The supply chain landscape within the data centre industry is undergoing a profound transformation. A recent panel discussion chaired by data centre consultant Simon Allen at Datacloud Global Congress 2025 explored the multifaceted challenges of de-risking the supply chain amid rising complexity, geopolitical uncertainty, and market volatility. The conversation highlighted how digitalisation, sustainability, and resilience are shaping the future of supply chains in this sector.

Speakers
Simon Allen, consultant - Data Centres
Brittany Miller, global SVP infrastructure development - NTT Global Data Centers
Michael Murray, group business development / power & renewables director - Kirby
Todd Zabelle, CEO - SPS
Mike Coleman, head of global data centre infrastructure & services - Bytedance
Diarmid Massey, CEO, data centres - ESR

Allen opened the discussion by framing the problem: supply chains today are “under immense pressure,” enmeshed in a web of “geopolitical shifts, tariffs, market volatility,” all while the industry is being asked to deliver faster and more sustainably. This dynamic environment demands practical, strategic, and sometimes uncomfortable conversations.

Brittany Miller, global SVP, infrastructure development at NTT Global Data Centers, reflected on how COVID-19 accelerated supply chain maturity. She explained that the pandemic forced suppliers to diversify sources globally, moving away from risky single-source dependencies. This evolution, she suggested, had a silver lining: “COVID actually helped us prepare to warm up for this exponential growth.”

The complexity of supply chains is compounded by regional nuances and varying standards, making global coordination a major challenge. Diarmid Massey, CEO of data centres at ESR, elaborated on this point by noting the difficulty in standardising infrastructure across diverse markets.

He said: “you don't get there without standardisation… but that’s also about customers having a standard build and that's just not necessarily going to happen in the short term.” He pointed out further obstacles such as differences in electrical standards and regulatory frameworks, explaining, “If I ship a generator from America to Japan, will the Japanese installer actually install it? And the answer is no.”

When asked to define efficiency, Mike Coleman, head of global data centre infrastructure & services at Bytedance, highlighted the volatility in current supply chains.

He stressed adaptability over rigid planning: “Being able to be flexible and adjust to the changes on a regular basis, adapt… is really the best way you can succeed.” His emphasis on flexibility was echoed by Massey, who noted that timely delivery, price, and reliability remain the core markers of efficiency in today’s climate.

Michael Murray, group business development and power & renewables director at Kirby, focused on the role of trust and collaboration.

He underscored “program certainty” through “early contractor involvement” (ECI) as a critical factor for efficient delivery. However, he also cautioned that early involvement has its barriers: “If you don't have your client, maybe power is still not there, planning might be upended.”

Todd Zabelle, CEO of SPS, summarised key supply chain attributes as reliability, transparency, flexibility, redundancy, and cost efficiency.

He drew lessons from the automotive industry, citing Toyota’s model of 80% reliance on one supplier and 20% on a secondary one to ensure redundancy. Zabelle explained how these suppliers compete continuously to increase their share, incentivising innovation and resilience.

The panel delved into the strain global networks place on supply chain visibility. Coleman observed a paradox where standardisation, intended to simplify, often exacerbates bottlenecks: “Standardisation… is actually kind of our biggest obstacle.” He described how reliance on a few suppliers for standard components creates single points of failure.

Massey added to this by highlighting regulatory and infrastructure inconsistencies across markets. These differences limit the fungibility of products across regions, posing logistical and operational challenges.

Allen pressed on tools available to improve visibility. Miller pointed to a notable lag in technology adoption within the supply chain sector.

Despite building data centres that power AI and advanced software, she noted the industry still relies heavily on outdated tracking methods: “I watch some people, we're still tracking supply chain things on Excel spreadsheets.”

She suggested a move towards AI-driven predictive tools to enhance supply chain transparency and responsiveness, emphasising that “being really embedded in your supply chain” is key to gaining predictability.

Zabelle extended this point, advocating for the development of “digital twins” of supply chains — virtual models that simulate scenarios and provide real-time data flows. He envisaged a future where supply chains evolve into “value chains,” with integrated tracking and analytics that move beyond mere logistics to strategic asset management.

Forecasting emerged as a particularly contentious topic. Zabelle was blunt: “Anything that has to do with the word forecast… is wrong.”

He argued that the industry should prioritise flexibility and agility over attempts at precise long-term forecasting, given the rapid shifts in demand and market dynamics. Instead, he suggested leveraging real-time data and adaptive models to navigate uncertainty.

Miller agreed on the need for flexibility but also stressed the importance of sharing long-term forecasts with suppliers to facilitate capacity planning: “We do share forecasts five years out… they’re probably wrong, but… that helps us tremendously.” She highlighted the shift from transactional to strategic partnerships, noting lessons learned from the semiconductor industry’s mature supply chain practices.

Murray described demand planning as “disjointed,” complicated by factors such as tenant uncertainty at the colocation level and the sheer scale of hyperscale growth.

He raised concerns about the balance sheets required to support the increasingly large projects being undertaken and implied that not all players may be able to meet these demands.

Panelists pointed to other industries for valuable lessons. Zabelle cited automotive and semiconductor sectors as benchmarks, given their advanced integration and innovation in supply chain management.

He detailed a story from Ford where workers from unrelated plants flew in to support a critical supplier outage, illustrating the deep integration and mutual support within automotive supply chains.

Miller emphasised how semiconductor supply chains, which have navigated crises and complexity for decades, offer valuable insights for the data centre sector, particularly regarding embedded partnerships and advanced supply chain technologies.

Although the panel did not directly focus on insolvencies, the underlying theme of risk mitigation permeated their discussion.

Zabelle’s emphasis on redundancy and flexibility clearly addresses the risk of cascading supplier failures. Murray’s comments on capacity slots being snapped up by larger clients highlight the vulnerability smaller players face, a potential catalyst for insolvency risks within the supply chain.

Miller’s call for deeper integration and real-time data sharing can be read as a strategy to mitigate domino effects by anticipating disruptions early and adjusting swiftly.

The panel at Datacloud Global Congress 2025 painted a complex picture of the data centre supply chain: a landscape marked by volatility, complexity, and the urgent need for innovation.

While challenges around standardisation, visibility, and forecasting remain formidable, the collective wisdom pointed to several critical enablers of resilience:

• Embedded partnerships across the supply ecosystem, moving beyond transactional relationships to strategic collaboration (Miller, Murray).
• Flexibility and agility to respond rapidly to changing market demands and disruptions, recognising that traditional long-term forecasts often fail (Zabelle, Coleman).
• Digital transformation, including adoption of AI tools and digital twins, to improve supply chain transparency and decision-making (Miller, Zabelle).
• Learning from other industries, notably automotive and semiconductor, to enhance redundancy, integration, and value chain thinking (Zabelle, Miller).

As Simon Allen concluded at the outset, the data centre industry must navigate this “web of complexity” with innovative, resilient supply chain models.