Freedom-to-Operate and Patent Clearance Toolkit

By Casey Scott McKay · 2024-01-17

A freedom-to-operate analysis answers one question: can we sell this without being sued successfully. It is the most requested and most misunderstood patent exercise, because clients want certainty and the honest answer is always a risk assessment. This toolkit runs a clearance from product scoping through search, charting, and the design-around decision, and routes each stage to the Marksy documents that do the work. It explains why the scope of the product defines the cost of the search, why pending applications and live continuations matter more than issued claims for a product with a long life, and what a written opinion actually buys given that failure to obtain advice of counsel proves nothing. It covers standard-essential patents, supply chain indemnities, and the diligence context. It closes with the reading path, the authorities table, and the forms.

IP and Technology > Patent Counseling Transactions | Toolkit | Published 17 January 2024 - Updated 13 February 2026 | Casey Scott McKay - marksy.us

Summary. A freedom-to-operate analysis answers one question: can we sell this without being sued successfully. It is the most requested and most misunderstood patent exercise, because clients want certainty and the honest answer is always a risk assessment. This toolkit runs a clearance from product scoping through search, charting, and the design-around decision, and routes each stage to the Marksy documents that do the work. It explains why the scope of the product defines the cost of the search, why pending applications and live continuations matter more than issued claims for a product with a long life, and what a written opinion actually buys given that failure to obtain advice of counsel proves nothing. It covers standard-essential patents, supply chain indemnities, and the diligence context. It closes with the reading path, the authorities table, and the forms.

Keywords: freedom to operate · patent clearance · product scoping · search strategy · claim charting · blocking patent · design around · written opinion · willfulness · advice of counsel · privilege · expired patents · pending applications · continuation risk · standard essential patents · licensing · indemnity · supply chain · diligence · budget

Start Here

Kettleworth Systems is eleven weeks from launching a networked industrial controller. On a Wednesday in May, the general counsel receives four things.

Engineering asks for "a patent clearance" on the product, which contains roughly two hundred distinguishable features, four third-party software components, and a communications module bought from a supplier.

The chief executive asks whether the company can represent to its largest customer that the product infringes no third-party patent.

A board member, who once sat on a company that lost a patent case, asks whether the company has an opinion of counsel.

And the head of sales mentions that the product implements a wireless standard, which nobody has framed as a legal fact.

Four requests, one exercise, and a common misunderstanding: that clearance produces a yes or a no. It produces a ranked risk assessment, a set of decisions, and a documented record — and the value is in the decisions.

This toolkit answers three questions.

What are we actually clearing? The scope of the search follows the scope of the product, and an unbounded scope produces an unbounded bill and an unusable result.
What did we find, and how bad is it? Claim charting against a small number of candidates, with a considered judgment on each.
What do we do about it? Design around, license, challenge, accept, or delay. Each is a real answer and the choice is commercial.

If you read only one thing, read Freedom to Operate. It explains what a clearance search can and cannot establish, which is the conversation that has to happen before the work is commissioned.

What Clearance Is, and Is Not

It is not a patentability search. Patentability asks whether an invention is new over the prior art. Clearance asks whether a product infringes claims in force. Different databases, different date logic, different output, and they are commissioned from the same vendors under confusingly similar names.

It is not a validity search. A blocking patent's validity is a separate question, answered only after the patent is identified.

It cannot be complete. Applications publish at eighteen months under 35 U.S.C. § 122, which means anything filed in the last year and a half is invisible. Continuations pending in published families can produce new claims drafted against a product shipping today. And searching is a human process against imperfect classification.

It does not confer a right. A patent is a right to exclude, not a right to practice. Clearing a product says nothing about whether the company's own patents are valid or valuable.

What it does produce. A ranked list of patents whose claims plausibly read on the product, a charted analysis of the ones that matter, and a set of decisions with a record showing they were made deliberately.

Stage One: Scoping

The single largest determinant of cost and usefulness.

Define the product precisely. Model, configuration, and the features actually shipping. A clearance on a roadmap is a clearance on a moving target.

Identify the inventive and unusual features. A search across every feature of a complex product is unbounded. A search across the features that are novel, distinctive, or known to be crowded is finite and answers the question that matters.

Identify what is bought rather than built. Components from suppliers may carry indemnities and may already be licensed, and clearing them at the company's expense duplicates work someone else has done.

Identify what is open source. Its patent exposure is separate from its copyright obligations, and some licenses carry patent grants and termination provisions that matter. See Running an Open Source Compliance Program.

Identify what implements a standard. Standard-essential patents are a separate analysis with a separate answer, addressed below.

Define the territories. Patents are territorial. A search covering jurisdictions where the product will not be sold is money spent on nothing, and one omitting a manufacturing country ignores where an injunction would bite hardest.

Define the timeline. Clearance takes six to twelve weeks done properly, and a request eleven weeks before launch is a request for a partial answer.

Deliverable. A scoping memorandum naming the product, the features to search, the territories, the timeline, and the budget.

Stage Two: Searching

Commission it properly. A professional searcher with subject-matter competence, briefed with the scoping memorandum rather than with a product brochure.

Search issued claims, not disclosures. Clearance is about what is claimed and in force.

Search pending applications too. Published applications signal what is coming, and a published application with a live continuation is a claim set that has not been written yet.

Check the assignee landscape. Who holds patents in this space, whether any are known assertion entities, and whether the company already has relationships or licenses with them.

Search the client's own knowledge. Competitors' marked products, patents cited in customer requests for proposal, art the engineering team already knows about. This costs nothing and it is skipped.

Screen the results. A raw search returns hundreds. The screening pass eliminates the expired, the abandoned, the plainly irrelevant, and the ones whose independent claims obviously do not read on the product. This is lawyer work and it is where a searcher's list becomes a legal work product.

Expect a small final set. A well-scoped clearance on a complex product typically produces five to twenty patents worth charting.

Stage Three: Charting

Chart independent claims only, at first. A product avoiding the independent claim avoids everything depending from it.

Break each claim at every limitation and map each to the product or record its absence.

Be honest about the contested element. The value of a chart is in the limitation that is close, not in the six that plainly read.

Check the file wrapper for anything close. The amendment that produced allowance identifies the point of novelty, and the arguments constrain the claim's scope in ways the claim language does not show. This is the step that resolves most close calls.

Check the term and the fees. Expiration under 35 U.S.C. § 154, terminal disclaimers under 37 C.F.R. § 1.321, and maintenance fee lapse under 35 U.S.C. § 41. A patent that expires within the product's life is a different risk from one with fifteen years left.

Check the family. A live continuation means new claims can be drafted later against the product as it actually ships.

Rank the results. Clear, close, or blocking. Only the last two need decisions.

Stage Four: The Decisions

For each patent that is close or blocking, five options and one hybrid.

Design around. Change the product so the claim does not read. Cheapest when done before launch and increasingly expensive afterward. Price it in engineering hours and schedule, and document the change and its date, because it establishes when exposure ended.

License. Approach the holder, which reveals the company's interest and starts a negotiation. Sensible where the holder licenses routinely, dangerous where the holder is a competitor who did not know.

Challenge. Assess validity. Where materially better art exists than the examiner saw, a post-grant petition may be available, though filing before any assertion is a strategic choice with its own consequences. See PTAB Practice Toolkit.

Accept. Where the claim is weak, the patent expires soon, the holder is inactive, or the exposure is modest. This is a legitimate answer and it should be a documented decision rather than an omission.

Delay. Ship in territories where the patent does not exist, or wait for expiry where it is close.

The hybrid that is usually right. Design around the ones where the change is cheap, accept the ones where the claim is weak, and reserve licensing for the small number that are both strong and unavoidable.

The Opinion Question

What the statute says. 35 U.S.C. § 298 provides that failure to obtain the advice of counsel, or failure to present such advice, may not be used to prove willful infringement or intent to induce infringement.

What that changed. It removed the practical compulsion to obtain opinions on every identified patent, which had made clearance ruinously expensive.

What an opinion still does. A competent opinion is affirmative evidence of good faith if the company chooses to rely on it. Reliance waives privilege on the subject matter, which is a real cost.

When it is worth it. Where the exposure is large, where the analysis is genuinely favorable, and where the company would want to put its state of mind in issue. Not as a routine output of clearance.

What competent means. Written, by counsel with relevant expertise, addressing the specific claims against the specific product, considering the file wrapper, and reaching a reasoned conclusion. An opinion that assumes facts or reaches a conclusion without analysis is worse than none.

The alternative. A documented internal analysis showing the patents were identified, charted, and decided upon. It does not carry the same evidentiary weight and it costs a fraction, and for most identified patents it is the right level of investment.

Enhanced damages under 35 U.S.C. § 284 turn on egregious conduct, and a company that identified a patent, analyzed it, and made a reasoned decision is not the profile that attracts enhancement — whether or not an opinion exists.

Standard-Essential Patents

Why they are different. A product implementing a published standard frequently practices claims declared essential to it. No amount of design-around avoids them, because the standard requires the technique.

What makes them manageable. Participants in standards development typically declare essential patents and commit to license them on defined terms. That commitment is what makes implementation possible.

What the analysis requires. Identify the standards the product implements, obtain the declaration databases, and understand which holders are active licensors. This is a licensing exercise rather than a clearance one.

What the commitment limits. Injunctive relief is constrained where a commitment exists and the implementer is willing to take a license, though the boundaries are contested.

Pool licensing covers many essential patents in one transaction for several major standards, and it is frequently the efficient answer.

See Standard-Essential Patents and FRAND Toolkit.

Supply Chain and Indemnity

Components carry their own exposure and their own protection. A module bought from a supplier may be covered by a supplier indemnity, and clearing it at the company's expense duplicates work.

Read the indemnity before searching. Scope, cap, procedures, and whether it survives. A cap tied to amounts paid for the component is common and is usually an order of magnitude below patent exposure.

Tender promptly if a claim arrives. Most indemnities require prompt written notice and late tender forfeits them.

Ask suppliers what clearance they did. A supplier that cleared its component and will say so in writing has reduced the company's scope meaningfully.

Watch the flow-down. A customer demanding an infringement warranty from the company should be met with the same warranty from its suppliers, and the mismatch between what is given and what is received is where companies absorb risk unknowingly.

The customer warranty question. A representation that a product infringes no third-party patent asserts a negative about an unbounded universe. The negotiated position is knowledge-qualified, capped, with defined procedures and carve-outs for customer specifications and combinations.

Clearance in Diligence

Buyers ask. Whether clearance was done, when, by whom, and what it found.

What a good answer looks like. A scoping memorandum, a search report, charts on the identified patents, and a decision record. The absence of clearance is a discount; the presence of clearance with unaddressed findings is worse.

What a bad answer looks like. "We are not aware of any infringement," which asserts ignorance as diligence.

Privileged material. Clearance analyses are typically privileged, and producing them in diligence risks waiver. The usual answer is a summary prepared for the purpose, with the underlying analysis available under a common interest arrangement if the transaction proceeds.

Timing. Clearance conducted during diligence is late, rushed, and expensive. A target that ran it before going to market has an asset.

Building the Program

A trigger in the product development process. New product, major feature, new territory, or a change to a cleared design. Routed to counsel automatically rather than on request.

A standing search relationship with a searcher who knows the technology.

A tiered response. Full clearance for flagship products and new categories; a targeted search for incremental features; a documented decision not to search where the exposure is genuinely low.

A watch service on key competitors and known assertion entities, monitoring published applications and issuances.

A decision register recording every identified patent, the analysis, the decision, and the date.

A design-around budget held against the possibility, because a change priced before launch costs a fraction of the same change afterward.

Supplier terms that require indemnity, require notice, and require suppliers to state what clearance they performed.

Refresh on a cycle. A clearance is a snapshot. For a product with a long life, an annual refresh against newly issued patents is the discipline that prevents a surprise in year four.

Kettleworth's Four Requests, Answered

"A patent clearance on the product." Two hundred features is not a scope; it is an unbounded bill. The answer is a scoping conversation with engineering that produces a list of ten to fifteen features that are novel, distinctive, or in known-crowded territory — the compression scheme, the failover mechanism, the calibration approach — plus the territories where the product will actually be sold and manufactured. The four third-party software components go to the open source analysis rather than to the searcher. The communications module goes to the supplier, with a request for whatever clearance it performed and a review of the indemnity in the supply agreement. What remains is a finite, answerable search.

The customer warranty. A representation that the product infringes no third-party patent asserts a negative about an unbounded universe, and no company can honestly give it. What can be given is a knowledge-qualified representation, an indemnity with a cap and defined procedures, and carve-outs for the customer's own specifications and for combinations the company did not supply. Where the customer will not move, the price should reflect what is being underwritten — and the company should hold matching indemnities from its own suppliers, because the gap between what it gives and what it receives is exactly the risk it absorbs.

The opinion question. 35 U.S.C. § 298 means the absence of an opinion proves nothing about willfulness, which removed the old compulsion to obtain one for every identified patent. The right answer is a documented internal analysis for each identified patent — charted, decided, dated — and a formal written opinion reserved for the small number where the exposure is large and the analysis is genuinely favorable. Reliance on an opinion waives privilege on the subject matter, which is a real cost and one the board member's question does not account for.

The wireless standard. This is the most important of the four and the only one nobody framed as a legal question. Implementing a published standard means practicing declared-essential claims, which no design-around avoids. The answer is not a clearance search; it is to identify the standard, obtain the declaration data, determine which holders actively license, and evaluate whether a pool covers the relevant patents. This is a licensing budget line rather than a clearance finding, and discovering it eleven weeks before launch is late.

The eleven-week problem. Clearance done properly takes six to twelve weeks. With eleven weeks to launch, the honest plan is a targeted search on the highest-risk features now, the standard-essential analysis in parallel, and a full clearance completed after launch with a design-around budget held in reserve. That is a worse outcome than starting six months earlier and it is better than either a rushed search nobody trusts or no search at all.

What It Costs

Scoping. A few hours of counsel time with engineering, and it determines everything downstream.

The search. Scales with the number of features and territories. A tightly scoped search on a dozen features in three jurisdictions is a modest, predictable cost. An unscoped search on a complex product is not, and the resulting report is unusable at any price.

Screening. Lawyer time reducing hundreds of results to a chartable set. Roughly a week for a well-scoped search.

Charting. The main expense, and it scales with the number of candidates rather than with the size of the search. Five to twenty patents charted properly, with file wrapper review on the close ones.

Opinions. Substantial per patent, which is why they are reserved rather than routine.

Design-arounds. Engineering cost and schedule, and the number that matters most in the whole exercise, because it caps the value of every blocking patent and it caps the settlement value of any assertion.

Refresh. An annual pass against newly issued patents, at a fraction of the initial cost.

The allocation rule. Spend on scoping and charting; economize on search breadth rather than on analytical depth; reserve opinions; and hold a design-around budget rather than an opinion budget. A company that identified ten patents, charted them, decided on each, and documented it has spent less and is better positioned than one holding two opinions and no register.

Designing Around, Properly

The design-around is the most valuable output of clearance and it is done badly more often than it is done well.

Design around the claim, not the patent. Read the independent claim, identify the limitation that is easiest to avoid, and change that. Changing something the claim does not recite accomplishes nothing.

Avoid the limitation completely, not partially. A modification that arguably falls outside the literal claim but performs the same function in the same way to achieve the same result invites an equivalents theory. The design-around should be different in a way that can be explained in a sentence.

Read the file wrapper first. Where the applicant narrowed the claim during prosecution to obtain allowance, the surrendered territory is unavailable to an equivalents theory, and a design-around that lands in that territory is unusually safe. This is the single most useful piece of information in a design-around analysis and it is public.

Check the dependent claims and the family. A design-around that avoids claim 1 but reads on a dependent claim, or on a claim in a sibling patent, has accomplished nothing.

Check whether a live continuation exists. A patentee with a pending continuation can draft claims against the design-around. This does not make the design-around pointless; it means the analysis has to be repeated when new claims issue.

Document the change and its date. The date establishes when exposure ended, which matters for the damages period if an assertion ever comes.

Price it honestly and early. A change made before tooling and certification costs a fraction of the same change afterward, and the price of the design-around is what caps the value of the blocking patent in any negotiation.

Do not describe it as a design-around in engineering documents without counsel's involvement, because those documents are discoverable and a memorandum saying "changed to avoid the Hallowell patent" establishes knowledge of the patent.

When Clearance Should Not Be Run

Not every product warrants a clearance, and saying so is part of the advice.

Where the field is uncrowded and the technology is genuinely novel, a targeted search on the two or three closest competitors may be sufficient.

Where the product is a minor variant of something already cleared, a delta search on the changed features answers the question at a fraction of the cost.

Where the commercial exposure is small. A product line with modest revenue does not justify a full clearance, and the honest recommendation is a targeted search plus a documented decision.

Where the timeline makes it useless. A clearance delivered after launch informs the next product rather than this one, and the client should know that before commissioning it.

Where the answer will not change the decision. A company that will ship regardless is buying knowledge it may prefer not to have, because identification is what creates the knowledge element for indirect infringement and for willfulness. This is uncomfortable and it is real, and the response is not to avoid knowing — it is to know, analyze, and decide, which is exactly what a documented clearance produces.

Where a supplier has already done it. A component cleared by its maker, with the clearance confirmed in writing and an indemnity behind it, does not need re-clearing.

In each case, document the decision not to search. A reasoned choice recorded contemporaneously reads very differently from a gap.

What Clearance Cannot See

The limits deserve stating explicitly, because a client that believes clearance is exhaustive will be surprised in a way that damages the relationship.

The eighteen-month window. Applications publish at eighteen months from priority under 35 U.S.C. § 122. Everything filed since is invisible, and a competitor working in the same area has filings the search cannot reach.

Non-published applications. Where a non-publication request was filed and no foreign filing occurred, an application can remain invisible until it issues.

Claims that do not yet exist. A published family with a live continuation can produce claims drafted specifically against the product as it ships. The search sees the disclosure; it cannot see the claim that will be written next year.

Foreign-language art in unindexed jurisdictions, and classification errors that put a relevant patent where nobody searched.

Method claims that read on how a customer uses the product rather than on the product itself. These require thinking about the customer's operation, not only the article, and they are missed routinely.

Indirect theories. A product that does not infringe directly may support an inducement or contributory claim depending on instructions, marketing, and the absence of substantial non-infringing uses. 35 U.S.C. § 271.

Patents held by parties nobody thought to search. Assignment records lag, portfolios are sold, and a patent cleared as held by an inactive company may be held next year by an entity that asserts.

What follows from this. Clearance reduces risk; it does not eliminate it. The correct framing for a client is a ranked risk assessment with decisions attached, refreshed on a cycle, and the wrong framing is a clean bill of health. Practitioners who promise the second create the problem that arrives in year three.

The Conversation to Have First

Before any work is commissioned, four things should be said to the client plainly, because saying them afterward sounds like an excuse.

Clearance produces a risk assessment, not a clearance. The deliverable is a ranked list with decisions attached. There is no document that says the product is safe, and anyone offering one is describing something they cannot deliver.

Scope drives cost, and the client controls scope. A list of a dozen features produces a finite, useful search. "The product" produces an unbounded bill and an unusable report. The scoping conversation is the client's decision to make, with counsel explaining the trade.

Identification creates knowledge. A patent found in clearance is a patent the company knows about, which matters for indirect infringement under 35 U.S.C. § 271 and for willfulness under 35 U.S.C. § 284. The answer is not to avoid looking; it is to look, analyze, decide, and document — which is what converts knowledge into evidence of good faith rather than evidence of disregard.

Timing is not negotiable. Six to twelve weeks for a proper clearance, and a request three weeks before launch buys a partial answer at full price. The right response is a targeted search now and a full one afterward, with a design-around budget held in reserve.

And one thing to ask the client. What will you do differently depending on the answer? A company that will ship regardless is buying a record rather than a decision, which is a legitimate purchase and a different one — and it changes what the deliverable should look like.

Refreshing a Clearance

A clearance is a snapshot of a moving landscape, and products outlive it.

Set the cycle by product life. Annual for a product shipping for a decade; at each major revision for anything iterating quickly.

Search only the delta. Patents issued or applications published since the last pass, in the same classifications, from the same assignee watch list. A fraction of the original cost.

Re-examine the ones previously accepted. A patent accepted as weak may have been reinforced by a continuation with better claims, or a holder previously inactive may have been acquired by an entity that asserts.

Re-examine the ones previously designed around, because a continuation can produce claims drafted against the design-around.

Update the register with the date, the scope of the refresh, and any new decisions.

Watch the assignee landscape, because portfolios move and a patent held by a customer's supplier last year may be held by an assertion entity this year.

Fold in what the business learned. Patents cited in customer requests for proposal, marked competitor products, and anything raised in a partner's diligence all belong in the next pass and none of them appears in a database search.

A Suggested Reading Path

If you are commissioning clearance:

If you found something:

If a letter has already arrived:

Primary Authorities

| Authority | Proposition | |---|---| | 35 U.S.C. § 271 | Infringing acts; the question clearance asks | | 35 U.S.C. § 154 | Term; when a patent stops mattering | | 35 U.S.C. § 122 | Publication at eighteen months; the invisible period | | 35 U.S.C. § 282 | Presumption of validity; defenses | | 35 U.S.C. § 284 | Damages; enhancement | | 35 U.S.C. § 285 | Fees in exceptional cases | | 35 U.S.C. § 286 | Six-year limitation | | 35 U.S.C. § 287 | Marking and notice | | 35 U.S.C. § 298 | Advice of counsel not required | | 35 U.S.C. § 311 | Inter partes review as a response | | 35 U.S.C. § 102 | Novelty; validity of a blocking patent | | 35 U.S.C. § 103 | Obviousness | | 35 U.S.C. § 112 | Definiteness; a blocking claim's boundaries | | 35 U.S.C. § 41 | Maintenance fees; lapse | | 35 U.S.C. § 261 | Ownership; who could assert | | 37 C.F.R. § 1.321 | Terminal disclaimers; early expiry | | 37 C.F.R. § 1.56 | Duty of disclosure; art found in clearance | | Fed. R. Evid. 502 | Waiver; opinion reliance and diligence disclosure |

Forms and Templates

The Portfolio Inventory Template doubles as the clearance decision register, tracking each identified patent, its claims, the analysis, the decision, and the refresh date — which is the document a buyer's counsel will ask for and the one that turns a series of judgments into a defensible record. The License Agreement Template is the starting point where the decision is to license rather than design around, and reading it before the approach clarifies what terms the company actually needs. The Cease and Desist Template is worth reading from the other side, because understanding how an assertion is framed informs how a clearance should be documented. The Assignment Agreement Template matters where a design-around produces new inventions the company will want to own.

Related Toolkits and Checklists

For the doctrinal background — what a claim covers and how infringement is judged — the Patent Fundamentals Toolkit is the companion. Where the product implements a standard, the Standard-Essential Patents and FRAND Toolkit handles the analysis clearance cannot. Where a blocking patent looks vulnerable, the PTAB Practice Toolkit covers the challenge. Where a demand letter has already arrived, the Patent Assertion Defense Toolkit takes over. And where the decision is to take a license, the Patent Licensing and Technology Transfer Toolkit supplies the negotiation.