the general-purpose text record, placed at a defined name under the domain. TXT is chosen because every DNS provider supports it, it accepts arbitrary strings, and existing conventions such as email authentication and site verification already work this way. A dedicated record type would be cleaner in principle but would need broader software support before it was practical to publish.
Does publishing such a record mean the domain is definitely available to buy?
No. A DNS record demonstrates control over the domain’s zone at the time of the query, not legal ownership or authority to transfer. The publisher might be a hosting provider, a former administrator with lingering access, or someone testing the convention. Any stated price is an asking figure with no binding force. Treat the record as a prompt to make contact and verify ownership through the registrar before money changes hands.
How would I check whether a domain publishes one?
You would query the domain for TXT records at the name the convention specifies, using a standard lookup tool such as dig or nslookup, or any web-based DNS lookup service. The answer, if present, is a text string containing the convention’s fields. Because the exact name and field format depend on which proposal you are following, check that specification first — a lookup at the wrong name simply returns nothing.
Is this an official internet standard?
Not as far as can be verified. Conventions of this type generally begin as a written-up proposal from an individual or a small project and only become standards if they are submitted through a formal process and gain implementation. Nothing about a DNS record requires official blessing to work — you can publish any TXT record you like — but interoperability, meaning other people’s software reading it the same way, does depend on shared agreement.
Will publishing a for-sale record attract spam?
Very likely, in the same way that any publicly readable contact address does. DNS records are open to anyone who queries them, and bulk querying of large domain lists is routine and cheap. Publishing a contact address in a machine-readable field makes harvesting trivial. The practical mitigation is to publish an address dedicated to this purpose, or a link to a form, rather than an address you rely on for anything else.
Why not just use WHOIS or a parking page instead?
Both remain available, and both have gaps. Registration data is now commonly redacted for privacy reasons, so WHOIS often reaches a proxy rather than the holder. A parking page requires the domain to be served over HTTP and is built for human readers, not scripts. A DNS record works without any web server and is machine-readable by default. The argument for the convention rests on filling those specific gaps.
Sources and further reading
- Public technical discussion threads on Hacker News, where the proposal was posted and debated by practitioners.
- IETF Request for Comments documents describing the DNS, TXT record usage and email authentication conventions built on them, as background on how such conventions are formalised.
- ICANN policy material on registration data access and the shift towards redaction, which explains why WHOIS-based contact has become less reliable.
- Documentation from DNS hosting providers and registrars on creating and verifying TXT records, useful for the practical steps.
Surfaced from the hackernews signal “DNS-based domain sale signalling”. AI-assisted draft, editorially reviewed.
