Alphabet reports Google Cloud annual revenue increased to $13B, up 46% YoY, on losses of $5.6B, up 21%, in FY 2020
Context & Ripple Effects
This FY2020 disclosure put hard numbers on Google Cloud's scale-versus-burn trade-off for the first time at annual granularity: $13B in revenue growing faster than the ad business, funded by $5.6B of operating losses that were themselves growing 21%. The subsequent earnings trail shows the arc this report set up — by mid-2021 the quarterly loss had narrowed enough to move the stock, with shares up 3%+ after Alphabet reported a $591M Cloud operating loss in Q2.
The pivot point came two years later, when Google Cloud posted its first quarterly operating income of $266M versus a $440M loss a year earlier (Reuters, October 2023) — turning the loss line this article documents into a profitability milestone investors now track. By late 2025 the same segment was reporting $15.16B in a single quarter, up 34% YoY.
First-order effects
- Alphabet investors get their clearest view yet of what hypergrowth costs: every point of Cloud's 46% revenue growth arrived with a widening loss, making the segment the company's largest disclosed drag on operating income.
- Enterprise buyers reading the $13B figure see confirmation that Google Cloud has the revenue base and sales motion to be a durable third hyperscaler rather than an experiment.
Second-order effects
- Cloud's loss trajectory becomes a market-moving metric in its own right — the pattern where a narrowed loss lifts GOOG 3%+ shows investors pricing the segment on its path to breakeven, not just its growth rate.
- Sustained losses at this scale force the rest of Alphabet to subsidize infrastructure buildout from ad profits, raising internal stakes on whether Cloud can convert AI-era demand — like the TPU compute it reportedly sells to Anthropic — into margin.
Third-order effects
- If the 2020-to-2023 pattern holds — losses peaking, then flipping to operating income as revenue compounds — hyperscale cloud settles into a structure where only companies willing to absorb multi-year, multi-billion-dollar losses can compete at the top tier.
- The eventual payoff layer is already visible in later disclosures: a $514B contracted-work backlog and DeepMind's compute shifting toward the cloud business suggest the loss-making years bought the capacity that AI customers are now monetizing.
The trend: Hyperscale cloud units are moving from a buy-growth-at-any-loss phase to a profitability-discipline phase, with Alphabet's Google Cloud as the template case from its $5.6B FY2020 loss to sustained operating income.