How to turn the data management problem into a reputation management opportunity

Last week I was joined for a panel discussion at Verge by Greg Mathe, Director, Electric Transmission Engagement, Permitting & Land for Dominion Energy, and Sarah Martinez, Vice President, ESG for Prime Data Centers, on the topic in the headline of this post. The discussion generated a lot of good insight, and I’m sharing it here. But first, let’s start with some context: 

There’s been a lot of news lately about the explosion of growth in data centers and, specifically, the amount of power needed to run them. (You can find a couple of examples here and here.)  

The data centers are needed to fuel the Artificial Intelligence (AI) boom we’re experiencing, and AI indeed uses a lot of power. (According to Goldman Sachs, one ChatGPT query uses as much electricity as it takes to power a light bulb for 20 minutes).  

This poses some real reputational threats – and opportunities – for both the tech and utility industries. Let’s start with the threats. At ERM Shelton, we have been surveying people in the US for 18 years about their beliefs and expectations related to energy. What we’ve seen consistently is: 

• People don’t know where their power comes from: For almost two decades now, we’ve never seen their understanding of how electricity is generated – and the fuel sources involved – match reality. They always think there is far more renewable energy on the grid than there actually is. And they expect their electric utilities to produce the bulk of their power via renewables. (PS: They don’t think of nuclear as renewable; they think of nuclear as scary and dangerous). 
• And while they’re very concerned about climate change, only 7% of people in America name “the energy we use in our homes and buildings” as the leading cause of climate change. In fact, electricity generation is now the number two driver for climate change, thanks to the industry’s decarbonization efforts. But for many years electricity generation was the number one driver, and we’ve seen awareness of that fact sit at about 7% for 18 years. 
• They’re kind of “meh” about their utility. Only half of people in America are satisfied with their electric utility. And what drives satisfaction – and concern about energy overall – is their monthly utility bill and reliability. 
• 66% of people in America say their monthly utility bills have gone up over the past year. That’s compared to 55% who answered in the same way back in 2021. 

Here’s why all of that matters as it relates to data centers and AI: New generation will have to be built to power all the data centers coming online. That will cause utility bills to go up (somebody has to pay for it), which will cause people to get mad – they won’t realize that it’s their queries to ChatGPT or views of funny AI-enhanced cat videos that are driving up demand. And, as Greg noted during our discussion, data centers require “always on” power, which solar and wind don’t provide. Also, as utilities increasingly look to nuclear as the low carbon, reliable option, people will get freaked out and protest that it can’t happen in their backyard. And, finally, if utilities aren’t able to supply the demand – which can sometimes happen when it’s really cold or really hot and everyone fires up their HVAC systems – we could see the kind of catastrophic blackouts that happened in Texas a few years back. 

These all represent reputational challenges that are worth getting in front of. 

The Tech industry has challenges as well. Despite the Edelman Trust Barometer noting that 76% of people trust the tech industry, both the Brookings Institute and the Public Affairs Council/Morning Consult noted a big drop in trust for tech companies during 2018-2021. In ERM Shelton’s ongoing tracking of consumer sentiment about which companies are deemed to be “good” and which are thought to be “bad,” and why, all the major tech companies make both lists. And the “why” behind those thought to be bad have to do with ESG-related reasons, like treating employees poorly, harmfully impacting youth, spreading misinformation, etc. 

As utilities and Big Tech increasingly get blamed for higher costs and, possibly, decreased reliability – while fighting the good fight to ensure the US wins the data center “arms race,” and, hopefully, not abandoning their work to lower their environmental footprint – how should they be communicating? Three insights came out of our discussion: 

1. Words matter: The easiest way to overcome concerns about nuclear is to talk about it differently, using terminology like “new nuclear” and “next-generation nuclear.” That immediately signals to the listener that there’s something different about the kind of nuclear we’re now talking about. If they hear that enough, they may be able to hear what’s actually different about it and then be open to hearing words like “carbon-free energy,” which nuclear provides. Both the tech industry and utility industry should align on this way of talking about today’s nuclear; and use the terminology consistently and without fail so the media picks up on it and uses it, too. Right now, the media just uses the term “nuclear,” which conjures up imagery of scary incidents in the past (especially now that Microsoft and Constellation are re-opening Three Mile Island.) 
2. Show up! More and more communities are banding together to block data centers from coming to their neck of the woods. On the one hand, it’s understandable…they don’t create as many jobs as, say, a manufacturing plant; there’s the construction traffic to deal with; and, again, the increased demand for energy (and sometimes also water). On the other hand, as Sarah pointed out, once the construction is done, you don’t have big trucks coming through communities as you would with a manufacturing plant or distribution center. And data centers generate a lot of tax revenue for the communities where they’re located. When people are upset, you can’t just tell them they’ve got it all wrong and expect them to listen. You have to show up in those communities, and really understand their concerns so you can find common ground and a way forward. 
3. It’s not a binary choice. Nor should we be talking about it that way. Almost every major utility and tech company has net zero goals. As Greg pointed out, “You don’t walk away from your goals just because you have a challenge.” Yet, the story of data/AI is often about how these industries will have to abandon their climate goals in order to meet energy demand. Instead, both industries should talk about all the ways in which they’re working to achieve both. 

If the utility industry and tech industry will work together proactively to show up, use consistent language, and paint a picture of how we can have all the computing power we need while still tackling climate change, both industries will see a reputational boost that will help them through the inevitably rocky waters ahead.