Law firms · Intellectual property partners
IP lateral partner recruiter
For the managing partner or IP chair underwriting one lateral. Who hires, what book shows up in the published data, how a guarantee should be sized, and which covenants can hold the start date.
IP is a different underwriting from a general lateral.
An IP lateral partner recruiter works for the firm that will pay the draw: the managing partner, the chair, or the head of IP. The question is whether a specific set of portfolios and client relationships will be at the new firm when the guarantee period ends.
The firm-wide method is written once. Brief, portability, the Lateral Partner Questionnaire, conflicts, and integration are on lateral partner recruiting. Repeating those five phases here would not change an IP decision. What changes the decision is the practice: a book that can be recurring prosecution or one-off litigation, panels that re-appoint counsel, technology-field conflicts, and open filings with deadlines that do not wait for a notice period.
The book figures below are the IP rows of Sartori & Partners’ published interview corpus of 2,600+ partner conversations, set out in partner book of business by practice. They are what partners reported in lateral conversations, not audited collections. The corpus skews to partners willing to discuss a move. It is not a census of every IP partner in the Am Law. Market facts from Pirical and Above the Law are labeled with the year in the same sentence.
The 2026 hiring read for this practice sits in legal hiring in 2026, read next to hours billed, not instead of them.
Five buyers. One of them is your brief.
IP laterals are opened by the shape of the firm, not by a generic partner requisition. Name the buyer before you name the partner.
- i.
IP boutiques
The practice is the firm. The hire is a technology field the bench does not cover, a litigation partner on a prosecution-heavy platform, or the reverse. The book is often a set of patent portfolios and the technical team that prosecutes them. Treat a solo lateral and a lift-out of a prosecution group as different underwritings.
- ii.
Full-service firms
IP is the service line that holds a technology or life-sciences client. The chair hires when corporate is losing pitches for lack of a patent litigator, when a licensing practice sits behind a transactional team, or when a conflict bars the sitting group from a field. A standalone IP collections figure understates the hire if the point is the client relationship the IP partner protects.
- iii.
Patent litigation platforms
A trial partner and a prosecution partner are two hires with one label. Litigation books are matter-driven and end with the case. Prosecution and portfolio books recur. The published IP range, $0.05 million to $20 million, is wide because those books share a practice name and not an economic shape.
- iv.
Trademark, brand and media practices
Enforcement, clearance and portfolio management for consumer and media clients are a different brief from patents. They seldom need a technical degree. They do need a written list of which brands a partner personally manages and which sit on a corporate panel.
- v.
European and London offices
The published EUR sub-sample for IP is small (n=5, median about EUR 1.75 million, range EUR 0.4 million to EUR 4 million), and the source page shows it only because the cell clears the n of five floor. Read it as direction, not as a benchmark. A European IP partner is underwritten on the local court and patent-office calendar, not on a US litigation résumé.
Partner moves by city, all practices, are counted on lateral partner hiring statistics for 2026. Pirical’s Q1 2026 Am Law 200 partner-title count, published 1 May 2026, put litigation at 388 joins, corporate at 217, and banking and finance at 136. IP was not in that published top three. Do not staff an IP chair off the litigation column.
The published IP number is about $2 million. The range is the decision.
A practice head who underwrites to the median and ignores the range will guarantee the wrong partner. These are reported books, banded, from one interview corpus.
- ~$2M
- median USD book reported by IP partners in the published interview corpus (n=52). Observed range: $0.05 million to $20 million.
- Sartori & Partners interview corpus, published on Partner Book of Business by Practice
- 79%
- of IP lateral partners hired by Am Law 101–200 firms in 2020–2022 were still there three years later. Cross-practice average on the same cut: 84%.
- Pirical, 30 April 2026
- 25 / 42 / 26
- percent of 84 assessed IP conversations rated highly portable, partially portable and low portability. A further 7 percent were institutional.
- Sartori & Partners interview corpus, published portability table
- 30–38%
- of lateral partners, across practices, leave the hiring firm within five years. Not an IP-only rate.
- Decipher Investigative Intelligence, via Above the Law, 6 May 2026
IP USD book data: n=52, median about $2 million, observed range $0.05 million to $20 million. Corporate and M&A in the same table: n=76, median about $2.15 million. Currencies are not cross-converted. Small cells under five conversations are not shown on the source page. Full table: partner book of business by practice. The 79 percent retention figure is a three-year look at Am Law 101–200 hires made from January 2020 through December 2022 (Pirical, 30 April 2026). The 30–38 percent leave rate is a five-year, all-practice figure (Above the Law, 6 May 2026, citing Decipher). They are not two readings of one cohort.
Above the Law’s 6 May 2026 account of a survey of managing partners and business-development leads, and of Decipher Investigative Intelligence, adds two operational points. Ninety percent of firms in that research report business-development challenges with laterals. The same account says a failed lateral can cost 200 to 400 percent of the lawyer’s annual compensation once recruiter fees, guarantees, and replacement are counted. Neither point is IP-specific. They are the failure math an IP guarantee still has to survive.
ABA Law Practice Magazine, July–August 2025, reports that some firms now require a lateral to show a $12 million book before they will open a conversation, and that capital contributions can exceed $1 million. Applied to IP, a $12 million screen excludes nearly the whole published range: the observed ceiling is $20 million and the median is $2 million. If the strategy is an IP chair, the screen has to be rewritten. If the strategy is an eight-figure originator who happens to hold patents, say that, and run it as a corporate or litigation lateral.
Four cuts before the number is allowed into a guarantee.
IP portability is a client-by-client and docket-by-docket question. The published mix says the practice does not sort itself into one answer.
Prosecution versus litigation
Separate the recurring portfolio work from the episodic case. A patent suit ends. A prosecution docket renews. The guarantee rests on the slice that recurs, and the two slices rarely belong to the same partner.
Client or docket
IP files can be tied to a docketing relationship and a panel, not to the lawyer. The published mix (25 percent highly portable, 42 percent partial, 26 percent low, 7 percent institutional) says the practice does not sort itself into one answer. Test each of the five largest clients.
Rate-card gap
Prosecution clients are often price-sensitive and run panels. A move onto a firm whose rates sit well above the client’s current panel can zero a relationship that is otherwise personal. The published book checklist stress-tests each client against a 15–25 percent rate uplift before the guarantee is drafted.
Team dependency
A patent partner who cannot practice without technical specialists and a prosecution team is a lift-out. Price the group, or do not price the partner as if the book moves alone. The firm-wide rule for mapping a group before anyone is approached is on the lateral partner recruiting page.
The 15–25 percent rate-uplift test is the checklist published on the book-of-business page, not a survey of what clients accept. Use it as a question to each client, then keep the answer. The firm-wide lateral page carries the cross-practice client-stay finding. Do not paste it onto an IP panel without reading the five largest files. The IP portability row is the practice-specific figure: 25 percent highly portable, 42 percent partial, 26 percent low, 7 percent institutional, on 84 assessed conversations. The published page reads that spread as the institutional nature of large-firm IP mandates.
Size the floor to the recurring slice. Then say what the firm is buying if that slice is small.
A guarantee is a promise to pay if collections miss. On a median reported book near $2 million with a range that starts at $0.05 million, the composition matters more than the headline.
- i.
What can sit under the floor
Originated, recurring prosecution and portfolio work that survives a rate-card test. That is the defensible base. Worked-not-originated revenue is the old firm’s client. A litigation fee ends with the case. Putting either of the last two under a multi-year floor is how IP guarantees miss.
- ii.
What the firm may be buying instead
A full-service firm sometimes needs the IP name so technology pitches stop failing. That is a strategic-gap hire, which the July–August 2025 ABA Law Practice article distinguishes from an opportunistic revenue hire. Pay for the gap in the offer narrative. Do not pretend it is IP originations that will hit the drawer in year one.
- iii.
Retention of the partner is a different series
Pirical’s 30 April 2026 ranking found intellectual property at 79 percent three-year retention among Am Law 101–200 laterals hired in 2020–2022, against an 84 percent average on that cut. Labor and employment was 86 percent and private client 87 percent. The ranking excludes firms with fewer than five lateral partner hires and excludes merger moves. It counts whether the person was still at the firm. It does not count whether the originations arrived. Decipher’s five-year leave rate of 30–38 percent, reported by Above the Law on 6 May 2026, is the longer clock, and it is not limited to IP.
- iv.
How the search is paid
IP partner mandates are retained. The fee is 25–30 percent of the hire’s total first-year compensation. Thirty percent of that fee is paid as a retainer at engagement and kept if the client cancels. Contingent search does not fund the client-by-client read this practice requires. Terms are in the letter before anyone is called.
The Lateral Partner Questionnaire is the instrument, and it is described on the firm-wide page. For IP, the columns that change the floor are realization and the matter list, not the headline origination. Prosecution work is often billed at panel rates below the firm’s rack rate, and litigation is written down when a case settles early. A three-year origination number that ignores realization will overstate what the new firm can collect. Ask for hours, rates, and collections on the ten largest matters or portfolios, then mark which are open, which are annual programs, and which ended. The guarantee uses the open and recurring lines only.
A group move is priced only after that read. If technical specialists and a prosecution team carry the docket, the partner’s reported book is the group’s book. Hiring the partner and hoping the team follows is how a reported median becomes a much smaller collection in year one. Either the offer includes the people the work needs, or the floor is cut to what the partner can staff alone. That choice is made before the compensation committee sees a number.
Four screens that decide whether the practice can start.
Discipline, book, client, and deed. In that order. A partner who clears the economics and fails the deed is not a hire with a delayed start. The start may be the problem.
- I Screen 01 — The discipline
Patent prosecution, patent litigation, trademark, or licensing. Pick one before you map.
IP is not one market. A patent litigator, a prosecutor with a technical background, a trademark and brand partner, and a licensing or technology-transactions partner hire from different benches and clear different conflicts. A résumé that says “IP” without the discipline is not a brief.
- State the technology field in the first conversation. Semiconductors and biologics do not share a bench, and a prosecution conflict follows the field.
- Technical credentials and patent-bar registration change who can do prosecution work. Say whether the seat needs them.
- Licensing and technology transactions often sit next to IP and are originated differently. Say whether the seat includes them.
- II Screen 02 — Composition of the book
Read the $2 million median as a mix, not a floor.
In the published corpus of 2,600+ partner conversations, IP partners with USD book data (n=52) reported a median near $2 million and a range of $0.05 million to $20 million. Corporate and M&A partners in the same corpus, the cross-practice anchor, reported a median near $2.15 million (n=76). Employment sits near $1.2 million (n=13). The IP median is close to the anchor, but the range under it is far wider than the median suggests.
- Ask for three to five years of originations, collections, realization, hours, and rates. That economic section is the Lateral Partner Questionnaire, described on the firm-wide page.
- Split the number into recurring prosecution and portfolio work, worked-not-originated, and litigation before anyone discusses a floor.
- A $12 million book screen, which some firms now apply before they will talk to a lateral, admits almost no one from the published IP range. (ABA Law Practice Magazine, July–August 2025.)
- III Screen 03 — Who the client will follow
The client is often a portfolio on a panel, not a relationship on a desk.
IP revenue often sits in an appointed panel or an inherited docket reviewed every year or two. The general counsel or the head of IP can move the file without moving the relationship the partner thinks they own. Of 84 assessed IP conversations in the published corpus, 25 percent were highly portable, 42 percent partially portable, 26 percent low portability and 7 percent institutional. The four shares sum to 100 percent.
- Test each of the five largest clients: loyal to the lawyer, to the platform’s technical coverage, or to a fee the new firm will not match.
- A portfolio staffed by the old firm’s prosecution team is low portability even when the partner owns the client call.
- Cross-sold transactional or litigation work that never appears on the IP partner’s origination report will not show up in the guarantee, and it may be the reason the full-service firm is hiring.
- IV Screen 04 — Covenants and the start date
A US non-compete usually fails. The notice clause and the file transfer often do not.
ABA Model Rule 5.6(a) provides that a lawyer shall not participate in offering or making a partnership, shareholders, operating, employment, or similar agreement that restricts the right to practice after the relationship ends, except an agreement concerning benefits upon retirement. The comment states the purpose: lawyer autonomy, and the client’s freedom to choose counsel. What delays an IP start date is usually notice, the client’s instruction on who holds each open filing, and a forfeiture clause dressed up as a retirement benefit.
- ABA Formal Opinion 06-444 (13 September 2006) reads the retirement exception narrowly: the benefits must be available only to a lawyer who is actually retiring, and the clause cannot forfeit income already earned.
- Open filings have deadlines. A transfer of a pending application or an instituted proceeding is the client’s decision and a docketing task, and it has to be sequenced before a start date, not after.
- California is the outlier. Rule 5.6(a) opens with “Unless authorized by law,” and Comment [1] points to Business and Professions Code section 16602 and Howard v. Babcock (1993) 6 Cal.4th 409. Do not export a California deed to New York.
What still binds, even where a non-compete does not: a notice period that is a duty to transition rather than a ban on practice afterward; the client’s right to choose, which cuts against both a firm lock-in and a partner who solicits while still inside the firm; and a staff non-solicit, which is a different clause from a ban on practice and has to be read under the governing state’s law. We do not treat “non-solicit” as a synonym for “unenforceable,” and we do not treat it as automatically valid. A London deed is a different regime. Garden leave and restrictive covenants in a UK partnership agreement are not Model Rule 5.6. Read the deed. Do not import a New York conclusion into it. During a US notice period the partner is still at the old firm. Client contact is a fiduciary question, not a recruiting one: who may speak to the client, who sends the choice-of-counsel letter, and which matters are mid-deadline or mid-proceeding and cannot move cleanly. An IP practice feels this because filings and proceedings run on dates the client does not control. Sequence the resignation, the client letter, and the start date so the new firm is not practicing the old firm’s files. The firm-wide integration sequence is on the lateral partner recruiting page. This page only adds the IP constraint: open applications, open proceedings, and docket transfers. Sources: ABA Model Rule 5.6, Formal Opinion 06-444, California Rule 5.6.
A partner-move count is not an IP hiring plan.
Partner-title joins in the first quarter of 2026 were led by other practices. A chair who reads that column as IP demand will open a search the market is not filling, or refuse a search the client base justifies.
Pirical, 1 May 2026, counted Am Law 200 partner joins for 1 January through 31 March 2026 and published litigation (388), corporate (217), and banking and finance (136) as the top three practice labels. IP was not among them. Use the 2026 lateral statistics note before comparing that quarter with a full-year, all-title series. Mixing a partner-title quarter with an all-title year produces a hiring plan that cannot be audited.
The retention series says something different. Pirical’s 30 April 2026 ranking puts IP at 79 percent three-year retention, five points under the 84 percent cross-practice average. That is a prompt about integration and docket transfer, not a reason to avoid the practice. If the gap is a missing patent litigator in a field where technology clients are asking, the brief supports opening the search. If the gap is “the market is hot, hire someone,” the Q1 partner-title count does not support it. The practice-heat read is on legal hiring in 2026.
Partner joins: Pirical, Q1 2026. Retention by practice: Pirical, 30 April 2026.
Retained, blind, and sequenced around the deed.
The five underwriting phases are not rewritten here. IP adds the screens above, then uses the same retained process.
Blind until you say otherwise
The firm’s name and the partner’s name stay down until both sides have a reason to proceed. An IP group is small enough that a leaked chair search is the search.
Evidence before the floor
Targeting uses the market map. The guarantee uses the LPQ economics and the four cuts on this page. Assertion is not a book.
Deed before the start date
Rule 5.6, the retirement-benefit exception, California’s separate line, and any UK garden-leave clause are read before a start date is offered to the partnership.
Integration is the year that counts
Pirical’s 79 percent is a three-year headcount result. The first year is when panel appointments and docket transfers are either completed in the new firm’s name or quietly left behind. We stay through that cycle.
Fee 25–30 percent of total first-year compensation. Retainer 30 percent of the fee at engagement. The replacement and off-limits terms are stated in the engagement letter, not after the first approach. Partners weighing a move of their own, rather than hiring one, should use the partner advisory.
The pages that carry the method, the book, and the 2026 market.
The firm-wide process and the full book table live on the pages below. The IP decision is the buyer, the book, the guarantee, and the deed.
Lateral partner recruiting
The firm-wide mandate: five underwriting phases, the LPQ, conflicts, and integration. IP does not get a second copy of that process.
Firm-wide lateral methodBook of business by practice
The IP row in full: USD median about $2 million, n=52, range $0.05 million to $20 million, and the portability mix on n=84.
Read the IP rowIs the 2026 market hiring
The 2026 practice-heat note: where the work is going and where law-firm hiring is softer.
Read the 2026 hiring noteLateral partner hiring statistics
Partner moves by city and practice in 2026, dated and sourced, for reading an IP search against the wider market.
Read the 2026 statisticsAssociate recruiting
The counsel and associate bench under an IP partner. A lift-out is priced here only after the partner economics are real.
Associate recruitingFor law firms
How a firm briefs a retained search, across partners, associates, and the leadership that runs the practice.
For law firmsWhat IP chairs ask before they commit
Our equity screen is a large number. IP partners do not clear it.
Drop the corporate floor or mis-hire. The published IP median is about $2 million (n=52 USD book reports), against about $2.15 million for corporate and M&A (n=76) in the same corpus. The range underneath, $0.05 million to $20 million, is what empties a slate written for a deal practice. We underwrite the portable slice of the IP book, not a firm-wide minimum.
The book is a docket. How much actually moves?
Split recurring from episodic before the floor. Of 84 assessed IP conversations in the published corpus, 25 percent were highly portable, 42 percent partial, 26 percent low and 7 percent institutional. We read the five largest clients against that split: portfolios that follow a person, and dockets that stay with the old bench.
What guarantee is defensible on an IP book?
The floor tracks evidenced collections. We do not publish a practice-wide guarantee percentage. The floor is the originated, recurring slice, tested for rate-card fit, for a defined ramp. A guarantee copied from a corporate template overpays a thinner book. Across practices, 30–38 percent of laterals leave within five years (Decipher, via Above the Law, 6 May 2026). Retention of the person is not retention of the originations.
IP laterals have the weakest retention we see.
Read the retention number for what it counts. Pirical put IP three-year retention at 79 percent for Am Law 101–200 hires made in 2020–2022, against 84 percent across practices (30 April 2026). It counts whether the person was still at the firm, not whether the originations arrived. We use it as a prompt to test integration and docket transfer, not as a verdict on the practice.
The partnership deed has a non-compete.
Read Rule 5.6 before you pick a start date. In most US jurisdictions a post-departure ban on practice is barred by Model Rule 5.6(a), with a narrow retirement-benefit exception (ABA Formal Opinion 06-444). Notice, client notice, and a real retirement forfeiture are different instruments. California’s Howard v. Babcock line is not the national rule. The cross-practice covenant sequence sits on lateral partner recruiting.
We need a patent litigator, not another prosecutor.
The discipline is the brief. Trial work, prosecution, trademark and licensing are not one practice with four headers. We map the bench for the discipline you name, and we say so before outreach. Mixing them produces a shortlist that cannot do the work the chair described.
Will the corporate client follow, or only the IP file?
Credit the relationship the firm is actually buying. On a full-service platform the IP partner is often hired so a technology client stays. That value does not appear as IP originations. We write it into the brief as a cross-sell condition, and we do not let it inflate the IP guarantee.
A prosecution history could conflict us out of a field.
Conflicts by technology field, before names move. We run a general-description conflicts pass against the matters that would actually bar the hire: a field where the partner prosecutes for a competitor of a client you hold, or a proceeding adverse to a client on your panel. The pass happens before the market knows you are looking.
We are replacing our IP chair. The market cannot know.
Blind until you release the name. A chair search in a small IP group leaks fast, because the bench is visible. The mandate stays confidential. Your firm’s name is disclosed only to screened candidates, under your control. Partners considering their own move use the partner advisory, which is a different conversation from this one.
A brief, not a commitment
An IP lateral is a book problem with a deed attached.
Brief the discipline, the technology field, and the gap. We map the IP field, read the reported book against the published range, and clear the covenant before a start date is put in front of the partnership. Retained.