5 Signs Your Invention Might Not Be Patentable

Not every invention can be patented. Before you spend $15,000+ on an application, watch for these five warning signs.

Sign 1: You Found Something Similar Online

You searched Google and found a product, video, or article describing something very close to your invention.

Why it matters: Prior art doesn’t have to be a patent. Products on Amazon, YouTube demos, academic papers, even old forum posts — all count as prior art that can block your patent.

What to do:

• Search thoroughly before concluding nothing exists
• Document exactly what you found
• Identify specific differences between your idea and the prior art
• Consider if those differences are substantial enough

Red flag level: 🚩🚩🚩 High — This needs serious evaluation

---

Sign 2: Industry Experts Say “That’s Obvious”

You pitch your idea and experts respond: “Oh yeah, that’s the obvious next step” or “Anyone in the industry would think of that.”

Why it matters: To be patentable, an invention must be “non-obvious” to a person skilled in the field. If experts consider it a natural evolution, the patent office might too.

The non-obviousness test:

• Would an engineer in your field easily think of this?
• Is it just combining two known things in an expected way?
• Does it produce surprising or unexpected results?

What to do:

• Document any unexpected results
• Identify aspects that go against conventional wisdom
• Look for evidence that others tried and failed

Red flag level: 🚩🚩 Medium-High — Might be overcome with the right framing

---

Sign 3: It’s Just “X But On a Computer/App”

Your invention takes an existing process and digitizes it, or does something “but with an app” or “but with AI.”

Why it matters: Abstract ideas implemented on generic computers often fail patent eligibility tests (particularly after the Alice decision). The addition of a computer or software must provide something more than just automation.

Examples that struggle:

• “Matching buyers and sellers, but with an app”
• “Calculating risk scores, but with machine learning”
• “Managing inventory, but in the cloud”

What might work:

• Technical improvements to how computers function
• Specific algorithms that solve technical problems
• Physical-digital integration with novel mechanisms

What to do:

• Focus on the technical implementation, not the business concept
• Identify specific technical problems your solution solves
• Document the technical architecture in detail

Red flag level: 🚩🚩🚩 High for pure software/business methods

---

Sign 4: You Already Told Everyone About It

You posted on social media, pitched at a demo day, published a paper, or showed it at a trade show — before filing anything.

Why it matters: Your own public disclosure can be prior art against yourself.

The rules:

• US: You have a 1-year grace period from first public disclosure
• International: Many countries have NO grace period — public disclosure kills foreign rights immediately

What counts as public disclosure:

• Social media posts (even “private” groups can be risky)
• Pitch events and demo days
• YouTube videos
• Academic publications
• Trade show demonstrations
• Conversations without NDA

What to do:

• File a provisional ASAP if you’ve disclosed within the last year
• Document exactly when and where you disclosed
• Consult an attorney if international rights matter

Red flag level: 🚩🚩🚩 Critical — Time-sensitive issue

---

Sign 5: It’s An Idea, Not An Implementation

You have a concept — “a device that does X” — but no specific design, mechanism, or method for how it works.

Why it matters: Patents protect implementations, not ideas. You must be able to describe your invention specifically enough that someone could build it.

Not patentable:

• “A teleporter” (concept only)
• “AI that perfectly predicts the stock market” (no specific method)
• “A cure for cancer” (outcome without implementation)

Patentable:

• “A quantum entanglement device using method Y with components Z” (specific)
• “A machine learning model using architecture X trained on data Y to predict Z” (specific)
• “A compound with formula X that inhibits protein Y” (specific)

What to do:

• Develop your concept into a specific implementation
• Document the mechanism, method, or structure
• Create drawings or diagrams
• Build a prototype (not required, but helps)

Red flag level: 🚩🚩 Medium — Usually fixable with more development

---

What to Do If You See These Signs

Don’t give up immediately. Many of these issues can be addressed:

1. Found similar prior art? Identify meaningful differences and design around.
2. Seems obvious? Document unexpected results and non-conventional aspects.
3. Software/business method? Focus on technical implementation details.
4. Already disclosed? File provisional immediately; assess international damage.
5. Just an idea? Develop specific implementation before filing.

Do get a professional assessment. A patent search and patentability opinion costs $150-$500 and can clarify whether these red flags are fatal or manageable.

---

The Harder Truth

Some inventions genuinely aren’t patentable. That doesn’t mean they’re not valuable:

• Trade secrets can protect processes you keep confidential
• First-mover advantage can matter more than patents
• Copyright protects code, designs, and content
• Trademarks protect branding
• Speed to market can beat legal protection

A patent is one tool. If it’s not available, use others.

---

Before You Spend Money

If you’ve seen multiple red flags:

1. Do a thorough prior art search ($150-$300)
2. Get an honest assessment of patentability
3. Quantify the differences between your invention and prior art
4. Decide if those differences justify $15,000+ in filing costs

Better to know now than after you’ve invested.

---

Check if your idea is patentable → Get an AI assessment