November 2026Every invention disclosure meeting ends at the same fork. File a patent and pay for a right to exclude. Or publish the idea so nobody else can ever own it. That is the whole of defensive publication vs patent, and most pages on the subject stop there, with the word "cheaper" and no numbers.
This guide adds what they leave out: what each venue actually costs next to real filing fees, why publishing in Europe can destroy your own patent on the day you hit upload, the third route most teams forget, and the question we ask from the deal side of the table. A patent is an asset someone may one day pay for. A defensive publication is a cost you never get back. Sometimes that is exactly the right trade.
A patent gives you a time-limited right to stop others making, using or selling your invention. It costs real money to get and keep.
A defensive publication gives you no right at all. You disclose the invention so it becomes part of the state of the art and can no longer be patented. As the Swiss patent office puts it, you can continue to use your invention, but so, too, can your competitors.
So the choice is not "strong protection vs weak protection". It is two different jobs. A patent keeps others out. A publication keeps others from locking you out. If you first want the basics, start with what a defensive publication is.
The patent is the only one of the two with a counterparty value. You can sue on it, license it by field or territory, or sell it outright. It is also the only one that shows up on a balance sheet or in a buyer's diligence folder.
A publication is a shield, not a sword. Researchers describe its logic as using disclosure to block the competitors from blocking oneself. That is valuable, but it is narrow. Chambers is careful to call it a risk-reduction mechanism, not a guarantee of freedom to operate against all present or future third-party rights. A patent someone else filed last year is untouched by what you publish today.
| Patent | Defensive publication | |
|---|---|---|
| Right to exclude others | Yes, after grant | None |
| Stops others patenting it | Yes | Yes, if enabling and dated |
| Can you sell or license it | Yes | No, there is nothing to transfer |
| Upfront cost | Thousands | $0 to a few hundred dollars |
| Ongoing fees | Renewals for the life of the patent | None |
| Reversible | You can abandon later | No, disclosure is permanent |
Here is the comparison nobody prints.
Publishing. Technical Disclosure Commons is free to publish and free to search, posts submissions within 72 hours if no revisions are needed, and is operated by bepress at the request of Google. The paid standard is IP.com: uploading a defensive publication costs $395 for up to 25 pages, and the USPTO subscribes for most of its 9,200 examiners and supervisors. The same report notes that Google pays employees a $1,000 bonus per TDCommons post, a tenth of what it awards for patent filings. That ratio tells you how one of the world's largest filers prices the two options internally.
Filing. At the USPTO, the basic filing, search and examination fees for a utility application come to $2,000 for a large entity, $800 for a small entity and $400 for a micro entity. In Germany the DPMA charges EUR 40 for electronic filing, EUR 300 for a search and EUR 150 or 350 for examination, depending on whether you already requested the search. Those are office fees only. Drafting is the real line item: our own worked example for a 10-claim German filing is EUR 4,000 plus official fees. Protect an invention in a few countries for 8 to 10 years and the bill can easily sum up to about 50,000 in pounds, euros or dollars.
| Route | Typical outlay to get started |
|---|---|
| Technical Disclosure Commons | $0 |
| IP.com Prior Art Database | $395 for up to 25 pages |
| US utility application, office fees only | $400 to $2,000 by entity size |
| German national filing with drafting | About EUR 4,000 plus DPMA fees |
| A few countries kept for 8 to 10 years | About 50,000 |
Then the patent keeps billing. Annual or staged patent renewal fees run for as long as you want the right to exist. The publication costs nothing after day one. For the full lifecycle on the filing side, see our breakdown of what it costs to file a patent.
Most guides present defensive publication as aimed at competitors. It hits you first.
US law treats anything described in a printed publication before your effective filing date as prior art. The exception in section 102(b)(1) is that your own disclosure made one year or less before you file does not count against you. So in the US, publishing buys you twelve months to change your mind.
Older advice says a US competitor can get around your publication by proving an earlier date of conception. That reflects the old first-to-invent rules. Today's section 102 measures prior art against the effective filing date, not the date someone had the idea.
The European Patent Convention allows only a six month window, and only for evident abuse or display at an officially recognised international exhibition. Your own blog post, paper or defensive publication does not qualify. The Nature Biotechnology analysis is blunt about the consequence: a decision to publish is a decision to irreversibly surrender potential patent rights, and outside the US that happens immediately.
For a team selling into the US and Germany at once, this is the deciding fact. Publish first and you have kept, at most, a US option that expires in a year.
A publication only blocks what it teaches. The same Nature piece explains that it must disclose every element of the later claim and enable a skilled person to make it. Anything you hint at but do not enable is still open for someone else to patent.
The European Patent Office's expectations, as summarised by PatentRenewal, are three: describe all relevant details, publish where search engines can find it under the relevant keywords, and state clearly when and where it was published. The European IPR Helpdesk adds that the drafting should reflect a patent claim and include a field of practical application, because a partial description lets competitors patent the features you left out.
Venue matters as much as text. A website or product catalogue is legally prior art, but these often sit outside the databases examiners search. A prior art no examiner finds still costs you a court case to use. That is the practical case for IP.com and TDCommons over your own blog. One small but useful feature: IP.com lets you publish with no authorship information, so a disclosure does not advertise which company is working on what.
There is a middle path. File a patent application, then abandon it. The IPR Helpdesk notes that the application is published 18 months after filing and stays in patent databases forever, which is exactly where examiners look first. You pay office fees instead of venue fees, and you get a filing date and a priority year in which to decide whether the invention deserves a full patent after all.
The cost is time. Until it publishes, the application is not prior art against anyone. Early publication can be requested, but IP.com, which has an obvious interest in the comparison, claims patent office early publication can take 14 weeks or more against minutes on its own platform. If a competitor is racing you to the office, a same-day publication closes the door faster.
The SERP answers this with "patent if the ROI is high". Fine, but nobody can measure ROI on an invention that has not been sold. We use a sharper question: would anyone pay for the right to stop someone else using this?
Publish when most of these are true:
File when any of these is true:
That last point is the one IP committees underrate. Chambers describes the habit of sorting every disclosure into "patent or park". Publishing is the third bin, and unlike parking it cannot be undone. Before you publish, check what others already own with a freedom to operate vs patentability search. A professional novelty search, which in our drafting work runs through Patsnap with filing by an independent licensed patent attorney of record, tells you whether there is anything left to patent. EX-IX is not a firm of patent attorneys, and a novelty search does not guarantee grant.
From the transaction side we see the other half. A patent can reach a buyer: ours go through a digital-twin validation gate of claim reconstruction, detectability analysis and validity probability before any buyer sees them, and brokerage is paid by the seller only on completion. Not every patent sells, and many never should be filed. But if you think there is a buyer, look at how much a patent can sell for before you give the idea to the public for $395.
Not what the publication enabled, and that includes you. The exception is the US, where your own disclosure does not count against you if you file within one year. Improvements the publication did not describe may still be patentable.
Legally it can be prior art if it is public and dated. Practically, examiners rarely search company websites, so a competitor's application may sail past it and you would have to prove it later. Use a venue examiners actually search.
Yes. IP.com offers a no-authorship option, so the disclosure enters the record without naming the company or the inventors.
No. It stops later patents on what you disclosed. It does nothing about patents others filed before your publication date, which you still need to search for.
We use essential cookies to operate this website and, with your consent, optional cookies to understand site usage. You can accept or decline non-essential cookies at any time. See our Impressum for our contact and legal details.