What Bloom Energy's Power Connect Really Fixes Is the Constraint the Backlog Hides

Ask anyone why Bloom Energy is a story and you'll get the same answer: the backlog. The company only prints a dollar figure for it once a year, and the figure it printed at the start of 2026 was $20 billion — roughly $6 billion of new fuel-cell hardware and about $14 billion of the service contracts that ride along with every sale. Backlog, the bulls will tell you, is growing faster than revenue. It's a big, round, impressive number. It is also, increasingly, the least interesting thing about the company.

The reason has to do with what backlog actually is. A backlog is a promise to build and install fuel cells. It is not revenue, and it is not cash. For a company that doubles its revenue in a year, the operative question is not whether customers want the machines. It's whether Bloom can actually get them installed — fast enough, cheap enough, and with skilled labor that data-center construction is busy consuming. That delivery problem, not demand, is the true bottleneck. And it is the thing the company's new Power Connect system is aimed at.

The constraint nobody quotes

Reading the second quarter makes the demand side of that story obvious. Bloom reported record revenue of $1.065 billion, up 165% from a year earlier, on product revenue that tripled to $935 million. Gross margin climbed to 33.4% from 26.7%, operating income swung from a small loss to $182 million, and the company recorded $196 million of net income and $226 million of operating cash flow — a swing of more than $400 million from a year ago, when the quarter burned $213 million. Management raised full-year revenue guidance to a $3.9–4.2 billion range, roughly a doubling. This is a company that just turned durably, measurably profitable for the first time after years of losses.

Yet here is the detail that frames the whole investment case: all of that hardware has to reach a site, get wired, get commissioned, and start producing electrons. Fuel cells have historically been treated as complex, bespoke electrical projects — long install campaigns, scarce field electricians, project risk that lands on the schedule. In an environment where data-center operators are fighting over power and speed-to-power, an install that drags is not just a cost; it is the constraint on converting that $20 billion promise into revenue and margin.

Why the factory beats the field

That is the wiring problem Power Connect, announced in mid-August, is meant to solve. The system moves critical electrical integration out of the construction site and into the factory, shipping units pre-connected, pre-wired and tested before they arrive onsite. Bloom says the standardized design can cut onsite installation time by more than 40%, and that concentrating complexity in a controlled manufacturing step relieves the shortage of skilled electrical trades — letting the available electricians spend their hours where they add the most value. It is, in effect, a shift from custom field assembly to repeatable, manufactured performance.

Dismiss this as an engineering nicety and you'll miss what it does to the model. Installation is a real component of cost of goods sold and a real consumer of working capital. Cut install time by 40% and you cut install cost per megawatt, raise the ceiling on how much throughput two thousand employees' factories and crews can push out each year, and shorten the gap between a signed order and the moment cash starts coming in. Against a balance sheet carrying $3.99 billion of debt and a business whose cash flow only recently turned positive, that is exactly the operating leverage a fuel-cell roll-out needs. It is the mechanism inside the backlog that actually produces cash — which is why, in my opinion, it matters more than the number everyone already quotes.

The price already knows the story

Which brings up the uncomfortable part, because none of this is free. Bloom sells for around $258, with a market capitalization near $76 billion. Against trailing revenue that's roughly 24 times sales and an earnings multiple north of 300; even against the raised forward guidance of roughly $4 billion, the stock trades near 19 times next year's revenue. There is no dividend here. The entire thesis is growth, executed at a full price, after a year in which the shares have already risen on the order of 200% and roughly 10x from their 52-week low.

The honest read is that Power Connect is the right lever on the right constraint, and one the current margin trajectory supports. But the 40% figure is a stated expectation, not a demonstrated result, and the margin gain that would justify the valuation is not yet in the numbers. What would change my conclusion is measurable: watch gross margin and, more importantly, free cash flow conversion across the next few quarters. If Power Connect shows up as faster installs, higher gross margin, and cash that keeps outgrowing revenue, the logic compounds. If throughput gains stall or the backlog converts more slowly than the sales line implies, then a stock priced for near-flawless execution has a long way to compress.

The cliché is that the market is buying a story and the backlog is proof. It's cleaner to say the backlog is the demand, and demand was never Bloom's problem. The problem is delivery — and that is precisely what Power Connect is built to fix. Whether it can do it fast enough to earn a price that already assumes it can is the question a holder, a watcher, and anyone chasing a 200% gainer is actually buying.