The Burden of Proof: Evaluating Energy and Industrial Solutions in the AI Era

AI is impacting every industry and reshaping how companies at various stages think about growth, defensibility and differentiation from competition. For energy and industrial sectors in particular, power demand growth is driving increased urgency for better speed, efficiency and accuracy across operations. This is an area where AI and other digital tools can help substantially, but the nature of these industries demands a slower, more controlled approach to adoption. At Energize, we believe successful innovation in these markets requires a thoughtful balance of emerging AI tools and trusted software partners - pairing industry context and expertise together with the latest agentic solutions to deliver outcome- and value-based results.

Companies operating in complex, highly regulated industries like energy and industrials have a high threshold for implementation of novel technologies. These companies manage critical infrastructure with very low tolerance for failure or downtime. Given this complexity, energy and industrial operators are less likely to quickly jump onto the AI bandwagon with brand new tools, instead preferring to lean into existing partners with industry expertise to help guide their adoption strategy. There is a clear and urgent need for AI solutions that can improve efficiency, accuracy and safety in today's power-hungry energy and industrial environment, and there is a significant opportunity for trusted software partners to be a part of that solution.

Evaluating both emergent AI and traditional software technologies in this rapidly changing environment is a challenge for operators and investors alike. Certain attributes that used to drive competitive advantage - like data access and management, for example - are now being disrupted by AI solutions. In other cases, established moats are sharpened by layering agentic tools on top of them. To help find clarity in this complex landscape, Energize has developed a simple framework for evaluating solutions - both traditional SaaS and AI-native - to help determine those with greater durability and commercial staying power.

Burdens of proof for defensible software and disruptive AI solutions

After reviewing and evaluating thousands of companies operating in the energy and industrials landscape, we've developed a set of criteria for assessing businesses based on their perceived exposure to AI displacement risks. The majority of companies sit somewhere on a spectrum across these various criteria - some elements of their technology may provide a strong moat, whereas others may carry more risk exposures. Our belief is that these parameters are helpful for evaluating companies based on where the technology sits today.

• Control vs. influence: Does the company's product control critical workflows as an embedded system, or only provide advice or analytics?
• Data advantage: Is the company's data proprietary and difficult to reach, or publicly available and now easily scraped by AI tools?
• Cost of failure: When the cost of an error is high, has it earned the trust to be relied on?
• Influence on outcomes: Can the company price based on value created (outcomes), or is the per-seat model too locked in?

Both traditional SaaS and AI-native platforms are in a race against time to prove value to customers. AI has set a new bar in terms of customer expectations around digital products: users both know they have at least some capability to build versus buy, and they also know that the software providers have AI tools at their own disposal, and generally expect more tailored, custom solutions than before.

For AI-native solutions, the race is to quickly earn and build upon customer trust. They start with a wedge, helping to automate and streamline critical workflows, and then can grow and expand coverage from there. Software platforms must maintain trust, proving they can leverage their existing data to generate unique insights and layer on AI products that are differentiated from those coming to market.

Key moats and risks in today's AI landscape

As we assess companies across this spectrum of criteria, several patterns have emerged regarding where critical moats and risks arise.

Key Moats:

• System of record over physical assets: A platform that holds a customer's asset data, and is trusted to sit on the equipment that generates it, gets the first crack at building AI on top. It moves from a system of record to a system of action - storing the data, then recommending what to do with it, then doing it. For example, a platform that can predict permitting delays by jurisdiction from patterns across its customers offers something that a single operator or general model couldn't assemble.
• Domain expertise and earned trust: In energy and industrial industries, where decisions are often high-stakes and deterministic, probabilistic AI is not good enough. These solutions need to be paired with trusted experts that can convert 80 to 95% accuracy to 99.999% accuracy in order to access and automate critical customer workflows.
• Proprietary operating data: A model cannot reach data that lives behind a meter, sensor or a decade of operations. Access to, and the ability to effectively aggregate, proprietary data sets gives software and AI solutions an edge. These datasets are often either captured over many years or through privileged operational access through trusted customer relationships.
• Outcomes-based pricing: AI often helps fewer people do the same work, in less time. When one individual can complete an entire teams' worth of tasks, this shrinks the number of seats needed for a technology platform. Solutions that can attach their pricing - and their revenue - to value of outcomes are better poised for sustainable growth in the AI era.

Key Risks:

• Replicable workflows: Rules-based, replicable workflows are very easy for AI to embed into and automate. Maintaining automations once had value, but agents are now making this action trivial.
• Public data: Solutions that could cull public data and draw actionable insights from that data were once differentiated, but now face steep competition from AI solutions that can easily reproduce this type of work. Publicly accessible data, no matter how fragmented and high-volume, is now a prime target for AI disruption.
• Seat-based pricing: The number of seats using a tool is drastically being reduced by AI. Solutions that cannot determine how to tie their value directly to customer outcomes (dollars or time saved, or accidents avoided, for example) will lose out on revenue opportunities and struggle to grow as AI solutions dominate.
• Team inertia: One of the biggest risks we're seeing for businesses operating in the AI era is less about the product or data set and more about the willingness to adopt the novel technologies. To survive and thrive in this environment, teams must be willing to endure discomfort and choose to aggressively build new products and deploy AI solutions rather than resting on existing moats. We've found that this mindset must be top-down, with senior executives championing change and driving innovation through the organization.

Specialization underpins defensibility

AI technology is constantly and rapidly evolving, and these success criteria will undoubtedly evolve with it. Across our evaluation of key risks and differentiators, there is one element in particular we believe has certain staying power: specialization. Across the moat-risk spectrum, solutions that exhibit domain expertise and industry context appear to have the strongest foothold with energy and industrial customers amidst this massive technological shift. AI-native solutions that lack this domain expertise end up competing directly with the foundation model labs and hyperscalers, with little to differentiate them. And traditional software platforms without it risk losing the privileged data access and customer trust that gave them their edge in the first place. In both cases, specialization is what separates a durable business from one exposed to displacement. The combination of pre-existing trust and workflow embedment with key customers, privileged data access, and team willingness to experiment with and deploy agentic solutions are the strongest indicators of durability in our current landscape.