# McKinsey "AI is everywhere. The agentic organization isn't — yet" → Umbra Studio (Foundation entry) **Source:** Alexis Krivkovich (Senior Partner, Bay Area; co-leader of People & Organizational Performance Practice) interviewed by Lucia Rahilly (Global Editorial Director). Edited transcript from *The McKinsey Podcast*, McKinsey & Company, April 2026. **Why this is entry #00, not #04.** Studio was conceptualized on April 16, 2026. This podcast — published earlier that month — is the paper Abe read that triggered Studio's existence. Numerically it predates the canon. Conceptually, it's the originating pulse: the framing, the vocabulary, and several of Studio's own product decisions trace directly back to Krivkovich's choice of words here. Treating it as #00 makes the canon honest: *we read this, we built Studio, then we absorbed everything that came after as validation*. --- ## Three discoveries that change how Studio reads its own positioning ### 1. The phrase *"lighthouse use cases"* is McKinsey's, not ours Krivkovich, on the kind of leader needed to make agentic real: > *"…leaders who can pick areas ripe for being lighthouse use cases."* — Krivkovich, p. 5 This is where the **Lighthouse Sprint** name lands. Studio's central product nomenclature is rooted in the foundational paper that triggered Studio. That's not coincidence — it's a direct lineage Studio should acknowledge in the deck and the framework intro, not pretend to have coined. **Action.** Add a short paragraph to `usl-framework.html` and the deck origin slide attributing the *lighthouse* metaphor to Krivkovich's framing. Honest provenance is a credibility asset, not a debt. ### 2. The *gen AI paradox* originates here, not in Marketing Workflow The Marketing Workflow (#02) line that became Studio's hero hook — *"the technology can be found everywhere except on the bottom line"* — is a downstream restatement. The original is here, plainer: > *"More than 80 percent of companies say they're not yet seeing impact on the bottom line…"* — Krivkovich, p. 2 This means stat #7 (90% experimenting / <10% capturing) and stat #14 (>80% seeing no bottom-line impact) are **the same finding observed at different altitudes** — the marketing-functional version vs. the cross-functional version. They should never appear in the same email/slide; they're not corroborating data points, they're the same data point. ### 3. *"Humans above the loop"* is new vocabulary not in #01/#02/#03 Krivkovich draws an explicit distinction: > *"Having a human above the loop suggests… the human's role becomes judgment on top."* — Krivkovich, p. 7 *Humans in the loop* = agents do parts of the process, then hand off to a human for the rest. *Humans above the loop* = agents do most or all of the core process; the human applies judgment after. This is a meaningful framing upgrade. Studio's Operations Manual currently uses "human in the loop" as a single concept; reality is two states with different governance, eval, and skills implications. Above-the-loop work demands *different* skills (oversight, pattern-recognition assessment, executive function review) than in-the-loop work (handoff fluency, prompt engineering, mid-process correction). **Action.** Operations Manual should adopt the in-the-loop / above-the-loop distinction and route the Team Skills Roadmap (#02 gap) accordingly. A reviewer of arbitration-style end-to-end agentic outputs needs different training than a copy-editor handing off draft variants. --- ## The five pillars of the agentic organization Krivkovich references *"five pillars that make up the new agentic organization"* but the transcript walks through them rather than enumerating with names. From the conversation, the five appear to be: 1. **Business model** — what near-zero marginal cost of delivery / hypersegmentation to a unit of one / customer-side agents do to your moat. 2. **Operating model & teams** — daily workflows, rituals, governance, the "hours of the day spent differently" question. 3. **Workflows (end-to-end reimagining)** — point solutions vs. whole-process redesign. Insurance underwriting, hire-to-onboard, arbitration case review as examples. 4. **Leadership** — how executives show up differently, the trust gap, the *"radically transform ourselves first"* line, the perpetual-state framing of change management. 5. **Talent & culture** — 75% of roles needing reshaping, the junior-talent / hazing-grunt-work problem, two-way doors, fluid talent flows, learning at the center rather than as a sidecar. The five pillars map cleanly to the six *Rewired* capabilities cited in entry #03 (strategic road mapping is implicit here; tech and data sit inside #2). The Manifesto's 12 themes are essentially the same pillars *enumerated*. So #00 introduces the conceptual scaffolding that #03 systematizes. --- ## Path B validation — the foundational paper agrees The strategic question from entry #03 — *how big a transformation does Studio claim to enable?* — was locked as **Path B (workflow that unlocks)** on 2026-05-03. The foundation paper independently corroborates this framing: - Krivkovich on what to redesign: *"workflow can be reimagined in its entirety… instead of point solutions"* (p. 4) — the workflow is the unit, not the agent. - On reuse: *"How do I make them reusable… deploy them in multiple places?"* (p. 5) — direct validation of the `@umbra/*` library thesis. - On who picks the target: *"leaders who can pick areas ripe for being lighthouse use cases"* (p. 5) — the workflow choice itself is the strategic act, which is what Studio's Discovery is for. - On the magnitude of the unlock: leadership transformation, talent reshaping, perpetual-state change management — *not* a Studio scope. Studio ships the workflow; the broader transformation is the client's. Path B exactly. This is the most important sanity check. Path B isn't borrowed McKinsey framing — it's the framing of the paper that made Studio exist. --- ## Validation (no gap — themes Studio already operates on) | Krivkovich's framing | Studio mapping | |---|---| | Workflows reimagined end-to-end > point solutions | Lighthouse Sprint thesis, verbatim | | Lighthouse use cases | Studio's Discovery deliverable selection logic | | Reusable agents across deployments | `@umbra/*` library + "third workflow cheaper than first" pricing | | Humans above the loop | Operations Manual review surfaces (needs distinction added — see Gap A) | | Trust gap, slop, hallucinations | Eval Workstream queued from #01 + Trust Deployment Gate from #03 | | Two-way doors, perpetual change | Governance Wrapper + Handoff scaffolding (Skills Roadmap from #02) | | 75% of roles need reshaping | Validates the relevance of the Team Skills Roadmap deliverable from #02 | --- ## Real gaps surfaced by #00 alone ### Gap A — Humans-in-the-loop vs above-the-loop is one concept in our framework, two in practice Studio's Operations Manual treats human oversight as a single concept ("human in the loop"). Krivkovich's distinction is operational: in-the-loop work and above-the-loop work require different skills, different review surfaces, different eval shapes, different governance. **Action.** Operations Manual gets a *Loop Position* section. For every redesigned step, classify the human role as in-the-loop or above-the-loop. The Team Skills Roadmap (queued from #02) then assigns the right skill emphasis per role. ### Gap B — Studio has no answer for the junior-talent / hazing-grunt problem Krivkovich calls this *"the billion-dollar question"* — if you eliminate the work that trained the next generation, you have a senior-only org with no pipeline ten years out. Studio's current Operations Manual does not address how the client maintains a junior-talent development pathway after we ship the agentic workflow. This isn't a P0 — clients in our $50M–$2B ICP usually don't have armies of junior staff to begin with. But for any engagement that touches a function with a deep junior bench (e.g., a marketing org with an in-house creative pool), Studio should have a pre-canned answer about preserving the apprenticeship pipeline. **Action.** Add an *Apprenticeship Preservation* note to the Operations Manual template. One paragraph: how the redesigned workflow keeps junior involvement in non-automatable judgment moments, even when the agentic system could technically run end-to-end. ### Gap C — *"Change management is no longer episodic — it's a perpetual state"* is not in any Studio handoff artifact Krivkovich's perpetual-state framing (p. 9) is the right tone for Handoff conversations but doesn't appear anywhere in our deliverables. The Operations Manual reads like a one-time setup document; reality is a workstream the client team continues forever. **Action.** Add a closing paragraph to the Operations Manual template framing post-Handoff operation as a perpetual workstream, not a project closeout. Ties to the *conviction-journey* candidate pattern from #03. --- ## New stats unlocked (two additions to the canon) | # | Stat | What it claims | PDF | |---|---|---|---| | 14 | **>80%** | Of companies say they're not yet seeing bottom-line impact from AI investments. | p. 2 | | 15 | **75%** | Of roles need fundamental reshaping right now. | p. 4 | These belong on the cross-functional / CEO altitude — both apply across business functions, not just marketing. Stat #14 is the *origin* of the gen-AI paradox framing; stat #15 is a sharper version of the role-reshaping argument that #03 covers as the 30–70 talent shifts. **Important rule (added to stats-sheet.md):** stat #14 (>80% no bottom-line impact, McKinsey podcast Apr 2026) and stat #7 (<10%/90% — McKinsey Marketing Workflow Apr 2026) are the *same finding* expressed at different altitudes. Never use both in the same surface; pick one based on audience function. --- ## Quote bank — for origin slide and CEO outreach Each is short (≤15 words) and works as standalone. | Use it for | Quote | Source | |---|---|---| | **Origin slide / framework intro** | *"Leaders who can pick areas ripe for being lighthouse use cases."* | Krivkovich, p. 5 | | The paradox (CEO/board) | *"More than 80 percent of companies say they're not yet seeing impact on the bottom line."* | Krivkovich, p. 2 | | Why workflow, not tool | *"Instead of point solutions… we're now talking about a whole workflow."* | Krivkovich, p. 4 | | The reuse argument | *"How do I make them reusable… deploy them in multiple places?"* | Krivkovich, p. 5 | | Above-the-loop framing | *"The human's role becomes judgment on top."* | Krivkovich, p. 7 | | Magnitude of role change | *"Seventy-five percent of roles need fundamental reshaping right now."* | Krivkovich, p. 4 | | Perpetual-state change | *"Change management is no longer an episodic thing. It's a perpetual state."* | Krivkovich, p. 9 | | Talent pipeline closing line | *"You may be replaced by someone who embraces AI before you do."* | Krivkovich, p. 6 | --- ## Asset update queue — what changes downstream | Priority | Asset | Change | |---|---|---| | P0 | **`usl-framework.html` + Studio deck origin slide** | Attribute "lighthouse use cases" phrase to Krivkovich; honest provenance. | | P0 | **Operations Manual** | Loop-position section (in vs above the loop) + perpetual-state closing paragraph. | | P0 | **`stats-sheet.md`** | Stats #14 + #15 added; the "same-finding-different-altitude" rule between #14 and #7 documented. | | P0 | **Studio.html** | "Lighthouse" provenance acknowledged in the methodology section. Even one line. | | P1 | **Team Skills Roadmap** (queued from #02) | Skills routing per loop position (in / above). | | P1 | **Operations Manual** | Apprenticeship Preservation note for engagements with deep junior benches. | | P2 | **CEO outreach pack** (queued from #03) | Open with stat #14 for cross-functional audiences; reserve stat #7 for marketing-functional. | --- ## Strategic posture — the canon order is now correct Reading the canon in numbered order tells the right story: - **#00 (Foundation)** — *Why* Studio exists. The paper that named "lighthouse." The paradox in plain English. *Workflow > tool* introduced. - **#01 (Tactical)** — *How* to build the agents well. Six Lessons from 50+ builds. Eval rigor. - **#02 (Functional)** — *Where* to apply it first inside marketing. 5-step process. 6 archetypes. - **#03 (Transformational)** — *What kind of org* this all adds up to. 12 themes. The Path-B validation. Every entry from #01 onward is a downstream specialization or systematization of #00. That's why the canon needed this entry: without it, the framework reads as borrowed-then-assembled. With it, the canon reads as *grounded in the conversation that made Studio possible*. --- ## How to use this doc going forward 1. **Origin attribution is non-negotiable.** When the deck or framework intro talks about "lighthouse use cases," cite Krivkovich. Studio's credibility comes from honest lineage, not from claiming originality. 2. **Don't mix stats #7 and #14.** Same finding, different altitudes. Pick one. 3. **In-the-loop / above-the-loop is now the framework's vocabulary.** Update Operations Manual deliverables to use both terms with their distinct skill profiles. 4. **The perpetual-state line is the right Handoff closing tone** — not "you're done," but "you've started a workstream that continues." 5. **Path B is locked.** This paper validates that decision. If anyone proposes drifting to A or C, point them here. — Synthesized May 3, 2026 (foundational entry, post-fact).