Switching Costs of Installed SystemsNarrow moat

Bloom Energy (BE) — moat facet

Once sited, wired, and integrated, the system stays — switching costs paid in concrete and permits.

The switching costs attached to Bloom's installed systems are the source of the moat's stickiness, and they are genuine. Installing Energy Servers is a substantial commitment: the customer sites them, connects them to fuel and electrical infrastructure, integrates them into its facility's power and operations, and builds a service relationship around them. Having made that investment and adaptation, the customer faces real cost, disruption, and risk in ripping the systems out to switch to a competitor's solution — so, within the life of the installation, customers tend to stay with Bloom, continue paying for service, and, when systems need replacing or capacity needs expanding, often turn to Bloom again. This is a real switching-cost moat of the kind that makes revenue sticky.

Service gross margin by quarter (%)1.3%Q1 259.3%Q2 2511.6%Q3 2516.9%Q4 2513.3%Q1 2618.7%Q2 26Bloom Energy 10-Qs and 10-K; Q1 2025, Q4 2025 and Q1 2026 derived from period totals
Customers that cannot easily switch are now being serviced at a profit, rising almost every quarter since the start of 2025.

The stickiness is reinforced by the service relationship and the operational trust it builds: a customer whose Bloom systems have run reliably, and who values the established support relationship, has incentives to stay beyond the pure cost of switching. This is the mechanism that could, as the installed base grows, give Bloom a durable hold on a large base of customers. The qualifier is that switching costs in this business are strong within a system's life but fade at its boundaries: when a service contract comes up for renewal, or when a system reaches the end of its useful life and must be replaced anyway, the customer is free to re-evaluate the whole decision against whatever alternatives then exist — grid, turbines, batteries, rival fuel cells, or new technologies. So the lock-in is real and valuable but cyclical rather than permanent, holding customers through a system's life but reopening the competition at each renewal or replacement. The switching costs are one of the best things about Bloom's business and a genuine source of moat — but they are the kind that must be re-won at intervals, not the kind that hold a customer forever — a distinction that matters under a ~$20B backlog of future re-winnings1.

Moat trajectory: Holding steady

Stable. Once sited and integrated, systems are costly to rip out, so customers stay through a system's life — real stickiness, but cyclical (it reopens at each renewal/replacement), so it holds rather than widens.

The number that tests this moat
Moat Explorer calc
Service gross margin
10.0% in 2025, from −0.7%

Only Bloom can service its installed fleet, which should make service profitable. A margin that has only just turned positive says the annuity is not yet paying; a rising margin would show the switching cost being monetised.

How it's calculated: Service revenue less service cost of revenue, divided by service revenue: ($228.3M − $205.4M) ÷ $228.3M in 2025; ($213.5M − $215.0M) ÷ $213.5M in 2024.
Source: Bloom Energy Form 10-K, FY2025 ↗
⚠ Threats to the moat
References
  1. ReportedA ~$20B backlog of future re-winnings.
    Bloom Energy Q2 2026 earnings press release — record revenue $1.065B (+166%), product revenue +215% to $935M, non-GAAP gross margin 34.3%, non-GAAP EPS $0.78; FY2026 guidance raised to $3.9–4.2B revenue / $800–900M operating income / $2.55–2.85 non-GAAP EPS; total backlog ~$20B; Brookfield financing expanded $5B → $25B — Q2 2026 · publ. August 2026 · source ↗
Sources
Generated September 23, 2026