---
id: "framework-incumbent-energy-playbook"
type: "framework"
source_timestamps: ["§ The Incumbent's Energy Playbook", "¶10", "¶11", "¶12", "¶13", "¶14", "¶15"]
tags: ["strategy", "playbook", "execution"]
related: ["action-make-energy-visible", "action-reduce-demand", "action-contract-optionality", "action-redesign-compute-location", "action-create-compute-council", "claim-incumbents-need-energy-access"]
steps: ["1. Make energy intensity visible: track energy cost per workflow and 'intelligence per watt' via dashboards.", "\\\"2. Reduce demand before buying supply: route simple tasks to smaller models", "cache queries", "compress prompts", "and batch nonurgent inference.\\\"", "\\\"3. Contract for optionality", "not ownership: use VPPAs", "green tariffs", "and bilateral contracts to hedge regional power price spikes and interconnection delays.\\\"", "\\\"4. Redesign where compute runs: shift flexible workloads to cloud regions with cheaper", "cooler", "and lower-carbon power.\\\"", "5. Make someone accountable: establish a cross-functional Compute and Energy Council with veto power over major AI deployments."]
sources: ["futures"]
sourceVaultSlug: "hbr-seg-futures"
originDay: 2
articleStem: "hbr-nm-101-energy-strategy-ai"
sourceUrl: "https://hbr.org/2026/06/your-company-needs-an-energy-strategy-for-ais-next-phase"
sourceTitle: "Your Company Needs an Energy Strategy for AI’s Next Phase"
---
# The Incumbent's Energy Playbook

## What It Is
A five-step strategic framework for **non-hyperscaler enterprises** to build distinctiveness in energy access and manage the rising energy costs of AI adoption. It executes the premise of [[claim-incumbents-need-energy-access]]: differentiate on *how you access energy*, not by building power plants.

## The Five Steps
1. **Make energy intensity visible.** Track energy cost per workflow and [[concept-intelligence-per-watt]] via dashboards. → [[action-make-energy-visible]]
2. **Reduce demand before buying supply.** Route simple tasks to smaller models, cache queries, compress prompts, quantize models, and batch nonurgent inference. → [[action-reduce-demand]]
3. **Contract for optionality, not ownership.** Use VPPAs, green tariffs, and bilateral contracts to hedge against regional power price spikes and interconnection delays. → [[action-contract-optionality]]
4. **Redesign where compute runs.** Shift flexible ([[concept-shiftable-vs-latency-sensitive|shiftable]]) workloads to cloud regions with cheaper, cooler, and lower-carbon power. → [[action-redesign-compute-location]]
5. **Make someone accountable.** Establish a cross-functional **Compute and Energy Council** with veto power over major AI deployments. → [[action-create-compute-council]]

## Sequencing Logic
The order is deliberate: *see it → shrink it → hedge it → relocate it → govern it.* Visibility (Step 1) is prerequisite to everything; demand reduction (Step 2) is the cheapest lever and should precede any supply-side contracting (Step 3); governance (Step 5) locks the discipline in place so the first four steps survive organizational turnover.

## Counter-perspective
Smaller enterprises that primarily *buy* SaaS AI rather than run large in-house models may find a formal veto council too heavyweight — CIO–CFO coordination plus ESG oversight may suffice. See [[contrarian-energy-is-strategic]] for the governance argument and its limits.