IP due diligence should run in the order that things kill deals, not in the order they appear on a checklist.
Chain of title first. A defect there stops everything, and no amount of commercial analysis matters if the asset cannot be transferred.
Then whether it is even alive. Then what encumbers it. Then whether it covers anything anyone does.
The first two stages are free and take hours, which is why running them first is both faster and cheaper.
The order
| # | Stage | Cost | Kills the deal if |
|---|---|---|---|
| 1 | Chain of title | Free | Gap |
| 2 | In-force status | Free | Lapsed |
| 3 | Encumbrances | Free to check records | Unreleased lien, exclusive licence |
| 4 | Scope and relevance | Analysis | Covers nothing anyone does |
| 5 | Validity exposure | Analysis | Challenged or vulnerable |
| 6 | Contracts and obligations | Document review | Restrictive terms |
Stages one and two eliminate the most for the least. That is the argument for the ordering.
Stage 1: chain of title
| Check | Detail |
|---|---|
| Do the inventors assign to the first entity? | The origin |
| Does each assignor match the previous assignee? | Continuity |
| Any gap? | Blocking |
| Any dissolved entity in the chain? | Frequently unfixable |
| Name changes recorded? | Common omission |
| Recorded in every jurisdiction? | Foreign rights are separate |
Three causes of a gap — an unrecorded transfer, an unrecorded name change, or a transfer by a party that did not own the asset. The first two are fixable; the third may not be.
A dissolved entity is the hard case, because the party needed to execute a corrective assignment no longer exists.
Use USPTO Assignment Search, free, and try many name variants. Misses are silent. See assignment search.
Stage 2: in-force status
| Check | Source | Time |
|---|---|---|
| Maintenance fees paid | USPTO Patent Center | 3 min |
| Expiry date | Front page + PTA | 5 min |
| Continuation chain | Related U.S. Application Data | 3 min |
| Terminal disclaimer | Front page | 2 min |
| Foreign members | National registers | Per country |
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
The continuation trap catches buyers constantly. A patent filed in 2022 as a continuation of a 2014 application expires in 2034, not 2042 — eight fewer years than the filing date suggests, and sellers miscalculate it without any dishonesty.
Stage 3: encumbrances
| Encumbrance | Visible in records? |
|---|---|
| Security interests | Yes — check for release |
| Exclusive licences | Frequently not |
| Non-exclusive licences | Frequently not |
| Obligations to assign | No |
| Government funding conditions | Sometimes |
| Litigation and PTAB proceedings | Yes, public |
| Co-ownership | Sometimes |
Exclusive licences are the invisible encumbrance. They restrict what an owner can transfer and appear nowhere, so they must be warranted rather than verified.
Unreleased security interests are visible and common. A lien recorded during a financing and never released stays on the record indefinitely.
Co-ownership is worth checking. Absent an agreement, each joint owner can license the whole invention independently, which changes what you are acquiring.
Stage 4: scope and relevance
| Question | Why |
|---|---|
| What do the claims actually cover? | Not the title or abstract |
| Do they read on products people sell? | The value driver |
| Is there documented evidence? | Turns belief into an asset |
| How much term remains? | Under 3 years is very hard to sell |
| Family depth and jurisdictions | Design-around difficulty |
| Prosecution history | Scope surrendered permanently |
Read the prosecution history. Scope surrendered during examination is barred from recapture through estoppel, and it is free to check from a public file wrapper. See prosecution history estoppel.
The employee and contractor problem
| Party | Ownership without a written assignment |
|---|---|
| Employee, agreement in place | The employer |
| Employee, no written assignment | Potentially the employee |
| Contractor, no written assignment | Frequently the contractor |
| Consultant or agency | Same risk |
| "Agrees to assign" | Obligation only, not a transfer |
This is the most common defect found in diligence. Paying for work does not buy the IP in it.
Present-tense language matters. "Hereby assigns" operates immediately; "agrees to assign" creates a promise requiring a further document. See intellectual property assignment.
By right type
| Patents | Trademarks | Copyright | Trade secrets | |
|---|---|---|---|---|
| Recorded chain | Yes | Yes | Optional | None |
| Renewal risk | Three fees | Use and filings | Minimal | Security |
| Special requirement | — | Goodwill must transfer | Termination rights | Secrecy through diligence |
| Main defect | Chain gaps | Non-use | Ownership | Disclosure |
Trade secrets are the diligence paradox. Showing a buyer the secret is how the secret stops being one if the deal collapses. Staged disclosure under NDA is the answer.
Trademarks assigned without goodwill can be damaged — an assignment in gross is a real risk rather than a formality.
Worked example: two portfolios
| Portfolio A | Portfolio B | |
|---|---|---|
| Patents | 62 | 14 |
| Assignment chains to verify | 51 families | 3 |
| Chain gaps found | 4 | 0 |
| Unreleased security interests | 2 | 0 |
| Lapsed patents in the set | 9 | 0 |
| Term overstated by seller | Yes — continuations | No |
| Evidence of use documented | None | Charts on 3 products |
| Diligence time | Weeks | Days |
Portfolio A's diligence burden was itself a discount. Fifty-one chains to verify before anyone values anything.
The four gaps were blocking. Two were fixable with corrective assignments; one involved a dissolved entity and could not be resolved.
Everything found in stages one and two cost nothing and would have been visible to the seller had they looked.
Warranties to ask for
| Warranty | Position |
|---|---|
| Ownership and authority to transfer | Ask for it |
| No undisclosed encumbrances or licences | Ask for it |
| Fees current at closing | Ask for it |
| All inventors assigned | Ask for it |
| Cooperation on foreign recordation | Needed after closing |
| Validity | Do not expect it |
| Non-infringement | Do not expect it |
Nobody warrants validity. A seller offering to is either inexperienced or pricing something worth examining.
Foreign recordation needs seller signatures after closing, which is why cooperation belongs in the terms rather than in goodwill.
Preparing your own assets
| Step | Cost |
|---|---|
| Verify your own chain of title | Free |
| Confirm all inventors assigned | Free |
| Check security interests released | Free |
| Verify in-force status | Free |
| Calculate term from the earliest parent | Free |
| Assemble file histories and family data | Free |
| Document evidence of use | Time |
Run this before you need it. A gap discovered under deal pressure costs far more than one found and fixed a year earlier. See how to purchase a patent.
IP due diligence: the checklist
- Chain of title first. A gap stops everything.
- Verify status second. 58.6% of US utility patents are abandoned before term.
- Calculate term from the earliest parent, never the filing date shown.
- Check for unreleased security interests. Common and visible.
- Warrant what records cannot show — exclusive licences especially.
- Check every employee and contractor assignment. The most frequent defect.
- Look for present-tense assignment language, not promises to assign.
- Read the prosecution history for surrendered scope.
- Verify against public records, not the seller's representations.
- Run it on your own assets first. Gaps found early are fixable.