A freedom to operate patent search asks whether you can sell, not whether you can patent.
Those are different questions with different scopes. Patentability looks at all prior art including expired patents. FTO looks only at in-force claims — expired patents are irrelevant to it entirely.
And it cannot stop early. A patentability search that finds one anticipating reference has answered the question. An FTO search that finds one blocking patent has found one of possibly several, and missing any of them is the failure mode.
That is most of the price difference, and it is why FTO runs $3,000 to $15,000 against $500 to $3,000.
FTO versus patentability
| Patentability search | FTO search | |
|---|---|---|
| Asks | Can I patent this? | Can I sell this? |
| Covers | All prior art | In-force claims only |
| Expired patents | Relevant | Irrelevant |
| Pending applications | Relevant as art | Relevant — may grant |
| Non-patent literature | Essential | Largely irrelevant |
| Can stop early | Yes | No |
| Reads | Whole disclosures | Claims |
| Typical cost | $500–$3,000 | $3,000–$15,000+ |
The reading changes too. A patentability search reads what a document teaches; an FTO search reads what a claim covers. Those are different parts of the same patent.
Non-patent literature barely matters for FTO. A journal article cannot be infringed. It matters enormously for patentability. See patentability search services.
What drives the cost
| Factor | Effect |
|---|---|
| Number of jurisdictions | Each is a separate body of live rights |
| Number of product features cleared | Each needs its own search |
| Technology crowding | More patents to review |
| Pending applications included | More documents, ongoing monitoring |
| Claim-by-claim analysis depth | Screening versus full mapping |
| Opinion included | Separate legal work |
Jurisdiction count is the largest multiplier. Clearing a product for the US, Europe, China and Japan is four searches, not one with a wider setting.
Feature scope is the lever you control. Clearing every feature of a complex product is prohibitively expensive, so scope is normally set around what is new and what competitors are known to have patented.
Scoping the search
| Question | Effect on scope |
|---|---|
| Where will it be made? | Manufacturing is an infringing act |
| Where will it be sold? | Each market separately |
| Where will it be imported? | Importing infringes |
| Which features are new? | Prioritise these |
| Which features do competitors patent? | Prioritise these |
| Which are standard commodity parts? | Usually deprioritise |
Manufacturing location is frequently forgotten. Making a product in a country where a patent is in force infringes there, even if you never sell it there.
Rights are national throughout. A US patent covers acts in the US only, so a product made and sold entirely in another country cannot infringe it. See freedom to operate.
Pending applications
| Status | Treatment |
|---|---|
| Granted, in force | The core of the search |
| Published application | Include — may grant |
| Application filed under 18 months ago | Invisible to everything |
| Non-publication request filed | Invisible until grant |
| Abandoned application | Ignore |
Published applications matter for two reasons. They may grant with claims covering your product, and pre-grant royalties can reach back to publication if the granted claims are substantially identical.
Claims usually narrow before grant, so a published application is a worst-case view rather than a forecast. Check the file history for amendments already made. See patent is pending.
The eighteen-month blackout is absolute. No search sees an application filed last year, which is why a clean result is not clearance.
Screening versus full analysis
| Level | What it involves | Cost |
|---|---|---|
| Screening | Identify candidates, check status, quick claim read | Lower |
| Full claim mapping | Element by element on each candidate | Higher |
| Opinion | Legal analysis and conclusion | Separate |
Most hits are resolved at screening. Expired patents, patents in markets you do not enter, and claims that plainly do not read on the product all close quickly.
Full mapping is reserved for the few that survive screening, which is what keeps the cost proportionate. See product mapping patent infringement.
Reading results
| Finding | Response |
|---|---|
| Expired patent | Ignore — public domain |
| In force, claims clearly avoid the product | Note and close |
| In force, claims read on the product | Escalate |
| In force, arguable | Map element by element; check file history |
| Published application, broad claims | Monitor; check amendments |
| Patent in a market you do not enter | Note the constraint |
Check status before analysing anything. Nearly three in five US utility patents lapse before term, so a substantial share of hits 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.
Remaining term changes the response. A blocking patent with two years left is a scheduling problem; one with twelve is an engineering or licensing problem.
What the search reads
FTO reads claims; patentability reads disclosures. The difference changes the work.
| FTO | Patentability | |
|---|---|---|
| Part of the patent read | The claims | The whole disclosure |
| Question asked | Does this cover our product? | Does this teach our invention? |
| A patent teaching the invention but claiming something else | Not a problem | A problem |
| A patent claiming broadly but teaching little | A problem | Less relevant |
A patent can be devastating for patentability and irrelevant for FTO, and the reverse. The same document behaves differently depending on which question is being asked.
Which is why the two searches are not interchangeable even where they cover the same technology area.
Worked example: scoping and results
A hardware product launching in the US and Germany, manufactured in Vietnam.
| Scope decision | Result |
|---|---|
| Jurisdictions | US, Germany (sales) + Vietnam (manufacture) |
| Features cleared | 4 novel elements, 2 competitor-patented areas |
| Commodity parts | Excluded |
| Pending applications | Included |
| Cost | $9,400 |
| Turnaround | 4 weeks |
What came back
| Finding | Jurisdiction | Status | Action |
|---|---|---|---|
| 11 patents screened out | All | Expired or clearly avoided | Closed |
| 1 patent, broad claims | US | In force, 9 years left | Escalate |
| 1 patent, same family | Germany | In force | Same analysis |
| 1 published application | US | Claims broad as published | Monitor |
| Nothing found | Vietnam | No filing | Manufacturing clear |
The blocking patent
| Step | Finding |
|---|---|
| Element-by-element mapping | 4 of 5 elements present |
| Fifth element | A specific mounting arrangement — present |
| Prosecution history | Claim narrowed to that arrangement to overcome prior art |
| Estoppel | Applies to equivalents around it |
| Design-around | Change the mounting arrangement — $40,000 engineering |
| Licence quote sought | Not pursued |
The design-around was safe because of the estoppel. The applicant surrendered non-conforming mounting arrangements to get the patent, so changing it cannot be captured by equivalents.
$40,000 against a nine-year exposure on a product line was a straightforward decision.
The Vietnam result mattered as much. No filing there meant manufacturing was clear regardless of the US and German positions.
The opinion is separate
| The search | The opinion | |
|---|---|---|
| Produces | Documents | Legal analysis |
| Performed by | A search firm | Counsel |
| Says | What exists | What it means |
| Priced | Separately | Separately |
| Protects against willfulness | No | Yes |
A written opinion obtained before launch is the principal defence against enhanced damages, which can reach three times the award.
It must predate the conduct to do its job. One obtained after suit addresses strategy rather than the state of mind that mattered.
Commission through counsel where possible, so the search results attract work product protection if they turn out unfavourable.
The search has a shelf life
| Time since the search | Reliability |
|---|---|
| Same week | Current |
| 6 months | Applications have published since |
| 12 months | Meaningful gap |
| 2 years+ | Says nothing about the present |
New applications publish continuously and patents grant weekly. An FTO opinion is accurate as of its search date and no later, which is why opinions state that date explicitly.
Monitoring is the practical answer. Watching your CPC classes catches new publications while design changes are still cheap. See patent monitoring.
If something blocks you
| Option | When it fits |
|---|---|
| Design around | The claim recites a changeable element |
| Licence | Designing around costs more than the royalty |
| Challenge validity | You hold strong documentary prior art |
| Wait | Short remaining term |
| Narrow the market | Sell only where it is not in force |
| Proceed with an opinion | Risk assessed and accepted |
Design-around is the cheapest resolution and the most overlooked. Check the prosecution history first — narrowing during examination makes a design-around safer by estopping equivalents.
Waiting is a real option. Litigation takes two to four years, so a patent with three years left may expire before any dispute concludes.
Documenting the search
| Record | Why |
|---|---|
| The search date | The opinion is accurate as of that day |
| Databases and jurisdictions covered | Defines scope |
| Search strategies and terms | Reproducibility |
| Documents reviewed and excluded | Shows the work |
| Stated limitations | Makes the opinion defensible |
An opinion that states its limitations is stronger than one that does not. Every FTO search has gaps — the eighteen-month blackout above all — and naming them shows the analysis was done properly.
Keep the record. If willfulness is ever raised, the search and opinion together are the defence, and they need to be locatable years later.
Freedom to operate patent search: the checklist
- Scope by jurisdiction first, including where the product is manufactured, not just sold.
- Prioritise novel features and competitor-patented areas. Clearing everything is prohibitive.
- Exclude expired patents. They are irrelevant to FTO, unlike patentability.
- Include published applications, and check their file histories for amendments.
- Verify in-force status on every hit before analysing it.
- Map blocking claims element by element, not by overall similarity.
- Read the prosecution history before relying on a design-around.
- Commission the search through counsel so results attract work product protection.
- Get a written opinion before launch. It is the principal willfulness defence.
- Treat the result as dated. Set up monitoring, because a clean search does not stay clean.