FTO meaning, in patents: freedom to operate. Whether you can make and sell your product without infringing somebody else's patent.
It is not the same as owning a patent, and the gap between those two ideas causes more expensive surprises than anything else in this field.
A patent is a right to exclude others. It is not permission for you to act.
(Outside intellectual property, FTO is also used for field training officer, full-time occupancy and other unrelated things. Everything below concerns the patent sense.)
The two questions people confuse
| Patentability | Freedom to operate | |
|---|---|---|
| Asks | Can I get a patent? | Can I sell my product? |
| Looks at | All prior art | In-force claims only |
| Expired patents | Relevant | Irrelevant |
| Focus | Your invention as claimed | Your actual product |
| Geography | Where you file | Where you make and sell |
| Answer | Novel or not | Clear, blocked, or uncertain |
Expired patents matter enormously to one and not at all to the other. A 1990 patent can defeat your application and cannot block your product.
Why a patent is not clearance
| Situation | Result |
|---|---|
| You patent an improvement | Valid |
| Someone holds a patent on the base invention | Also valid |
| Your improvement requires the base | You infringe theirs |
| They cannot practise your improvement | They infringe yours |
| Outcome | Cross-licence, or neither ships |
Blocking positions are ordinary. Improvement patents that cannot be practised without a licence to the underlying invention exist in every field.
Both patents are valid throughout. Nothing has gone wrong; the two rights simply overlap. See patent rights.
What an FTO search looks for
| Include | Exclude |
|---|---|
| In-force patents | Expired patents |
| Pending applications | Abandoned applications |
| Claims that could read on your product | Claims that plainly cannot |
| Countries where you make, sell or import | Countries you will not enter |
| Recent continuations | — |
Pending applications matter more than people expect. A competitor with a continuation pending can write claims aimed at your product once they see it.
Status verification comes first. Nearly three in five US utility patents are abandoned before term, so a meaningful share of scary-looking results are already dead.
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
Who FTO matters to
| Party | Why |
|---|---|
| Product companies | Launching into a patented field |
| Startups raising money | Investors ask in diligence |
| Manufacturers | Making for others carries risk too |
| Importers | Importing infringes independently |
| Licensees | A licence covers one patent, not the field |
| Acquirers | Buying a product line buys its exposure |
"We have a patent" is not an answer to an FTO question, and investors know it. The two are different analyses with different outputs.
Contract manufacturers infringe too. Making an infringing product for someone else is direct infringement, regardless of who designed it.
The process
| Step | What happens |
|---|---|
| 1 | Define the product precisely — features, components, methods |
| 2 | Identify the markets and manufacturing locations |
| 3 | Search by function and classification |
| 4 | Screen results on jurisdiction and status |
| 5 | Verify in-force status in Patent Center |
| 6 | Map surviving candidates element by element |
| 7 | Check prosecution history for estoppel |
| 8 | Conclude, with limitations stated |
| 9 | Obtain a written opinion where risk is real |
Step one determines everything downstream. A vague product definition produces a search that misses what matters and returns what does not.
Step five is free and eliminates the most. Three minutes per patent in USPTO Patent Center. See patent status.
Searching by function
Search what your product does, not what you call it.
| Approach | Why |
|---|---|
| Several vocabularies | Your term is not the industry's term |
| CPC classification browsing | Groups by function, catches what keywords miss |
| Competitor assignee searching | Who files in your space |
| Citation networks | Follow references from close hits |
| Foreign families | Rights are national; check each market |
What you call a flow stabiliser, a patent may call a damping baffle. One vocabulary finds one slice.
Element-by-element, never overall similarity
| Test | Correct? |
|---|---|
| Every claim element present in your product | Yes — that is infringement |
| Products look similar overall | No — not the test |
| Your product has extra features | Does not help — "comprising" is open |
| One claim element absent | No literal infringement |
Extra features never avoid a claim. Almost every claim uses "comprising", so a product containing everything claimed plus more still infringes.
One absent element ends literal infringement, and then the doctrine of equivalents and prosecution history estoppel decide the rest. See patent infringement analysis methodology.
Grading the result
| Grade | Meaning |
|---|---|
| Clear | No in-force claim reads on the product |
| Design change advised | A cheap modification removes the risk |
| Licence advisable | Blocking claim, no practical design-around |
| Uncertain | Turns on claim construction |
| Unverifiable | Public evidence insufficient |
| High risk | Claim reads squarely, patent robust |
Grade honestly, including "unverifiable". An analysis presenting a guess as a conclusion is worse than one stating what could not be determined.
Most results are not binary. The useful output is a ranked risk list with actions against each, not a yes or no.
The opinion and what it protects
| Form | Written |
| Timing | Before the accused conduct, not after suit |
| By | Qualified counsel |
| Content | Construes the claims, maps the product, reasons |
| Protects against | Enhanced damages for willfulness — up to 3× |
| Guarantees | Nothing about being sued |
Timing is the requirement most often failed. An opinion obtained when a complaint arrives says nothing about the four preceding years.
A conclusory opinion is worth little. One stating a conclusion without construing claims or mapping the product does not evidence the investigation it exists to prove. See willful infringement.
Manufacturing and import
| Activity | Where the patent must exist |
|---|---|
| Making | The country of manufacture |
| Importing | The destination country |
| Selling | The country of sale |
| Using | The country of use |
A US patent reaches goods made abroad at the border. Manufacturing in a country where the patentee holds nothing does not avoid US infringement on import.
Which means FTO must cover the destination market, not just the factory location.
FTO can never be certain
| Source of uncertainty | Why |
|---|---|
| 18-month publication delay | Applications are invisible when filed |
| Pending continuations | New claims can be written toward you |
| Claim construction | Predictive, not certain |
| Doctrine of equivalents | Extends beyond literal wording |
| Search completeness | No search finds everything |
| Foreign families | Different claims in each country |
Anything filed in the last eighteen months is invisible. That gap cannot be closed by any amount of searching.
FTO analysis manages risk. Anyone promising certainty is selling something.
When to do it
| Stage | Value |
|---|---|
| Concept | Highest — changing the design costs nothing |
| Detailed design | High |
| Before tooling | Last cheap moment |
| Pre-launch | Useful; options narrowing |
| After a notice letter | Late — and now urgent |
| During litigation | Damage control |
The cost of a design change rises steeply. At concept it is a conversation; after tooling and inventory it is a write-off.
Investors ask. A startup raising money in a patent-dense field will face FTO questions in diligence, and "we have a patent" is not an answer.
Worked example: the same product, two countries
A device launching in the US and Germany.
| United States | Germany | |
|---|---|---|
| Candidates after screening | 18 | 6 |
| In force | 11 | 4 |
| Reads on the product | 2 | 0 |
| Estoppel bars one | Yes | — |
| Result | 1 residual risk | Clear |
The same product, opposite conclusions. The blocking family had no German counterpart, which is ordinary — applicants file where they expect markets.
This is why FTO is per country and why a single global answer does not exist.
If something blocking is found
| Option | Detail |
|---|---|
| 1. Verify it is in force | Free — resolves many cases |
| 2. Check the actual claims | Broad summaries mislead |
| 3. Check remaining term | Continuation chains expire early |
| 4. Design around | Cheapest if done early |
| 5. Check prosecution history | Estoppel may narrow it |
| 6. License | Certainty, at a price |
| 7. Challenge validity | IPR or litigation |
| 8. Wait for expiry | 23.2 points drop at the third fee |
| 9. Accept the risk, with an opinion | Documented decision |
Step three catches people. A patent filed in 2023 as a continuation of a 2015 application expires in 2035, not 2043 — which may be sooner than your product cycle.
Waiting is a real strategy where the term is short. Paying nothing and launching later can beat licensing.
Worked example
A company plans to launch a filtration module in the US and EU.
| Step | Finding |
|---|---|
| 1. Product defined, 6 features | — |
| 2. Search, CPC + keyword, 2 jurisdictions | 240 results |
| 3. Screened on jurisdiction | 150 |
| 4. Screened on obviously absent elements | 18 |
| 5. Verified in-force status | 11 live, 7 lapsed |
| 6. Mapped element by element | 2 concerning |
| 7. Prosecution history checked | 1 barred by estoppel |
| 8. Remaining risk | 1 patent, 4 years left |
| Option chosen | Reason |
|---|---|
| Design change to one feature | Cheap at design stage |
| Written opinion on the residual | Willfulness protection |
| EU position | Different — no counterpart filed |
Step five removed seven patents for nothing. Free status checks did more than any paid analysis.
Step seven removed another one for nothing. Estoppel is established from a public file wrapper.
The EU result differed entirely, which is why FTO is country by country.
Documenting the analysis
| Record | Why |
|---|---|
| Date of every source | Products and patents change |
| Search queries and classes used | Shows scope |
| Patents screened out, and why | Defensible method |
| Construction adopted on disputed terms | Makes conclusions checkable |
| Evidence gaps | Honest limitations |
| Who performed it, and when | Privilege considerations |
Analyses done in anticipation of litigation may be privileged; routine ones may not. Involving counsel early affects that.
Cost
| Level | Scope |
|---|---|
| Self-screening | Free — Google Patents, Patent Public Search, Espacenet |
| Professional search | Search firm, one jurisdiction |
| Full analysis | Counsel, mapping, construction |
| Written opinion | Counsel — the willfulness defence |
| Ongoing monitoring | Catches new grants and continuations |
Screen yourself first. Free tools do the elimination, and paid work then focuses on the handful that survive.
Monitoring matters because FTO is a snapshot. New patents grant continually, and an opinion from three years ago may no longer describe the position. See patent monitoring.
Monitoring keeps it current
| Watch | Why |
|---|---|
| New grants in your classes | The position changes |
| Published continuations | Claims may be aimed at you |
| Assignment transfers | Patents moving to assertion entities |
| Competitor filings | Direction of travel |
An FTO opinion is a snapshot. New patents grant continually, so an opinion from three years ago may no longer describe the position.
Common mistakes
| Mistake | Consequence |
|---|---|
| Assuming a patent gives clearance | The core error |
| Searching only your own vocabulary | Misses the blocking art |
| Not checking in-force status | Worrying about dead patents |
| Comparing overall similarity | Wrong in both directions |
| Ignoring pending applications | Blindsided later |
| Treating it as one-off | Position changes |
| One jurisdiction only | Rights are national |
| Doing it after tooling | No cheap options left |
What it costs to skip
| Skipped | Possible consequence |
|---|---|
| FTO before launch | Redesign after tooling, or a licence on their terms |
| Written opinion | Enhanced damages up to 3× |
| Per-country analysis | Blocked in a market you entered |
FTO meaning: the checklist
- FTO means freedom to operate — can you sell, not can you patent.
- A patent is not clearance. It excludes others; it permits nothing.
- Only in-force claims matter. Expired patents are irrelevant here.
- Search by function in several vocabularies, plus CPC classification.
- Verify status before worrying. 58.6% of utility patents are abandoned early.
- Compare element by element, never overall similarity.
- Check prosecution history. Estoppel is free to establish and often decisive.
- Do it per country. Rights are national and results differ.
- Get a written opinion before launch, not after a letter arrives.
- Re-run it periodically. FTO is a snapshot, and new patents grant constantly.