Carve-Out and Divestiture IP Checklist: Asset Identification and Splitting, Shared and Retained Rights, Transitional Licences and Services, Recordation and Registry Updates, and Brand Migration Milestones

By Casey Scott McKay · 2026-03-15

A ten-phase working checklist for the intellectual property workstream in a carve-out, divestiture, or spin-off, usable on either side. Phases one and two run the asset identification interviews and build the four-bucket schedule. Phases three through five settle the house mark, the patent cross-licences, and the software, data, and know-how layers. Phases six and seven cover the transition services terms and the separation agreement provisions. Phases eight through ten cover recordation, brand migration, and post-closing governance. Each phase closes with a gate.

IP and Technology > IP and IT in Corporate Transactions | Checklist | Published 15 March 2026 - Updated 12 August 2026 | Casey Scott McKay - marksy.us

How to use this checklist

Two facts organise everything below.

The register does not know which business owns anything. It says the parent company. The allocation must come from the operating business, which makes Phase 1 the critical path and the only part of the workstream that cannot be compressed.

Every continuing arrangement needs an end date and a consequence. Transitional licences, transition services, and cross-licences without them become permanent, and a separation whose purpose was independence ends with two companies still contractually joined.

Start Phase 1 at the first serious conversation about a divestiture, not at signing. A seller that begins before a buyer is identified negotiates from knowledge; one that begins after exclusivity negotiates from discovery, and every discovery is a concession.

The doctrinal background is Splitting a Company in Two; the operational treatment with worked engagements is Executing an IP Carve-Out; the cluster is assembled in the Carve-Out, Divestiture, and Brand Separation Toolkit.

Phase 1. Asset identification

[ ] Appoint a single workstream lead with authority to convene the operating business, since a workstream that can only ask will not get diary time.
[ ] Interview by function: research and development, engineering, manufacturing, quality, regulatory, marketing, sales, and information technology. Each holds a different part of the answer.
[ ] Ask what they use, not what they own. "What would stop working on Monday if you lost access to it?" produces a better list than any request for an intellectual property schedule.
[ ] Ask specifically about the unregistered: specifications, formulations, test methods, calibration data, tooling designs, supplier lists, process parameters, and the spreadsheet that runs the plant.
[ ] Ask directly about the shared, since nobody volunteers it and it is the bucket that drives the negotiation.
[ ] Ask who made it: own employees, contractors, a joint development partner, a university, or an acquired business.
[ ] Inventory the software with its licensor and whether anybody has read the assignment clause.
[ ] Interview the divested business's own operating people, not only group functions, since a schedule built from group-level knowledge misses exactly the operating dependencies that become wrong-pocket claims.
[ ] Record interviewee and date, since a later dispute about completeness turns on who said what.
[ ] Reconcile against three sources: the register extract, the contract database, and the finance system's capitalised intangibles.
[ ] Run the schedule as a project with named interviewees, dates, and a weekly completion percentage reported to the deal team.
[ ] Prepare the interview record under privilege where available, since it contains candid statements about what the business does not own.
[ ] [Gate] Every function has been interviewed and the schedule is built from what the business uses rather than from what the register lists.

Phase 2. The four-bucket schedule

[ ] Transfers — used exclusively by the divested business. Identify, schedule, assign, record. Verify the retained business really does not use it.
[ ] Retained — used exclusively by the retained business. Verify from the other direction.
[ ] Shared — needed by both: the house mark, shared platforms, patent families covering common technology, customer databases, operating manuals, enterprise licences. This is where the negotiation and the drafting time go.
[ ] Unresolved — ownership cannot be established. Unregistered rights with no documented creator, software with no licence record, contractor output with no assignment, undocumented know-how, and material from prior acquisitions with broken chains.
[ ] Try to resolve the resolvable before accepting the bucket: locate contractors for confirmatory assignments, check whether predecessor entities still exist, and look in finance systems for licence records. A fortnight typically halves it.
[ ] Choose a mechanism for the remainder: a mutual non-assert, a broad mutual licence, or an allocation with a licence back.
[ ] Say who defends if a third party asserts against an unresolved asset.
[ ] Size the bucket as a number and report it to the deal team weekly, since it is a price input.
[ ] Circulate the schedule to both businesses for challenge, since showing an operating manager a list that omits what they depend on is the fastest way to find a gap.
[ ] Version and date it, since it will change and the agreement must schedule a specific version.
[ ] Keep a decisions log recording each allocation, the reason, and who made it.
[ ] [Gate] Every material asset sits in one of the four columns, and the unresolved bucket has been sized and reported.

Phase 3. The house mark

[ ] State the constraint. A trademark cannot be assigned separately from the goodwill it symbolises — 15 U.S.C. § 1060 — so it cannot simply be allocated to whichever side wants it. See Trademarks in the Deal.
[ ] Choose among three structures: one side keeps it and the other rebrands under a transitional licence; a field or territory split with coexistence terms; or a neutral holder licensing to both.
[ ] Resist a field split between businesses in the same category, since it fails the first time either extends its range.
[ ] Price the rebrand into the transaction where the divested business must change name, since it is a real and quantifiable cost.
[ ] Draft the transitional licence with: goods and services covered; permitted uses including packaging, signage, digital properties, and legal entity names; quality standards actually inspected; territory; term with a priced extension mechanism; and run-off treatment for inventory manufactured before and sold after expiry.
[ ] Include real quality control, since an uncontrolled licence risks abandonment under 15 U.S.C. § 1127 and the Barcamerica International USA Trust v. Tyfield Importers, Inc. exposure runs to the licensor.
[ ] Assume the licence runs long — regulatory approvals, packaging inventory, certifications tied to a name, and signage volume all extend migration — and build extensions in at a price.
[ ] Separate product names and sub-brands from the house mark, since they frequently transfer even where the house mark does not and materially ease the migration.
[ ] [Gate] The mark allocation is documented, the licence has a term and a consequence, and the rebrand cost is in the model.

Phase 4. Patents and cross-licences

[ ] Map the families before allocating, since a family with members directed at both businesses shares a specification and a priority chain.
[ ] Prefer title plus cross-licence to splitting members, since splitting creates two owners of a common priority chain with divergent interests and a shared invalidity exposure.
[ ] Avoid joint ownership, since 35 U.S.C. § 262 permits either owner to exploit and licence without accounting — meaning each side can licence the other's competitors. See Whose Invention Is It.
[ ] Define fields by what the businesses do, not by claim scope, since claims move on amendment and product categories do not.
[ ] Test each field against both businesses' five-year plans, since the argument arrives when one side moves into an adjacent space.
[ ] Avoid negative definitions, which allocate all future uncertainty to one side.
[ ] Address overlap explicitly: permitted, consent-based, or subject to consultation.
[ ] Write a worked example into the recitals, since an illustration of what falls inside and outside is worth several paragraphs of definition.
[ ] Settle enforcement: standing, cooperation, control of litigation, allocation of recoveries, and what happens if the owner declines to enforce in the licensee's field.
[ ] Settle maintenance and abandonment, with a step-in right and an assignment obligation where the owner wants to abandon a member the licensee relies on.
[ ] Address improvements — whether they flow back, in which direction, and on what terms.
[ ] Handle applications in flight: who instructs, who pays, who receives the grant, communicated to the attorneys of record before the next office action.
[ ] Build the renewal schedule with a named payer, since the first post-closing cycle is where transferred rights lapse. Assignment and recordation formalities are at 35 U.S.C. § 261.
[ ] Include a five-year review mechanism — an obligation to discuss rather than to agree — giving both sides a route that does not begin with a letter.
[ ] [Gate] Every shared family has an owner, a field, an enforcement position, a maintenance payer, and an improvements rule.

Phase 5. Software, data, and know-how

[ ] Inventory every application with licensor, transferability, and consent requirement.
[ ] Start consent conversations early, since enterprise agreements are typically entity-specific and vendors know when the deadline is.
[ ] Establish the contractor assignment position for own-developed software, since transfer requires the signed writing at 17 U.S.C. § 204 and the work-for-hire categories at 17 U.S.C. § 101 rarely cover an independent contractor's code. See Who Owns the Work.
[ ] Run an open source inventory for anything shipping in a divested product, since copyleft obligations travel to the acquirer. See Copyleft and Consequences.
[ ] Plan the shared platforms: duplicate, separate under transition services, or leave with one side and grant access.
[ ] Treat data as a separate workstream, with the privacy analysis running alongside since transferring personal data is a processing operation requiring a basis and, in many frameworks, notice. See The State Privacy Wave and the Data Licensing Checklist.
[ ] Collect the digital estate — domains, social accounts, app listings, analytics properties — which are frequently registered to individuals or held by agencies and are cheap to sort out before closing.
[ ] Enumerate the know-how before allocating it, since a business that has never written down what it treats as confidential cannot allocate it and has not met the reasonable measures standard at 18 U.S.C. § 1839 in any event. See Building a Trade Secret Program That Survives Litigation.
[ ] Accept that people carry general skill and knowledge either way, and allocate only specific documented material.
[ ] Deal with cross-holdings of each other's confidential information: return, destroy, or hold under continuing obligation, with a schedule and a deadline.
[ ] Negotiate the residuals clause deliberately, limiting it to unaided memory, excluding deliberate memorisation, and excluding enumerated categories.
[ ] Time-limit the mutual non-solicitation and check enforceability. See Where an Employee Can Go and the Trade Secret Protection and Departure Checklist.
[ ] [Gate] Every application has a consent position, every dataset has a basis, and the know-how exists as a schedule rather than as institutional memory.

Phase 6. Transition services, intellectual property terms

[ ] Identify what is actually provided: shared system access, use of retained software under the seller's licences, hosting, shared data, and continued use of the seller's marks in interfaces and customer communications.
[ ] Resolve the vendor consent question, since a seller providing services using third-party software may be sublicensing without the right and the vendor's agreement frequently prohibits provision to third parties.
[ ] Settle the data roles: controller, processor, permitted use by the seller, and what happens at exit.
[ ] Address improvements made during the term to shared software the buyer uses.
[ ] Write real exit criteria — a defined technical acceptance test with a date, not "migration to be completed within eighteen months".
[ ] Cover reverse services, since the divested business frequently provides something back.
[ ] Prevent the TSA becoming a licence, since indefinite use of retained technology because migration never completed is a perpetual licence nobody negotiated or priced.
[ ] Price extensions rather than allowing them by default.
[ ] [Gate] Every service has an exit criterion with a date, and none of them is aspirational.

Phase 7. Separation agreement provisions

[ ] Incorporate versioned schedules covering transferred rights, retained rights, cross-licensed rights, and licensed marks.
[ ] Include the assignment mechanic, with an obligation to execute jurisdiction-specific instruments and to cooperate in recordation, surviving closing.
[ ] Include the wrong-pocket provision: an asset found to have gone to the wrong side transfers for no additional consideration, with a cooperation obligation. One paragraph, and without it every discovery is a negotiation.
[ ] Attach the transitional trademark licence with the Phase 3 terms.
[ ] Attach the cross-licences with fields, enforcement, maintenance, and improvements settled.
[ ] Include the confidentiality cross-obligations and the residuals position.
[ ] Scope the warranties to what the seller can verify, with a survival period the buyer can live with. Sellers should resist warranting completeness of unregistered rights; buyers should press for a specific warranty on assets the business actually depends on.
[ ] Include a further assurance clause, which is what makes post-closing cooperation enforceable.
[ ] Include the governance mechanism — a named contact on each side and a standing meeting.
[ ] [Gate] The agreement contains a wrong-pocket clause, a further assurance clause, and versioned schedules.

Phase 8. Recordation

[ ] Scope it honestly: a mid-sized portfolio produces thousands of recordations across dozens of jurisdictions, each with its own procedure, forms, fees, and language requirements — patents under 35 U.S.C. § 261, copyright transfers under 17 U.S.C. § 205, and trademarks under the applicable registry rules.
[ ] Take the executable suite before closing, since some registries require original signatures, notarisation, or legalisation and obtaining them afterwards is materially harder.
[ ] Prioritise by exposure: rights in enforcement, rights up for renewal, and rights in jurisdictions where recordation affects enforceability or priority against a subsequent purchaser.
[ ] Expect chain of title breaks, since this exercise finds the prior acquisition whose assignments were never recorded and the intermediate entity that was dissolved.
[ ] Plan the remediation, which requires locating people at predecessor entities and is frequently the longest post-closing item.
[ ] Budget it as a separate line in the post-closing plan rather than absorbing it into deal costs, since it will otherwise be unfunded when the deal budget closes.
[ ] Name an owner, since the deal team disperses and recordation runs for two years.
[ ] Track it visibly, since the buyer's next transaction will ask and an incomplete programme is a diligence finding against the buyer. See the IP Due Diligence Toolkit and the Trademark Portfolio Management Toolkit.
[ ] [Gate] The executed suite exists, the programme has an owner and a budget, and the priority order is set.

Phase 9. Brand migration

[ ] Inventory the estate: signage at every site, vehicle liveries, uniforms, packaging in every format and language, moulds with embossed marks, labels, manuals, safety documentation, invoices, contracts in flight, and marks stamped into products already in the field.
[ ] Map the regulatory dependencies, since in regulated sectors a name change requires re-approval and the approval timeline sets the migration timeline regardless of the licence term.
[ ] Negotiate inventory run-off rather than assuming it.
[ ] Start the naming project early, since clearance can fail and a name that fails in month nine of a twelve-month migration leaves no time.
[ ] Clear across the whole footprint, since a name available at home and blocked in three other markets produces a fragmented brand.
[ ] File before announcing, since the announcement is a disclosure and an invitation in first-to-file jurisdictions. See the International Trademark Toolkit.
[ ] Secure the digital estate simultaneously, since domains and handles are taken within hours of an announcement.
[ ] Consider the endorsed transition — "X, formerly part of Y" — which carries recognition across and is a licensed use requiring permission and a duration.
[ ] Address third-party usage by distributors, resellers, directories, and mapping services, which is a control problem rather than a legal one.
[ ] Budget the search and digital work, since redirects, listings, and app entries carry the old identity for years.
[ ] Set dated milestones with a named owner on each side and a financial consequence for missing them.
[ ] [Gate] The migration plan has dates, owners, and consequences, and the rebrand budget is in the operating plan rather than the legal one.

Phase 10. Post-closing governance

[ ] Name an owner on each side with authority and a budget, since deal teams disperse within months.
[ ] Hold a standing quarterly meeting covering recordation progress, migration milestones, TSA exit criteria, and wrong-pocket items.
[ ] Keep a joint issues log, since the alternative is two parallel accounts of the same disagreements.
[ ] Track the renewal calendar for transferred rights through the first two cycles.
[ ] Escalate before deadlines, since an expired transitional licence puts the buyer in unlicensed use.
[ ] Run a twelve-month schedule review against what each business actually turned out to use, while the wrong-pocket clause is live and people remember.
[ ] Run the same identification exercise on the retained side, which costs a fraction of the work already done and answers the question the seller faces at its next transaction.
[ ] Close out formally when the transitional arrangements end, so no obligations survive by accident.
[ ] [Gate] Somebody on each side is accountable for the obligations that run for years after the deal team has gone.

Structure variants

[ ] Private equity carve-outs add a standalone build; transition services are extensive and the schedule must be complete enough to operate a company that did not exist last week.
[ ] Spin-offs to shareholders remove the adverse counterparty, which sounds easier and is riskier, since nobody negotiates hard enough and both sides live with an ambiguous allocation.
[ ] Regulatory divestitures run to an authority's deadline, may require standalone viability, and may have the package specified in the remedy. Continuing links attract scrutiny; information flows during migration may require clean team arrangements.
[ ] Joint venture unwinds are harder than they look, since the venture agreement's ownership terms are usually thin.
[ ] Insolvency sales remove the seller's cooperation entirely: no wrong-pocket clause, no assistance, no missing documents obtainable. See When Your Licensor Goes Bankrupt and Protecting a Trademark License Against Insolvency.
[ ] Internal reorganisations look trivial and are not: moving rights between group entities is an assignment requiring recordation and is the transaction most likely to be done informally and found as a chain of title break years later.
[ ] [Gate] The checklist has been adapted to the structure rather than applied generically.

The nine buy-side diligence requests

[ ] A schedule of registered rights indicating which business uses each, with an acknowledgement of where that is uncertain.
[ ] A list of what the divested business uses that the seller will retain, which produces the cross-licence schedule and which sellers frequently have not asked themselves.
[ ] The software inventory with licence transferability, since consents are the item most likely to delay closing.
[ ] The contractor assignment position for own-developed software and content.
[ ] The open source inventory for anything shipping in a product.
[ ] The know-how enumeration, or an honest answer that none exists.
[ ] The chain of title for anything acquired, since prior acquisitions are where the breaks are.
[ ] Renewal schedules with next dates, since the first post-closing cycle is where transferred rights lapse.
[ ] The brand estate inventory — sites, vehicles, packaging, moulds, digital properties — which is what prices the migration.
[ ] [Gate] A seller that can answer all nine is prepared; one that cannot has shown the buyer where to concentrate.

The timeline

[ ] Asset identification: six to twelve weeks, and it cannot be compressed by adding lawyers.
[ ] Allocation decisions: four to eight weeks, requiring both businesses to engage while running the operation.
[ ] Document suite: four weeks once allocation is settled.
[ ] Execution: longer than expected, because notarised and legalised documents need signatories, apostilles, and couriers.
[ ] Recordation: one to three years, running past closing.
[ ] Brand migration: two to five years in a regulated sector, one to three otherwise.
[ ] [Gate] The workstream started at the first serious conversation about a divestiture, not at signing.

What a completed separation holds

[ ] A versioned four-bucket schedule built from operating interviews.
[ ] A transitional licence with quality standards, a term, and a priced extension.
[ ] Cross-licences with fields, enforcement, maintenance, and improvements settled.
[ ] A software inventory with consents obtained.
[ ] A know-how enumeration.
[ ] Transition services with real exit criteria.
[ ] An executed and notarised assignment suite.
[ ] A wrong-pocket clause.
[ ] A recordation plan with an owner and a budget.
[ ] A migration plan with dated milestones and consequences.
[ ] [Gate] All ten exist. A separation with them completes; one without the schedule never knows what it transferred, one without the wrong-pocket clause negotiates every discovery, and one without migration milestones stays joined.

Common failures

[ ] The schedule built from the register rather than from the business. Phase 1.
[ ] The unresolved bucket nobody sized, discovered as a price adjustment. Phase 2.
[ ] The house mark split by field between two businesses in the same category. Phase 3.
[ ] The patent family split by member, leaving two owners of one priority chain. Phase 4.
[ ] The software consent conversation started six weeks before closing. Phase 5.
[ ] The TSA with aspirational exit criteria. Phase 6.
[ ] The missing wrong-pocket clause. Phase 7.
[ ] The recordation programme with no owner and no budget. Phase 8.
[ ] The migration plan with no dates and no consequences. Phase 9.
[ ] The transferred right that lapsed in the first renewal cycle because nobody was named to pay. Phases 4 and 10.

Costing the workstream

[ ] Interview and schedule build — the largest single item, largely time rather than expertise, and materially reduced where the client provides internal resource for the logistics.
[ ] Allocation negotiation — expertise-intensive and unpredictable, since it depends on how much sits in the shared bucket.
[ ] Document suite — largely template work once allocation is settled, with the transitional licence and cross-licences as the bespoke items.
[ ] Assignment instruments and execution — scaling with jurisdiction count, administratively rather than analytically expensive.
[ ] Recordation — a per-right cost with official fees, local agent fees, and translation, budgeted as a separate post-closing line.
[ ] Brand migration — largely an operating cost rather than a legal one, but estimated by the legal team because nobody else will.
[ ] Post-closing governance — a small standing cost preventing a large intermittent one.
[ ] [Gate] The budget separates the recordation and migration lines, so they survive the closing of the deal budget.

Three worked applications

Sell-side, before a buyer exists

[ ] Start the interviews now, since it is the only part that cannot be compressed later.
[ ] Build the four-bucket schedule and circulate internally for challenge.
[ ] Fix quietly what can be fixed: chain of title breaks, missing contractor assignments, unrecorded prior assignments, lapsed renewals.
[ ] Enumerate the know-how, which the business should have done anyway.
[ ] Model the rebrand cost, since the buyer will price it and the seller should know the number first.
[ ] Prepare answers to the nine diligence requests.

Buy-side, in a compressed process

[ ] Ask for the four columns, and treat the schedule's absence as the finding.
[ ] Concentrate on the shared bucket, where the operating dependency on the seller sits.
[ ] Ask what the divested business uses that the seller retains.
[ ] Check software transferability and the contractor assignment position.
[ ] Price the migration from the brand estate inventory.
[ ] Insist on the wrong-pocket clause, which is the cheapest protection against the diligence there was no time to do.

Post-closing, eighteen months in

[ ] Check the wrong-pocket clause, which converts a discovery into an administrative step.
[ ] Check the cross-licence fields, since the right may already be licensed and nobody read the schedule.
[ ] If neither applies, recognise the buyer is negotiating from established dependency, which is why the clause and the schedule matter.
[ ] Run the twelve-month review that should have happened, while the clause is live.
[ ] Run the mirror review on the retained side, which finds the equivalent problems there.

A note on proportion

[ ] A small divestiture with no shared assets needs Phases 1, 2, 7, and 8 only.
[ ] Never defer Phase 1, since it is the critical path and cannot be compressed.
[ ] Never omit the wrong-pocket clause, which costs a paragraph and prevents an entire category of dispute.
[ ] Never leave recordation unowned, since transferred rights lapse in the first cycle and the loss is irreversible.
[ ] Treat Phases 3, 4, 5, 6, and 9 as conditional on the business actually having a shared mark, entangled families, shared systems, transition services, or a rebrand.
[ ] [Gate] Client and adviser have agreed in writing which phases are in scope and why the others are not.

The interview script

Phase 1 works or fails on the quality of the interviews, and a standard script makes them comparable across dozens of conversations.

[ ] "Describe what your function does, in five minutes." Establishes context and surfaces activities the organisation chart does not show.
[ ] "What would stop working on Monday if you lost access to it?" The single most productive question in the whole exercise.
[ ] "What do you use that you did not create?" Surfaces third-party dependencies including software nobody in legal knows about.
[ ] "What did you create that somebody else uses?" Surfaces the shared bucket from the supply side.
[ ] "Where is it kept, and who else can reach it?" Establishes both the location for transfer and the access position for the know-how enumeration.
[ ] "Who wrote it, and were they an employee at the time?" The ownership question, asked in a form somebody can answer.
[ ] "What do you send to suppliers or customers?" Surfaces disclosed specifications and the confidentiality position around them.
[ ] "What is the thing nobody else could replicate?" Surfaces the genuinely valuable know-how, which is rarely the thing the organisation formally documents.
[ ] "What would you take with you if you moved to the other business?" Uncomfortable and revealing, and it identifies the assets that travel with people rather than with paperwork.
[ ] "Who else should I ask?" Closes the loop and finds the people the organisation chart missed.
[ ] [Gate] The same ten questions were asked of every function, and the answers are recorded against named interviewees with dates.

The weekly report to the deal team

The intellectual property workstream is invisible to a deal team unless it reports in terms they act on, and a one-page weekly format does that.

[ ] Interview completion percentage against the planned schedule, which shows whether the critical path is moving.
[ ] Assets identified to date, split across the four buckets, so the shape of the negotiation is visible early.
[ ] The unresolved bucket as a number, since it is a price input and the deal team needs it before the negotiation rather than after.
[ ] Open allocation decisions older than two weeks, escalated by name, since these do not resolve themselves.
[ ] Third-party consents outstanding, with the count and the longest-running, since these have external dependencies and can delay closing.
[ ] The estimated rebrand cost, updated as the estate inventory develops.
[ ] The recordation estimate, so it enters the post-closing budget rather than being discovered afterwards.
[ ] One risk of the week, stated in a sentence, since a report with everything flagged flags nothing.
[ ] [Gate] The deal team has seen the unresolved bucket's size before the price is agreed.

The brand migration milestone set

A migration plan needs milestones somebody can verify, and the same set works across most separations.

[ ] Month zero: the new name is cleared and filed in every market on the footprint, with domains and handles secured.
[ ] Month three: the visual identity is approved and the asset library exists, since every downstream milestone depends on having artwork.
[ ] Month six: customer-facing digital properties migrated — website, application, email domain, and social accounts — with redirects in place.
[ ] Month nine: all new packaging and labelling artwork approved, including any regulatory submissions the change requires.
[ ] Month twelve: all sites re-signed, which is the most visible milestone and the one a monitoring trustee or a counterparty will check first.
[ ] Month fifteen: vehicle liveries and uniforms complete.
[ ] Month eighteen: old-brand inventory depleted or written off, with a stock report evidencing it.
[ ] Month twenty-four: moulds, tooling, and embossed marks replaced on any product still in production.
[ ] Ongoing: third-party listings corrected — directories, mapping services, resellers, and industry databases — reported quarterly.
[ ] Final: a certificate of completion delivered to the licensor, listing what was done and confirming no remaining use.
[ ] [Gate] Each milestone has a named owner, a date, and a financial consequence for slipping, since a plan without those extends indefinitely.

A closing note

Almost nothing in this checklist is doctrinally difficult. The doctrine occupies a handful of provisions: no assignment in gross, quality control to preserve the mark, a signed writing for copyright transfers, and the joint ownership default that makes co-ownership unattractive.

What makes separations hard is that they require an organisation to describe itself accurately, on a deadline, for the first time — and most organisations cannot, because nobody ever needed them to.

Which is why Phase 1 carries the whole exercise, why it cannot be compressed, and why the single most valuable thing an adviser can do is start it before anybody has agreed a signing date. Everything else in this document is what to do with the answers.

Five things to say at the first meeting

[ ] "The register does not know which business owns anything."
[ ] "Everything continuing needs an end date and a consequence."
[ ] "The rebrand is a real number and it belongs in the model."
[ ] "We need a wrong-pocket clause."
[ ] "Recordation is a two-year project starting at closing, and it needs an owner and a budget."
[ ] [Gate] All five were said at the first meeting rather than at signing, where they become a list of things that should have happened.
[ ] And one to hear back: the business should be able to name, without research, the person on each side who owns the post-closing obligations. If it cannot, Phase 10 has not been done.

Key Authorities at a Glance

The trademark constraint is 15 U.S.C. § 1060, prohibiting assignment in gross, with 15 U.S.C. § 1127 making an uncontrolled transitional licence an abandonment risk — the exposure illustrated in Barcamerica International USA Trust v. Tyfield Importers, Inc.. Enforcement after the split runs through 15 U.S.C. § 1114 and 15 U.S.C. § 1125, with cancellation grounds at 15 U.S.C. § 1064.

Patent allocation runs through 35 U.S.C. § 261 and is constrained by the joint ownership default at 35 U.S.C. § 262. Copyright transfers require the signed writing at 17 U.S.C. § 204, with categories at 17 U.S.C. § 101, initial ownership at 17 U.S.C. § 201, and recordation at 17 U.S.C. § 205. Know-how depends on 18 U.S.C. § 1839 and 18 U.S.C. § 1836. Licence survival in insolvency is Mission Product Holdings, Inc. v. Tempnology, LLC.

| Authority | Phase | | --- | --- | | 15 U.S.C. § 1060 | 3 — no assignment in gross | | 15 U.S.C. § 1127 | 3 — uncontrolled licence risks abandonment | | Barcamerica International USA Trust v. Tyfield Importers, Inc. | 3 — the naked licensing exposure | | 15 U.S.C. § 1114 | 9 — enforcement after migration | | 15 U.S.C. § 1125 | 9 — unregistered marks and false association | | 15 U.S.C. § 1064 | 3 — cancellation after a defective assignment | | 35 U.S.C. § 261 | 4, 8 — assignment and recordation | | 35 U.S.C. § 262 | 4 — why joint ownership fails | | 17 U.S.C. § 101 | 5 — work made for hire categories | | 17 U.S.C. § 201 | 5 — initial ownership | | 17 U.S.C. § 204 | 5 — signed writing for transfers | | 17 U.S.C. § 205 | 8 — recordation of copyright transfers | | 18 U.S.C. § 1836 | 5 — trade secret claim over allocated know-how | | 18 U.S.C. § 1839 | 5 — the enumeration precondition | | Mission Product Holdings, Inc. v. Tempnology, LLC | Variants — licence survival in insolvency |

Further reading is collected at carve-out asset schedule, transitional trademark licence terms, cross-licence field definition, wrong pocket clause, and post-closing recordation programme.