---
id: "framework-three-breakdowns"
type: "framework"
source_timestamps: ["\\\"§ Breakdown 1: Learning is informal", "while delivery is relentless.\\\"", "§ Breakdown 2: Incentives reward the wrong behaviors.", "§ Breakdown 3: Leaders and managers operate in different realities."]
tags: ["failure-modes", "organizational-analysis"]
related: ["concept-triple-burden", "action-adjust-incentives", "action-visible-leadership"]
speakers: ["Julia Shin", "Sandra J. Sucher"]
steps: ["Identify if learning time is protected or swallowed by delivery pressure.", "\\\"Audit incentive structures to see if AI-enabling behaviors (coaching", "sharing) are rewarded over pure utilization.\\\"", "Bridge the perception gap by having senior leaders engage directly in operational working sessions."]
sources: ["reskilling"]
sourceVaultSlug: "hbr-seg-reskilling"
originDay: 10
articleStem: "hbr-sig-50-adoption-overloading-managers"
sourceUrl: "https://hbr.org/2026/06/ai-adoption-is-overloading-your-middle-managers"
sourceTitle: "AI Adoption Is Overloading Your Middle Managers"
---
# The Three Breakdowns of AI Adoption

**The Three Breakdowns** are a diagnostic lens: three distinct ways the middle layer fails under the weight of AI adoption. Organizations struggling to realize AI value can use them to locate where they are breaking.

1. **Learning is informal, while delivery is relentless.** Time saved by AI is immediately swallowed by client work. Without protected learning time or a [[concept-centralized-internal-hub]], teams repeatedly solve the same problems in isolation. → Fixes: [[action-protect-learning-time]], [[action-build-centralized-hub]].
2. **Incentives reward the wrong behaviors.** Evaluation systems still reward billable hours and individual output (see [[prereq-consulting-business-model]]). Behaviors critical to AI scaling — sharing prompts, coaching, contributing to internal tools — go unrewarded, so managers default to measured metrics. This is why the [[concept-triple-burden]] is unsustainable. → Fix: [[action-adjust-incentives]].
3. **Leaders and managers operate in different realities.** Executives are highly optimistic while managers bear the operational friction. [[entity-bcg-d50|BCG]] survey data shows executives are roughly **twice as likely** as individual contributors to describe employees as enthusiastic about AI. Without firm-wide direction, managers make isolated calls on quality standards and client transparency (see [[question-client-transparency]]). → Fixes: [[action-visible-leadership]], [[action-train-ai-oversight]].

**Diagnostic steps.**
1. Identify whether learning time is protected or swallowed by delivery pressure.
2. Audit incentive structures to see if AI-enabling behaviors (coaching, sharing) are rewarded over pure utilization.
3. Bridge the perception gap by having senior leaders engage directly in operational working sessions.

**Enrichment context.** Each breakdown maps to independently documented tensions: Salesforce (managers lack protected time/training), the AI-resistance literature (legacy metrics reward the wrong things), and multiple surveys showing executive optimism outrunning frontline reality. The three-part framing is proprietary but conceptually sound.