Most organizations find out in the numbers anyway, because nothing is instrumented to see it earlier. TL Partners solves that problem. The briefs, method, and field notes below show how.
Running a company: start with the briefs. Backing companies: start here.
Sanitized briefs from real engagements, field notes, and the point of view underneath them. The thinking spans 2021 to 2026; the write-ups were standardized as one library in July 2026.
The method has been in field use since 2025; these scenario briefs are its current state, revised July 2026.
Every company has someone accountable for talent. Almost no company has someone accountable for work. Work is now produced by people, by AI, and by combinations of the two that change quarter to quarter; the org chart tracks the people, and nobody tracks the work. The question that matters: is the work we need done actually getting done? Headcount, engagement, and attrition will not answer it.
The gap between what leadership said matters and where day-to-day work actually goes shows up late, usually once it is in the numbers. By then you are explaining rather than steering. The drift was visible months earlier, in behaviors, and nobody was instrumented to see it. So: alignment is a measurement problem before it is a communication problem, and leadership is trainable once the measurement tells you where. Instrument the work first. Then develop the leaders exactly where the instrument points.
A mid-market software PE firm moving from a deal-only model to an operating model had a playbook for everything except leadership. Over two years, from a standing start: a leadership diligence framework run on four acquisitions pre-close, a 100-day talent plan each new company adopted post-close, and an executive scoring system (intrinsics, outcome metrics, work-flow metrics) tracked as talent density across eight portfolio companies in the US, Israel, and Latin America.
The scoring made moves possible that are invisible without it: proven operators redeployed across the portfolio to the asset that needed them, search briefs written against measured gaps rather than departures, workforce planning run as a value-creation lever with each CHRO, engineering hubs stood up in Latin America at four companies, planned headcount growth cut at two others, and go-to-market changes with CEOs that lifted revenue per rep 15 to 22 percent.
Read the full paper: The Portfolio Talent Operating System (PDF)
The people leader who ran annual cycles beautifully may be the wrong hire for a company whose thesis assumes an AI-driven margin structure, and the assessment playbook has not moved with the job. The brief gives sponsors six criteria for the AI-ready people leader, each with the interview question that exposes it, the tell that separates real from rehearsed, and the reference twin that cannot be coached.
Then the honest limit: every answer is still narration, and the leader who narrates well and the one whose organization executes are different people more often than anyone admits. That is the gap the Executive Assessment closes, in two stated modes, scored against your thesis rather than a generic bar. It ends on one run, sanitized: the search that did not open.
You diligence financials to the second decimal and model the thesis under three downside cases. Then you assess the team that has to deliver all of it with a few partner dinners and a reference call, and when the numbers wobble you reach for the two tools you have: replace an executive or reorganize. Those two moves treat different diseases, and almost nothing in the standard toolkit tells you which disease you have.
Replace the CRO when the real constraint is a broken handoff and you burn twelve months onboarding a successor into the same broken system. Reorganize when the real constraint is one seat and you put the whole company through change management to avoid one hard conversation. Either mistake costs a year, and most holds only have five or six of those. The paper names three misreads and the two instruments that answer them.
Enablement budgets buy courses, because a skills gap implies a course and a course can be procured. In the rooms we run, skill is the smallest of three barriers. A support rep with forty tickets a day has no slack to experiment, and the professionals hired to be right, in finance, legal, and operations, learned long ago that being wrong feels like failing. The thing that makes you great with AI, the willingness to experiment and get it wrong, is the opposite of what made you great at your job.
The paper names the five mistakes behind most stalled AI programs, none of them a technology problem, and the sequence that fixes them: anchor AI to the strategy you already have, leaders first, teams doing real work, evidence instead of activity. It ends with the five numbers an AI program can defend to a board.
The full library, every paper and brief → Including the research franchise, Mastering AI for HR, second edition, and the complete set of scenario briefs.
Every value-creation plan reduces to three questions: WHAT is the value hypothesis? WHO can carry it? HOW is the work actually moving toward it?
The WHAT is yours; we never set it. Our instruments answer the other two, scored against your thesis rather than a generic model, which is what lets you tell a people problem from a system problem before you fix the wrong one.
The thesis: where you have decided this company is going and why it will be worth more when it gets there. We never set it. Everything we measure is scored against it, which is what makes the instruments yours rather than generic.
Intrinsics, outcome metrics by function, and a read of how work flows through each leader's organization, rolled into one executive quality score and tracked over time as bench density. Run pre-close on acquisitions and across the hold. Every input is invited: structured sessions and opt-in signals, never passive collection.
A five-dimension pulse, a system map of the likely constraint, what drift costs monthly, and a 30/60/90 roadmap. Run at day 60 to 90 post-close and at every inflection, one company or comparably across a portfolio. And when WHO and HOW are both broken, the pairing sequences the fix: repair the system enough to see the person, then judge the person against the thesis.
Two weeks, fixed fee, one to two senior vantage points. The job: tell you whether alignment is your constraint at all, what the drift costs monthly, and whether deeper validation would change any decision you would make. Sometimes the answer is no, and the brief says so. Buy this when you suspect a problem and want the cheapest honest read before committing anyone's calendar.
Six to ten executive interviews, perception analysis across functions, and a leadership calibration session. The job: replace one vantage point with shared reality, name the constraint at Strong Signal confidence, and hand the team operating recommendations they have already argued through in the room. Buy this when the downstream decision is expensive: a reorg, an executive change, a reset you cannot afford to aim wrong.
Most client relationships begin with the two-week read. Some end there, on purpose. The ones that continue do it phase by phase, each with its own stop point, and the phases tend to add up to 18 to 24 months. Nobody commits to that on day one; it accumulates one good decision at a time.
Self-reported scores usually overstate readiness. The qualitative evidence underneath them is where the constraint shows, which is why the pulse is the start of the audit and never the conclusion.
The instruments are half the method. Leadership is trainable, like any expert skill, so once the reading says where, the work moves into the room: executive team offsites built around the specific misalignments the audit named, executive coaching against the assessment, and development programs for the managers who carry strategy to teams. Diagnosis without that follow-through is a report. Follow-through without diagnosis is a guess. The method is both, in that order.
Anatomy of an audit: day 0 to day 90 (PDF) What working with us is actually like, walked through a composite engagement, including the finding the CEO did not want and the moment we say stop.
Where the method turns into engagements. One thesis underneath: read the work, then develop the leaders where the reading points.
The largest share of our work, and the part where clients renew: executive offsites designed around the specific misalignments the audit found rather than a facilitation template, executive team coaching against the assessment, succession maps scored on the same instrument, and development programs for the managers who carry strategy to teams. Everything is bought in short phases with exits built in; clients keep choosing the next phase because the room work compounds. Recent example: re-anchoring leadership spans on measured decision load, not headcount math, after a reduction.
AI resets change who does what faster than org charts follow. We start with the work inventory that finds the 15 to 20 percent of surviving work with no remaining consumer, publish the agent boundary per function, then build the assessments and upskilling behind the new model. Across two enterprise AI-enablement programs in 2025 and 2026, more than 80% of participating managers reached target proficiency.
The audit is a snapshot. Alyn is the product we build so the reading never stops: it sits in Slack, reads where on-the-ground work is drifting from executive-stated priorities, and surfaces it while course correction is still cheap. alyn.app
Our founder built the portfolio talent function at Turn/River Capital, a software-focused private equity firm, from a standing start: leadership diligence run on four acquisitions before close, a 100-day people plan each new company adopted, and executive scoring across eight portfolio companies. Two offerings come out of that system: pre-close talent diligence that tests a management team against the investment thesis before the check is written, and portfolio alignment audits, the same audit run comparably across holdings, newly offered to funds.
The one-page version of this argument (PDF), as presented at fund partner meetings. We bring fifteen minutes of material; you bring the questions.
A note on names: our private equity clients and their portfolio companies are never named, here or anywhere. Publicly traded clients we have served include Google, Workday, Sony, and Upwork. The portfolio talent operating system in Thinking describes our founder's own instruments, built in-house and since generalized; no fund's deals, companies, or economics appear anywhere on this site.
TL Partners is a partnership of senior operators, globally distributed, working primarily with PE-backed and late-stage software companies from growth stage to multi-billion-dollar public. The firm is built on one question: does the work an organization produces match the strategy its leaders set? Everything on this site, the instruments, the briefs, the papers, is that question made operational.
The method has a provenance. Founder and Principal Thomas Igeme built the portfolio talent function at Turn/River Capital, a San Francisco software private equity firm, from a standing start. That work is the provenance of the instruments in Method, since generalized across advisory clients. Before that: the Boston Consulting Group; LinkedIn, where he built the LinkedIn Way, its first global sales methodology, and its sales-enablement product ecosystem; and Trybe.ai, the startup he co-founded before a shift to ServiceNow, where he led global scaled learning through hypergrowth. He has led TL Partners since 2023.
Stanford University, B.S. Management Science and Engineering. Instructor, Peak Performance Leadership, Stanford Continuing Studies. Founding advisor, PeopleTech Partners. Board member, Envision Education, FACES SF, and the Faith and Justice Network. Host, Venture Visionaries. LinkedIn
If the thinking here sounds like a moment your company is in, 25 minutes is the right amount of time to find out. Nothing to prep. And if you back companies rather than run one: we present this work at fund partner meetings, fifteen minutes of material, your questions after.
Tell us where to send it and the download starts right away. We send the occasional new brief, nothing else, and we never share your details.