type: pattern tags: [backlog, revenue-acceleration, cloud, enterprise, leading-indicator, growth-rate-differential] confidence: medium created: 2026-04-01 source: AMZN earnings-review Q4_FY25 persona: atlas provenance: legacy source_analysis_path: null source_paragraph_quote: null source_transcript_span: null source_loss_log_path: null

Backlog Growing Faster Than Revenue Signals Upcoming Revenue Acceleration

When an enterprise or cloud company's contracted backlog grows materially faster than its current revenue (ratio ≥1.5x), this divergence is a leading indicator that the revenue growth rate will increase in forward quarters — not just that revenue will grow. The mechanism: backlog accumulation precedes conversion; a widening backlog-to-revenue growth spread means the pipeline is filling faster than it can be delivered, creating a conversion tailwind that pulls the revenue growth rate higher as capacity comes online. This is distinct from the coverage ratio signal (backlog as a multiple of annual revenue), which addresses downside risk; the growth-rate-differential signal addresses forward acceleration.

Evidence

• AMZN Q4 FY25: AWS contracted backlog 244B, growing + 401.5B → +1.6B → +2.1B → +$2.6B over the last four quarters.
• Management framing: capacity "monetized as fast as we install it" — demand is backlog-limited, not market-limited. The backlog growth rate therefore directly proxies demand growth, which leads revenue by the average backlog conversion lag.
• SNOW Q4 FY25: RPO (Remaining Performance Obligation) grew +42% YoY vs. total revenue +30% YoY — a growth ratio of 1.40x, sustaining for at least two consecutive acceleration quarters. RPO/TTM revenue = 2.08x (over 2 years of contracted backlog). Revenue had troughed at 25.8% in Q1 FY25 and re-accelerated to 30.1% by Q4 FY25. Critically, SNOW uses a consumption-based model — RPO represents contracted minimums, not ceilings. Actual revenue realized from a given cohort regularly exceeds the contracted floor because customers consume more than their committed minimums when the platform delivers value. This means the divergence signal in consumption models is stronger per unit of growth differential than in committed-contract models: the backlog understates likely revenue conversion, so the forward acceleration signal is at least as strong at 1.4x growth ratio as it would be at 1.5x in a pure committed-contract model.

Implication

For any cloud, infrastructure, or enterprise software company disclosing a contracted backlog: (1) compute both the absolute coverage ratio (existing signal) AND the YoY growth rate of the backlog; (2) if backlog growth ≥1.5x revenue growth for 2+ consecutive quarters, treat this as a forward acceleration signal with 2-4 quarter lag; (3) pair with capacity data — if constrained by supply (data center capacity, headcount), the conversion lag extends and the signal weakens; if demand-constrained, backlog growth directly predicts revenue acceleration; (4) apply a lower threshold for consumption-based models — because RPO/backlog is a contracted floor and actual revenue regularly exceeds it, a 1.3-1.4x growth ratio in a consumption model is analytically equivalent to a 1.5x ratio in a committed-contract model. A widening backlog-growth-to-revenue-growth spread is one of the cleaner leading indicators available in enterprise/cloud reporting.