Freedom-to-Operate Checklist: Scope, Search, Claim Charting, and Design-Around

By Casey Scott McKay · 2024-06-21

A freedom-to-operate analysis fails in predictable places: the wrong search is run, the product is never decomposed, hits are charted before they are screened, the file wrapper is never read, and the report arrives after the design is frozen. This checklist runs the exercise in twelve phases to prevent each of them - scope, decompose, define jurisdictions, search, screen, chart, read the file wrapper, assess invalidity where it matters, design around, decide about opinions, report and allocate, then refresh. Each box gives the reason, the authority, and the trap. A worked matter follows one product from scoping to launch. The companion article explains what the exercise can and cannot establish, and the companion guide walks the method.

IP and Technology > Patent Counseling Transactions | Checklist | Published 21 June 2024 - Updated 13 October 2025 | Casey Scott McKay - marksy.us

Summary. A freedom-to-operate analysis fails in predictable places: the wrong search is run, the product is never decomposed, hits are charted before they are screened, the file wrapper is never read, and the report arrives after the design is frozen. This checklist runs the exercise in twelve phases to prevent each of them — scope, decompose, define jurisdictions, search, screen, chart, read the file wrapper, assess invalidity where it matters, design around, decide about opinions, report and allocate, then refresh. Each box gives the reason, the authority, and the trap. A worked matter follows one product from scoping to launch. The companion article explains what the exercise can and cannot establish, and the companion guide walks the method.

Keywords: freedom to operate checklist, FTO scoping, feature decomposition, patent screening, claim charting, file wrapper review, design around, doctrine of equivalents, prosecution history estoppel, continuation monitoring, opinion of counsel, privilege, supplier indemnity, product launch gate, jurisdiction selection, risk assessment, invalidity screen, standards declaration, refresh cycle, report drafting

How to use this checklist

| Phase | What it covers | Who owns it | When | |---|---|---|---| | 1 | Scope the engagement | Counsel + client | First | | 2 | Define the product | Counsel + engineering | Before searching | | 3 | Decompose and rank features | Counsel + engineering | Before searching | | 4 | Define jurisdictions and horizon | Counsel + business | Before searching | | 5 | Search | Counsel or searcher | Weeks 2–4 | | 6 | Screen | Counsel | Immediately after | | 7 | Chart | Counsel + engineering | On the short list | | 8 | Read the file wrapper | Counsel | On each charted patent | | 9 | Assess invalidity, selectively | Counsel | Real risks only | | 10 | Design around | Engineering + counsel | Before design freeze | | 11 | Report, opine, allocate | Counsel | End | | 12 | Refresh | Counsel | Ongoing |

Phases 1 through 4 determine the cost. Phase 6 determines whether the budget reaches the patents that matter. Phase 10 is the only phase that removes risk rather than measuring it. Boxes marked [Gate] should clear before the next phase begins.

The matter. Kellerman Instruments planned a launch in five months for a portable spectrometer with a new calibration routine. Counsel was asked for "a patent search." The engagement began with the question of which search, and the answer changed the project.

Phase 1. Scope the engagement

[ ] [Gate] Establish which question is being asked.
- Why. A patentability search examines disclosures and asks whether the client can obtain a patent. An FTO analysis examines in-force claims and asks whether the client can sell. They find different documents.
- Authority. 35 U.S.C. § 271(a) defines the acts an FTO analysis charts against.
- Trap. "We already did a search," which almost always means the other one. See Prior Art and Patentability Checklist.
[ ] Establish the trigger and the timeline.
- Why. A launch in five months permits a design-around; one in five weeks does not.
- Trap. Accepting a deadline that makes the exercise pointless without saying so.
[ ] Explain the four service tiers and their prices.
- Why. A landscape scan, a knockout screen, a full analysis with charts, and written opinions are different products.
- Trap. Quoting one and delivering another.
[ ] Set the expectation in writing: risk assessment, not clearance.
- Why. No search can establish that nothing exists, because eighteen months of filings are unpublished under 35 U.S.C. § 122(b).
- Trap. A client who will read any report as permission.
[ ] Identify who will act on the report.
- Trap. A report written for a board that engineering cannot use.

Phase 2. Define the product

[ ] [Gate] Obtain the actual technical specification.
- Why. Marketing descriptions omit implementation detail and overstate function in both directions.
- Trap. Charting against a datasheet, which is the most common way an FTO analysis reaches a confident wrong answer.
[ ] Assign an engineer to the project.
- Why. The analysis is unworkable without one, at every phase from decomposition to design-around.
- Trap. An engineer nominally assigned and never available.
[ ] Separate what is built from what is supplied.
- Why. Supplied components under indemnity are a contract question rather than a search question, and separating them shrinks the scope immediately.
- Trap. Searching a commodity module from a major vendor.
[ ] Identify what will change before launch.
- Trap. Analyzing a configuration slated for redesign.
[ ] Record the configuration analyzed, precisely.
- Why. It becomes the scope statement and it bounds the report's conclusions.

Phase 3. Decompose and rank features

[ ] [Gate] Break the product into eight to fifteen features.
- Why. This is the largest cost lever in the exercise. A search scoped to "the product" is simultaneously unaffordable and too shallow to be useful.
- Trap. Skipping it because the product feels like one thing.
[ ] Rank each on novelty, supply status, landscape density, and commercial necessity.
- Why. Risk is not evenly distributed, and depth should follow rank.
- Trap. Uniform depth across features.
[ ] Allocate depth by rank.
- Why. Commodity supplied features get a note. Novel features in dense landscapes get full charts. The middle gets a screen.
- Trap. Charting everything, which exhausts the budget on the wrong patents.
[ ] Record the ranking and the reasoning.
- Why. It becomes the scope statement, and it is the defence if something later surfaces in an unsearched feature.

Phase 4. Define jurisdictions and horizon

[ ] List every country where the client makes, uses, sells, offers, or imports.
- Authority. 35 U.S.C. § 271(a); patents are national.
- Trap. A single search assumed to cover a multi-market launch.
[ ] Add markets planned within the horizon.
- Why. A design-around adopted now is far cheaper than one adopted at market entry.
[ ] Check the supply chain for cross-border exposure.
- Authority. 35 U.S.C. § 271(f) for components supplied abroad for combination; 35 U.S.C. § 271(g) for products made abroad by a patented process.
- Trap. Contract manufacturing abroad, assumed to be outside U.S. reach.
[ ] Set the time horizon against the product's commercial life.
- Trap. Treating a patent expiring in eight months the same as one with a decade to run.

Phase 5. Search

[ ] Search claims, not disclosures.
- Trap. Reusing a patentability search strategy.
[ ] Search by classification and by concept.
- Why. Each misses what the other catches.
[ ] Search in-force patents and published applications.
- Why. A published application may issue with claims covering the product, and that is the forward-looking half of the exercise.
- Trap. Ignoring applications because they are not yet enforceable.
[ ] Search assignee names for known competitors and assertion entities.
[ ] Search citation networks around the closest hits.
[ ] Search each jurisdiction separately, in the appropriate language.
[ ] Document databases, queries, classifications, and dates as you go.
- Why. It becomes the scope statement and the evidence of diligence.

Phase 6. Screen before charting

[ ] [Gate] Check status: expired, lapsed, or abandoned.
- Authority. 35 U.S.C. § 41(b).
- Trap. Charting a patent that lapsed for unpaid maintenance fees.
[ ] Compute expiry properly, accounting for terminal disclaimers and term adjustment.
- Authority. 35 U.S.C. § 154(b).
[ ] Read the independent claims only.
- Why. A product avoiding the independent claim avoids everything depending from it.
- Trap. Reading dependent claims first and alarming the client.
[ ] Exit on any single clearly missing element.
- Why. Infringement requires every limitation.
[ ] Check the assignee and whether they assert.
- Why. An operating competitor, a university, a standards participant, and an assertion entity present four different risks and four different resolutions.
[ ] Check for a standards declaration.
- Why. A commitment to license on fair, reasonable, and non-discriminatory terms changes both injunction risk and royalty range.
- Authority. See Standard-Essential Patents and FRAND Toolkit.
[ ] Check post-grant history.
- Why. A patent that survived an inter partes review is materially stronger; one with a pending petition may resolve itself.
- Authority. 35 U.S.C. § 311. See The Second Look.
[ ] Reduce to a short list of ten to thirty.
- Trap. A project that charts everything it finds.

Phase 7. Chart the short list

[ ] One chart per independent claim, element by element.
- Trap. An impressionistic assessment of whether the patent "covers" the product.
[ ] [Gate] Chart with the engineer, against the product.
- Trap. Charting alone against a document.
[ ] State plainly where nothing corresponds.
[ ] Resolve the preamble question.
- Why. Whether a preamble limits depends on whether it recites essential structure or merely states purpose.
[ ] Identify means-plus-function limitations.
- Why. A functional limitation may be construed to the disclosed algorithm or structure and its equivalents, which is far narrower than the words suggest.
- Authority. 35 U.S.C. § 112(f). See The Bargain of Disclosure.
[ ] Chart the doctrine of equivalents separately for each missing element.
- Why. Avoiding the literal claim is not enough if the substitute performs substantially the same function in substantially the same way for substantially the same result.
- Authority. See Proving Patent Infringement.

Phase 8. Read the file wrapper

[ ] Find every narrowing amendment.
- Why. Surrendered territory is generally unavailable to the patentee under prosecution history estoppel, which narrows the claim and forecloses equivalents in that space.
- Trap. Skipping this because it is not in the search results. It is the cheapest leverage in the exercise.
[ ] Find the arguments distinguishing prior art.
- Why. Statements about what the claim does not cover are binding characterizations, and they frequently describe the client's product.
[ ] Find disclosed but unclaimed subject matter.
- Why. Material disclosed and not claimed is generally dedicated to the public, which supports a design-around aimed at a disclosed alternative.
[ ] Find any terminal disclaimer.
- Authority. 37 C.F.R. § 1.321. It affects expiry and conditions enforceability on common ownership.
[ ] [Gate] Check for pending continuations in the family.
- Why. A live continuation lets the patentee draft claims against your product once it is public, and no snapshot of issued claims reveals it.
- Authority. 35 U.S.C. § 120. See Inside Patent Prosecution.
- Trap. The risk clients are least prepared for.

Phase 9. Assess invalidity, selectively

[ ] Only for patents that chart as a real risk.
- Why. Invalidity analysis is expensive and does not clear the path.
- Authority. 35 U.S.C. § 282.
[ ] Look for prior art the examiner did not consider.
- Why. It is the strongest invalidity material.
[ ] Look for Section 112 exposure where the claim is much broader than the disclosure.
- Authority. 35 U.S.C. § 112(a). See Section 112 Compliance Checklist.
[ ] Look for eligibility exposure in software and diagnostic claims.
- Authority. 35 U.S.C. § 101. See What Can Actually Be Patented.
[ ] Note which forum would hear which ground.
- Authority. Prior art to an inter partes review under 35 U.S.C. § 311; eligibility and Section 112 to court or to post-grant review under 35 U.S.C. § 321.
[ ] Use it as a risk input, not a conclusion.
- Trap. Telling a client a patent is invalid when what you mean is that it has exposure.

Phase 10. Design around

[ ] Work from the chart, element by element.
- Why. The objective is to remove one element entirely, not to make the product generally different.
- Trap. Five changes that leave every claimed element in place.
[ ] Pick the element cheapest to change and hardest to call equivalent.
- Why. Removing a step is stronger than substituting a component.
[ ] [Gate] Test every candidate against the doctrine of equivalents before recommending it.
- Trap. A design-around that reads as an insubstantial difference.
[ ] Aim at surrendered territory where the file wrapper permits.
[ ] Cost each option with engineering.
- Trap. A legally elegant solution nobody can build.
[ ] Re-chart after implementation.
- Why. Engineering drifts between the decision and the shipped product, and a design-around implemented differently is not one.
- Trap. Signing off on the design intent rather than the build.

Kellerman, Phase 10. Two patents charted as real risks. One resolved on a narrowing amendment in the file wrapper that had surrendered exactly the configuration Kellerman used. The other required moving a computation from runtime to calibration time — three engineering weeks, adopted eleven weeks before design freeze, and unavailable at any price four months later.

Phase 11. Report, opine, and allocate

[ ] State the scope prominently: configuration, jurisdictions, databases, classifications, dates, exclusions.
- Trap. A reader who assumes the search was total.
[ ] State the limits prominently.
- Why. Unpublished applications, claims not yet drafted, classification and vocabulary gaps, and untested constructions.
- Authority. 35 U.S.C. § 122(b).
[ ] Organize by feature, not by patent.
[ ] Give each patent a risk level with a one-sentence reason.
- Trap. "Low risk" with no reasoning, which is not defensible later.
[ ] Decide about a written opinion, selectively.
- Why. It is evidence of reasonable conduct relevant to enhanced damages, and relying on it waives privilege over it and potentially related communications.
- Authority. 35 U.S.C. § 284; 35 U.S.C. § 298 provides that failure to obtain advice is not evidence of willfulness.
- Trap. Commissioning opinions routinely, or using litigation counsel to write them.
[ ] Mark the report, limit distribution, and separate legal analysis from business planning.
- Trap. Wide internal circulation, which is how privilege is lost.
[ ] [Gate] Allocate residual risk by contract.
- Why. The analysis identifies risk; the contract decides who carries it.
- Authority. See IP Insurance and Risk Transfer Toolkit.
- Trap. Identifying supplier-originated risk and never asking the supplier to indemnify it.
[ ] Instruct engineers on written commentary.
- Trap. Internal emails speculating about infringement, created by people nobody told.

Phase 12. Refresh

[ ] Date the report and state its shelf life.
- Trap. A four-year-old report still relied on as current.
[ ] Monitor the specific families identified as risks.
- Why. Cheap, and it catches continuation claims drafted against your shipped product.
[ ] Refresh before each significant product change, market entry, or investment.
[ ] Refresh annually for a long-lived product, at lighter depth.
[ ] Record refresh decisions, including decisions not to refresh.
[ ] Feed findings back into the design-around library and the feature-risk map.
- Why. Most products share architecture with their predecessors, and re-deriving solutions wastes engineering time.

Phase 13. Sector-specific screens

The general phases hold; the risk profile and the practical moves differ by field.

Software and platform products

[ ] Decompose by function and by layer, not by product module.
- Why. Patents in this space claim protocols, data structures, and processing techniques rather than product features, and a decomposition mirroring the engineering org chart will miss them.
- Trap. Searching "our analytics platform."
[ ] Check whether the claim is a method performed across parties.
- Why. A method claim requiring steps by both a server operator and an end user is materially harder to assert.
- Authority. See Building or Defeating a Patent Infringement Case.
[ ] Screen every functional limitation for a means-plus-function construction.
- Why. Where the specification discloses no algorithm, the claim may be indefinite; where it discloses one, the claim narrows to it.
- Authority. 35 U.S.C. § 112(f).
[ ] Screen software claims for eligibility exposure.
- Authority. 35 U.S.C. § 101. A patent that would not survive a motion is a different risk from one that would.
[ ] Check open source dependencies separately.
- Why. Some licenses carry patent grants and defensive termination provisions that change the analysis entirely.
- Authority. See Software, Data, and Open Source Toolkit.

Life sciences and medical devices

[ ] Search composition, method-of-making, method-of-use, and formulation claims separately.
- Trap. Clearing the compound and missing a method-of-treatment claim.
[ ] Check regulatory exclusivities alongside patents.
- Why. Data and market exclusivities operate independently of patent rights and can block a launch on their own timeline.
[ ] Check whether any listed patent covers the product.
- Why. Listing carries its own procedural consequences on approval pathways.
[ ] Assess divided infringement for method-of-treatment claims.
- Why. Steps performed by a physician rather than by the manufacturer raise real proof problems for the patentee, and inducement becomes the theory.
- Authority. 35 U.S.C. § 271(b).

Hardware, components, and manufacturing

[ ] Search the process as well as the product.
- Authority. 35 U.S.C. § 271(g) reaches importation of a product made abroad by a patented process.
- Trap. Clearing the device and importing it from a plant using a patented method.
[ ] Obtain indemnities from component suppliers and verify they reach patent claims.
- Trap. A general indemnity that excludes IP, or one from an entity that cannot pay.
[ ] Check standards participation for every interface the product implements.
- Authority. See Standard-Essential Patents and FRAND Toolkit.

Phase 14. Cost, and where it goes

| Item | Relative cost | What determines it | |---|---|---| | Scoping and product definition | Low | One or two meetings | | Feature decomposition and ranking | Low | The single largest cost lever | | Landscape scan | Low | Breadth of the technology area | | Knockout search, per feature per jurisdiction | Low to moderate | Number of features times jurisdictions | | Screening | Low | Volume of hits, and it removes most | | Claim charting, per patent | Moderate | Claim complexity and product complexity | | File wrapper review, per patent | Low | High return per hour | | Invalidity assessment, per patent | High | Depth of the prior art search required | | Design-around development | Variable | Engineering effort, not legal | | Written opinion, per patent | High | Reserved for genuine risks | | Report drafting | Moderate | Number of features and patents covered | | Annual refresh | Low to moderate | Much cheaper than the original |

Three observations that shape the budget.

The decomposition determines everything downstream. Fifteen features across three jurisdictions is not five times the cost of three features across one — it is closer to fifteen times, because search, screening, charting, and reporting all scale with the product of the two.

Screening is the cheapest phase and it removes the most work. A project that reaches charting with three hundred patents has skipped it.

The design-around is the only phase that reduces risk. Everything else measures it. A budget that funds a thorough analysis and no engineering time to act on it has bought a description of a problem.

And the timing point, which is not about money. The same analysis run at design freeze minus one quarter and at design freeze plus one month produces identical findings and entirely different options. Practitioners cannot control when they are engaged, and they can say plainly at the first meeting what the difference costs.

Phase 15. The FTO gate in a product development process

For a company shipping products regularly, FTO should be a gate rather than a project, and the design of that gate determines whether it works.

[ ] Place the gate at design freeze minus one quarter.
- Why. Late enough that the configuration is stable, early enough that a design-around is three engineering weeks rather than a tooling change.
- Trap. A gate at launch readiness, which produces findings nobody can act on.
[ ] Make the gate a required artifact, not a recommendation.
- Why. Checks that depend on someone remembering are skipped on exactly the programs under the most schedule pressure.
- Trap. A policy in a memo and not in the stage-gate template.
[ ] Maintain a standing feature-risk map across products.
- Why. Most products share architecture with their predecessors, and last cycle's analysis covers much of this cycle's product.
- Trap. Starting from zero each time, which makes the gate unaffordable and therefore skippable.
[ ] Maintain a design-around library.
- Why. Solutions developed once frequently transfer, and re-deriving them wastes engineering time.
[ ] Monitor the families already identified as risks.
- Why. Watching a dozen specific families for new continuations and issuances costs very little and catches the risk that snapshots structurally miss.
- Authority. 35 U.S.C. § 120.
[ ] Standardize supplier indemnity language at procurement.
- Why. The contract allocation should happen when the component is sourced, not when the analysis finds the risk.
[ ] Set the escalation threshold in advance.
- Why. Which findings go to the general counsel, which to product leadership, and which resolve at engineering level. Without a threshold, everything escalates or nothing does.
[ ] Train engineers on written commentary.
- Why. Internal speculation about whether a competitor's patent is infringed becomes a discoverable document, created by people nobody instructed.
- Trap. Discovering the emails in litigation.
[ ] Report the program's outputs annually.
- Why. Design-arounds adopted, risks retired, indemnities obtained, and families being monitored. It is what keeps the program funded, because it shows outcomes rather than activity.

Phase 16. When the analysis follows a demand letter

An FTO analysis run after contact is a defense rather than a plan, and four things change immediately.

[ ] [Gate] Recognize that knowledge is now established.
- Why. Inducement and contributory infringement both require knowledge, and the letter supplies it. A component supplier with no direct exposure may now have indirect exposure it did not have last week.
- Authority. 35 U.S.C. § 271(b); 35 U.S.C. § 271(c).
[ ] Recognize that willfulness exposure has begun.
- Why. Continuing conduct after notice without a reasoned basis is the fact pattern enhanced damages address.
- Authority. 35 U.S.C. § 284.
- Trap. Waiting three months to analyze while shipping continues.
[ ] Assess whether a declaratory judgment action is available and desirable.
- Why. A letter asserting rights and demanding cessation generally creates a case or controversy, which lets the recipient choose the forum.
- Authority. 28 U.S.C. § 2201. See Where Patent Cases Are Fought.
- Trap. A short window, missed while the analysis proceeds.
[ ] Issue a litigation hold.
- Why. Litigation is now reasonably anticipated.
[ ] Do not respond substantively immediately.
- Why. Confirm receipt, identify the patents, and take the time to chart properly. An early substantive response made without a chart is a position you will be held to.
[ ] Tender to insurers.
- Why. IP coverage exists in more policies than clients assume, and late notice is the most common ground of denial.
- Authority. See IP Insurance and Risk Transfer Toolkit; Who Pays for the IP Lawsuit?.
[ ] Look upstream immediately.
- Why. If the accused feature came from a supplier, the indemnity may move the whole problem — and its notice provisions have deadlines of their own.
- Trap. Notifying the supplier after the indemnity's notice window closed.
[ ] Then run Phases 6 through 10 on the asserted patents specifically.
- Why. Screening, charting, file wrapper review, invalidity assessment, and design-around are the same steps, now aimed at a known target rather than a landscape.
- Authority. See Defending a Patent Assertion; Patent Assertion Defense Toolkit.

A closing note on how this checklist gets used badly. The most common misuse is running Phases 5 through 8 thoroughly and Phases 1 through 4 not at all — a rigorous search of a product nobody defined, in jurisdictions nobody listed, at a depth nobody allocated. The result looks like diligence and answers no question the business asked. The second most common is running everything except Phase 10, producing an excellent description of risk with no engineering time budgeted to remove any of it. And the third is running the whole thing once, filing the report, and treating it as permanent, when new patents issue continuously and the families identified as risks are precisely the ones with continuations pending. Scope deliberately, act on the findings, and refresh — and the exercise repays itself many times. Do the middle phases beautifully and skip the ends, and it is an expensive document.

Kellerman, outcome. The product shipped on schedule. Of the fourteen features analyzed, nine required no work, three were supplied under indemnities that were verified rather than assumed, and two produced charts warranting attention. One of those resolved on the file wrapper. The other produced a three-week engineering change adopted before design freeze. Total legal cost: a fraction of one month of the litigation it did not have.

The company now runs the gate at design freeze minus one quarter on every program, and maintains the feature-risk map between cycles.

Key Authorities at a Glance

| Authority | Proposition | Phase | |---|---|---| | 35 U.S.C. § 271(a) | Making, using, offering, selling, importing | 1, 4 | | 35 U.S.C. § 271(b) | Inducement; requires knowledge | 11 | | 35 U.S.C. § 271(c) | Contributory infringement | 3 | | 35 U.S.C. § 271(f) | Components supplied for combination abroad | 4 | | 35 U.S.C. § 271(g) | Products made abroad by a patented process | 4 | | 35 U.S.C. § 101 | Eligibility as an invalidity ground | 9 | | 35 U.S.C. § 112(a) | Enablement and written description | 9 | | 35 U.S.C. § 112(f) | Means-plus-function construction | 7 | | 35 U.S.C. § 120 | Continuations; the invisible risk | 8 | | 35 U.S.C. § 122(b) | Publication at eighteen months | 1, 11 | | 35 U.S.C. § 154(b) | Term adjustment | 6 | | 35 U.S.C. § 282 | Presumption of validity | 9 | | 35 U.S.C. § 284 | Damages and enhancement | 11 | | 35 U.S.C. § 285 | Fees in exceptional cases | 9 | | 35 U.S.C. § 298 | Failure to obtain advice not evidence of willfulness | 11 | | 35 U.S.C. § 311 | Inter partes review grounds | 6, 9 | | 35 U.S.C. § 321 | Post-grant review grounds | 9 | | 35 U.S.C. § 41(b) | Maintenance fees | 6 | | 37 C.F.R. § 1.321 | Terminal disclaimers | 8 | | 28 U.S.C. § 2201 | Declaratory judgment | 11 |

The five things people get wrong

One: they run the wrong search. A patentability search examines disclosures and asks whether a patent is obtainable. An FTO analysis examines in-force claims and asks whether the product can be sold. Reporting the first as clearance for the second is the most consequential confusion in this practice.

Two: they never decompose the product. Searching "the product" produces a project that is unaffordable and shallow at the same time. Eight to fifteen ranked features, with depth allocated by rank, is what makes the exercise both useful and payable.

Three: they chart before screening. Status, expiry, independent claims only, one missing element. Most hits die there for almost nothing, and the budget then reaches the patents that matter.

Four: they skip the file wrapper. Narrowing amendments limit equivalents and frequently convert an alarming claim into a manageable one. It costs almost nothing and it is omitted because it does not appear in search results.

Five: they run it too late. After tooling, contracts, and inventory, the findings are identical and the options are gone. The entire value of the exercise is the window in which design changes are still cheap. See Running a Freedom-to-Operate Analysis.