In 2025, the energy landscape faces a crossroads: rising demand coupled with strained supply.
The levers are moving in opposite directions. Grid Strategies forecasts a staggering 128 gigawatts of additional load growth in the U.S. over the next five years, largely driven by AI and electrification. That’s like adding a new Texas grid in half a decade. At the same time, greenfield development of new projects can take five to 13 years. The math ain’t mathing.
These compounding factors will create significant headaches and complexity for energy buyers across the country. And as a result, businesses have to start thinking about energy scarcity for the first time in decades. Energy management will become central to managing the bottom line.
But what does that look like in practice? In my experience, a company’s energy strategy comes down to three competing priorities, often in this order:
- How do I buy the cheapest energy?
- How do I ensure my next kilowatt-hour is secured?
- And how do I source the cleanest energy?
The cheapest energy
Of course, no business wants to overpay for energy. But ensuring you’re buying the cheapest energy and hedging your costs requires a proactive approach — not just taking whatever the utility offers.
Too many businesses today rely on a single tactic for energy management, whether it’s buying VPPAs, installing solar, or locking in long-term wholesale supply contracts. To secure the most cost-effective energy, however, businesses need a diverse strategy that looks at every avenue for savings by blending onsite and offsite power generation with tariff and demand charge optimization.
It’s clear that declines in the cost of onsite energy production offer new opportunities for savings, but businesses also need tools to optimize for the rapid rise of dynamic pricing. Price signals and dynamic pricing have exploded in the last decade and will continue to do so. A commodity that disappears in milliseconds should not be priced the same every hour of every day — especially with more and more intermittent renewables being added to the grid.
Arcadia’s Signal database tracks every rate in the U.S., and over the past 5 years, we’ve seen a 400% increase in commercial time-of-use rates offered by utilities.
And nearly half of those rates have a 2x arbitrage opportunity — i.e. the difference between the most and least expensive rate — that can yield significant value from load shifting and demand management. For example, one AI hyperscaler we work with saved nearly $1 million a year in a single data center by understanding complex industrial tariff options and optimizing their demand around the times when power was the cheapest.
The next kilowatt-hour
Securing reliable energy supply is increasingly a top priority. Some of our commercial customers at Arcadia are deeply concerned that their utilities won’t be able to keep up with their business growth.
As evidenced by last week’s announcement of “Stargate” — a joint venture between OpenAI, Oracle, and SoftBank to finance and build the world’s largest data center supercomputers, which could ultimately expand to $500 billion — AI demand is moving faster than ever. To power more than 100 GW of new capacity will require new grid infrastructure, copper, critical minerals, and labor, while further straining supply chains. The threat of hyperinflation in energy prices is real and could get worse.
To hedge against this risk and ensure future supply, businesses must diversify their energy sources and make long-term commitments. This means adopting a mix of behind-the-meter generation, virtual-net metering programs like community solar, and long-term power purchase agreements.
These decisions are not simple, and require unified energy data and complex engagement from finance teams to understand risk and take a view on wholesale and retail costs over time.
For example, if a cold storage company wants to build a new facility, should they look to onsite generation to lock in costs for decades? Or contract for a five-, 10-, or 20-year wholesale supply agreement that locks in costs, even if it’s a slight premium to today’s price?
And could fixed energy costs ultimately be an advantage against their competitors who can’t secure the next kilowatt-hour or pay hyper-inflated prices?
Every growing business will realize in the next decade how crucial managing electricity costs is to their bottom line — and they need a long-term, and comprehensive strategy to ensure they can continue to grow in an energy-scarce and volatile market.
The cleanest energy
Historically, pollution has been tracked but not taxed. And this regulatory gap has allowed many industries to delay actually addressing their environmental impacts. It was clear from the attendance at Trump’s inauguration that Big Tech cares more about the next kilowatt-hour to build AI, than whether that AI runs on clean energy.
While sourcing clean energy has unfortunately moved down the priority list for some businesses, it remains intimately and technologically tied to both cost and reliability. Solar, storage, virtual power plants, and community solar can help fix costs and provide solutions that can be deployed within one to two years — versus alternatives like new combined-cycle gas turbines or nuclear, which can take close to a decade to deploy.
The new administration is not wrong to predict a looming energy crisis, and we should applaud any industrial policy that creates energy abundance and dominance. But attempting to address this crisis without prioritizing decarbonization would be a serious mistake.
Accelerating clean energy abundance is essential — not only to combat climate change, which is imperative, but also to avoid falling further behind China, a nation already decades ahead in this space. Energy buyers face a daunting future, but the ones who prioritize the cleanest kilowatt-hour on par with cost and reliability will ultimately be winners over the long term.
Kiran Bhatraju is the CEO and founder of Arcadia, a global utility data and energy solutions platform. The opinions represented in this contributed article are solely those of the author, and do not reflect the views of Latitude Media or any of its staff.
