Pim de Morree's Post

Pim de Morree · article · 2026-09-12 Reader 🖥️

Enterprise AI Adoption & Readiness Hierarchy & Org Charts Outcome-Based Work Design (vs. Fixed Jobs)Team Autonomy

Bayer shows that when an organization is already mid-transformation, the better path is to fold AI into that existing change effort rather than launch AI as its own parallel initiative with separate leadership, budget, and governance.

practical · expository — A short LinkedIn teaser making a practical recommendation (fold AI into an existing transformation rather than run it separately), illustrated with a single unelaborated case (Bayer) and explicitly deferring its actual evidence and argument to a separate Thursday newsletter. It is closer to an abstract/trailer than a self-contained argumentative piece.

Structure

Part	Role
Opening anecdote: "Here we go again" as the predictable reaction to a new transformation announced on top of an existing one.	context / motivating problem
Description of Bayer's existing transformation: halved management layers, management positions down two-thirds, teams working in 90-day cycles and setting their own outcomes.	evidence / background
The stated choice: Bayer could have run AI as a separate initiative (own leader, budget, steering committee) but decided not to, folding it into the transformation already underway instead.	conclusion / thesis statement
Teaser reference to one concrete example from a phone call with Bayer's team, which the author says raises a bigger question: whether organizing people around fixed jobs still makes sense.	forward pointer / incomplete evidence
Sign-off tagline "Work is broken. Let's fix it."	brand framing, not argumentative

Key terms

90-day cycles Outcome-Based Work Design (vs. Fixed Jobs)
Bayer's operating rhythm in which teams work in bounded quarterly-ish periods and set their own outcomes within them, replacing (or displacing) a more static annual/role-based planning structure.
fixed jobs Outcome-Based Work Design (vs. Fixed Jobs)
Organizing people around stable, defined job roles/positions, as opposed to organizing around outcomes a team owns — the framing the author uses to gesture at Bayer's deeper structural bet, without stating what the alternative looks like.
transformation
The specific, already-underway de-layering and outcome-based restructuring effort at Bayer (not a generic label for "digital" or "AI" transformation) — the vehicle AI is being folded into rather than a separate thing AI itself constitutes.

Propositions

explicitBayer chose to fold its AI rollout into an already-underway organizational transformation rather than launch AI as a separate initiative with its own leader, budget, and steering committee.conf: high
explicitBayer's pre-existing transformation had already halved management layers, cut management positions by roughly two-thirds, and moved teams to 90-day cycles in which teams set their own outcomes.conf: high
refinesThe engineering manager role is splitting in two · p6 — That article claims most companies have not deliberately chosen a flattening model and are flattening as an unplanned byproduct of layoffs; Bayer is a counter-instance of an organization that deliberately designed de-layering and outcome ownership as a structural program, ahead of and independent of its AI rollout.
explicitAnnouncing a new change program on top of an existing one predictably produces resistance and fatigue ("here we go again"), and this is a normal, expected reaction rather than a sign of failure.conf: high
explicitcontestedA single concrete example from Bayer's AI-in-practice work surfaces a question bigger than AI itself: whether organizing people around fixed jobs still makes sense.conf: medium
contradicts ←The person your firm needs and cannot name · p8 — This doc's thesis is that you must formalize and name a role/title before you can hire for it; the Bayer piece raises the broader question of whether organizing people around fixed jobs/titles still makes sense at all — the two pull in opposite directions on the value of titling.
implicitAn organization's readiness to absorb AI without a dedicated AI-transformation track depends on already having a flexible, outcome-owning operating model (de-layered management, self-set team outcomes) rather than on AI-specific governance or technology maturity.conf: medium
extendsHow Fast-Growing Companies Can Make Better Decisions · p6 — Bayer's practice of teams setting and being accountable for their own outcomes in 90-day cycles is a concrete, at-scale instance of the general prescription that employees should be assessed on outcomes rather than on process/role compliance.

Arguments

Organizations already undergoing structural transformation should fold AI adoption into that existing transformation rather than stand up a separate AI initiative. strength: low

Explicit premises:

A second, separately-branded change program on top of an existing one triggers predictable resistance ("here we go again").
Bayer already had management layers halved, positions cut, and teams operating in self-set 90-day outcome cycles.
Bayer chose to integrate AI into that existing structure instead of launching a parallel AI transformation with its own leadership and budget.

Implicit premises:

An organization's capacity to absorb change is finite, so minimizing the number of distinct change vehicles reduces total resistance.
An operating model already built around team-level outcome ownership and short cycles is flexible enough to absorb a new technology without requiring its own dedicated governance structure.
The benefits of avoiding a duplicate initiative (leadership, budget, steering committee) outweigh whatever AI-specific oversight a dedicated AI transformation track would have provided.

Inferences:

If outcome-owning, de-layered operating models are what make folding AI in possible, organizations that have NOT already restructured this way may not be able to simply fold AI into "business as usual" the same way Bayer did — the de-layering may be a precondition, not incidental context.
Treating AI governance, security, and risk oversight as fully absorbable into a general transformation structure (rather than warranting some AI-specific oversight) is an assumption the post never tests.

Narratives

Bayer, a German life-sciences company with 85,000+ employees, already mid-transformation (halved management layers, positions down two-thirds, 90-day self-set-outcome cycles), decided to fold its AI rollout into that transformation instead of running a separate AI initiative.
→ Serves as the sole evidentiary support for the unity claim — it is the case that is supposed to show folding-AI-in works, though the mechanism of how is not actually shown here.
A specific, unnamed example from a phone call with Bayer's team, which the author says gets into whether organizing around fixed jobs still makes sense — explicitly withheld and deferred to Thursday's newsletter.
→ Functions as a narrative stub: it is invoked to establish stakes ("something much bigger than AI") without being told, so its persuasive work is almost entirely promissory.

Questions addressed

✅ Should an organization already undergoing a structural transformation run AI adoption as its own separate initiative, or fold it into the existing transformation?
Fold it into the existing transformation — illustrated by Bayer's decision not to launch a separate AI transformation with its own leader, budget, and steering committee.
⭕ Does organizing people around fixed jobs still make sense, given how AI and outcome-based team structures interact in practice?
Explicitly flagged as an open, bigger-than-AI question, with the answer deferred to a separate newsletter rather than resolved here.
⭕ What concretely changed at Bayer when AI was folded into the 90-day-cycle model (tooling, workflows, team composition, headcount effects)?

Gaps

incompleteThe post withholds the one concrete example it invokes as proof of stakes ("gets into something much bigger than AI") — the actual mechanics of how Bayer folded AI into its 90-day-cycle model are entirely absent, deferred to a separate Thursday newsletter. As written, this is a teaser for an argument rather than the argument itself.
Read next: Read the referenced Thursday newsletter (linked via https://lnkd.in/emN_R95n) for the actual case detail and the fixed-jobs argument.
incompleteNo evidence is given for the claim that Bayer's de-layered, outcome-owning structure is what made folding AI in (rather than running a separate initiative) work — no comparison case, no account of friction encountered, no measure of success. The central causal mechanism is asserted, not shown.
Read next: Case studies with before/after evidence on AI adoption inside already-restructured (de-layered, outcome-based) organizations versus AI adoption bolted onto conventional hierarchies, to test whether operating-model maturity is actually the enabling variable.
incompleteThe post never addresses whether AI adoption has governance, security, or risk needs distinct enough from a general organizational transformation that folding it in without dedicated oversight could create blind spots (e.g. model choice, data handling, security review) that a separate AI initiative would have had a mandate to own.
Read next: Material on AI-specific governance/guardrails requirements versus general change-management practice, to test whether "no separate AI initiative" trades away necessary AI-specific oversight.
incompleteThe "fixed jobs" question — whether organizing people around fixed jobs still makes sense — is raised as the post's biggest idea but given zero content: no alternative structure is named, no criteria for when fixed jobs stop making sense, and no acknowledgment of what is lost (career ladders, accountability clarity, skill specialization) if jobs are replaced by rotating outcome ownership.
Read next: Literature on outcome-based/results-only work design (e.g. ROWE) and on the trade-offs of dissolving fixed roles in favor of team-owned outcomes, plus this corpus's existing material on team autonomy and structured empowerment.